Patterns of Adult Cross-Racial Friendships: A Context for Understanding Contemporary Race Relations
Abstract
Objectives
This study examined patterns, characteristics, and predictors of cross-racial friendships as the context for understanding contemporary race relations.
Methods
A national survey included 1,055 respondents, of whom 55% was white, 32% was black, and 74% was female; ages ranged from 18 to ≥ 65 years. Focus groups were conducted to assess societal and personal benefits. Participants (n=31) were racially diverse and aged 20–66 years.
Results
After accounting for multiple covariates, regression analysis revealed that Asians, Hispanics, and multiracial individuals are more likely than their white and black counterparts to have cross-racial friends. Females were less likely than males to have eight or more cross-racial friends. Regression analysis revealed that the depth of cross-racial friendships was greater for women than men and for those who shared more life experiences. Increasing age was associated with lower cross-racial friendship depth. Qualitative analysis of open-ended questions and focus group data established the social context as directly relevant to the number and depth of friendships. Despite the level of depth in cross-racial friendships, respondents described a general reluctance to discuss any racially charged societal events, such as police shootings of unarmed black men.
Conclusion
This study identified salient characteristics of individuals associated with cross-racial friendships and highlighted the influence of the social, historical, and political context in shaping such friendships. Our findings suggest that contemporary race relations reflect progress as well as polarization.
Despite increased racial integration in U.S. cities and in the workforce, Americans remain segregated in their social patterns (Hoenig, 2013; Nesbitt, 2012). Considerable research has demonstrated that the degree of intersectionality and the opportunity for contact between social positions (class, age, and education level) either constrain or facilitate social relations between racial ethnic groups. The opportunity for contact further depends on whether or not members of smaller groups live near or work with one another (Blau, 1977; Hallinan & Smith, 1985; Hallinan & Williams, 1989).
Standard segregation measures show that American cities are more integrated today than they have been since 1910, but spatial racism still exists (Glaeser & Vigdor, 2012; Williams & Collins, 2001). Spatial racism is a form of racism that depicts a pattern of housing development in which racially and economically segregated suburbs or gentrified areas of cities are created (George, 2001; Brokaw, 1997). Although racial diversity has increased in the workforce, coworkers experience limited contact with each other outside the work environment (Burns, Barton, & Kerby, 2012; McPherson, Smith-Lovin, & Brashears, 2006; Pettigrew & Troop, 2006; Dunsmuir, 2013; Toossi, 2012). The resulting cultural encapsulation may lead to ethnocentrism and limited understanding of another’s worldview (DiPrete, Gelman, McCormick, Teitler, & Zheng, 2011; McCubbin & Bennett, 2008; Wrenn, 1962).
A Reuters poll (Dunsmuir, 2013) found that white Americans are far less likely to have friends of another race than non-white Americans, with about 40% of white Americans having only white friends, while about 25% of non-white Americans is surrounded only by people of their own race. Another study using “egocentric network analysis” designed to assess the scope of diversity of Americans’ social networks found that Americans’ social networks are primarily comprised of people from the same racial or ethnic background especially for white Americans. Among white Americans, 91% of the people in their social networks were white. Among black Americans, 83% percent of people in their social networks were black. Among Hispanic Americans, approximately two-thirds (64%) of the people in their social networks were also Hispanic (Pubic Religion Research Center (PRRC), 2013).
Despite increased opportunity to live and work in mixed-race settings, Americans remain racially segregated in their friendship patterns. Spatial racism, cultural encapsulation, and personal choice are contributing factors for patterns of same-race friendships. Current neuropsychology research further provides an explanation for the lack of cross-racial friendships. Behavioral neuroscience indicates that specific brain regions, particularly the amygdala, or emotional center, are active in initiating and maintaining racial bias (Phelps, O’Connor, Cunningham, Funayama, Gatenby, Gore, & Banaji, 2000; Wheeler & Fiske, 2005). Our shared neuropsychological makeup may also predispose us to aversive racism (Dovidio & Gaertner, 2004), a contemporary, subtle form of prejudice in which racial groups apply stereotypes to explain their negative attitudes toward others of a different race. Aversive racism operates through an unconscious mechanism by which racist behavior is rationalized on the basis of an attribute other than race. Cross-racial friendships may “potentially alter who is a we and who is a they, undermining a contributing factor to aversion racism” (Dovidio & Gaertner, 2004, p.30). Other research on implicit bias has suggested that to rid oneself of stereotypes that interfere with fostering cross-racial friendships, an individual must train his or her brain by providing contradictory stimuli (Banaji & Greenwald, 2013; Vedantam, 2013). Exposure to other races, particularly in friendship, can be a powerful way to reduce bias and change cultural beliefs (Page-Gould, Mendoza-Denton, & Tropp, 2008).
Outside of studies focused on school-age children’s developmental pattern in forming cross-racial friendships and studies of romantic interracial relationships, little is known about the nature of cross-racial friendships among adults. Understanding the nature of cross-racial friendships in a wide array of relationships, ranging from the intimacy of marriage to platonic friendships to casual work relationships would deepen our understanding of societal race relations (McPherson, Smith-Lovin, & Cook, 2001). In addition, the potential of such individual cross-racial friendships to advance race relations in the broader society is worth exploring. A substantial body of evidence has suggested significant benefits of socializing across racial lines (Aboud, Mendelson, & Purdy, 2010; Dovidio, Kawakami, & Gaertner, 2002; Page-Gould, Mendoza-Denton, & Tropp, 2008; Page-Gould, 2010), including reduced racial isolation in communities, the creation of a better informed citizenry, expansion of the concept of citizenship to a global level, improvement of team performance in organizations, and an increase in innovation in organizations and the arts (Devine & Vasquez, 1998; Plummer, 2004, 2009). Social connections, in general, have been linked to improved academic performance, income, and overall quality of life (Lieberman, 2013). However, as evidenced by the preponderance of same-race friendships (PRRC, 2013), not all Americans have benefited from these connections.
Thus, the purpose of this study was twofold. First, the study investigated the nature of adult cross-racial friendships and attempted to identify the individual factors that increase the probability of having cross-racial friends. Second, the study looked at the identified individual factors in the context of contemporary race relations to explore their potential to advance positive race relations. We designed two related studies for this research. The first examined the frequency and quality of cross-racial friendships and identified the demographic characteristics of individuals most likely to have cross-racial friends. In the second study, we further investigated the depth or degree of familiarity of the cross-racial friendships and identified factors that distinguished cross-racial friendships from same-race friendships. In both studies, we explored the social and historical context as distinguishing factors of cross-racial friendships. We then examined the potential for these friendships to influence positive race relations. We hypothesized that participant characteristics such as race, ethnicity, age, gender, education, and attitudes about contemporary race issues are related to the number and depth of participants’ cross-racial friendships. Our inquiry across both studies was informed by Blau’s (1977) macrosociological theory of social structure, which posits the critical importance of intersectionality of social structure and life experiences.
STUDY 1
The first study was designed to assess the nature and patterns of friendships that cross racial lines based on national administration of the Friendship Survey via the Internet and paper distribution.
Methods
Participants
We received 1,055 responses to the national survey. We eliminated respondents who did not live in the United States (n=20) and those with incomplete responses (n=27), resulting in a final sample size of 1,008 respondents. Respondents were mostly female (73%) between the ages of 18 and 55 years (72%), and most were white (55%) or black (32%). The respondents were generally highly educated, with only 6.5% attaining an educational level of high school or less. Reported annual household income ranged from $9,000 or less to more than $200,000. Respondents were from a broad range of geographic areas across the United States. Table 1 provides the details of respondent characteristics.
Table 1
Participant Sociodemographic Characteristics, Overall and by Number of Cross-Racial Friends, National Cross-Racial Friendship Cohort, 2013–14 (n=1008).
| Number of Cross-Racial Friendships | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall | None | 1–4 | 4–8 | ≥8 | z-score1 (p-value) | ||||||
| n | (%) | n | (%) | n | (%) | n | (%) | n | (%) | ||
| Overall | 1008 | 70 | (6.9) | 280 | (27.8) | 184 | (18.3) | 474 | (47.0) | ||
|
| |||||||||||
| Age | |||||||||||
| 18 – 35 years | 270 | (26.9) | 19 | (7.0) | 84 | (31.1) | 43 | (15.9) | 124 | (45.9) | 0.81 (0.418) |
| 36 – 55 years | 453 | (45.2) | 29 | (6.4) | 126 | (27.8) | 84 | (18.5) | 214 | (47.2) | |
| ≥ 56 years | 280 | (27.9) | 22 | (7.9) | 68 | (24.3) | 56 | (20.0) | 134 | (47.9) | |
|
| |||||||||||
| Race/ethnicity | |||||||||||
| White/European American | 548 | (54.9) | 41 | (7.5) | 177 | (32.3) | 114 | (20.8) | 216 | (39.4) | 3.16 (0.002) |
| Black/African American | 317 | (31.8) | 22 | (6.9) | 80 | (25.2) | 54 | (17.0) | 161 | (50.8) | |
| Asian American/Pacific Islander | 34 | (3.4) | 0 | (0.0) | 7 | (20.6) | 2 | (5.9) | 25 | (73.5) | |
| Hispanic/Latino/Latina | 49 | (4.9) | 4 | (8.2) | 8 | (16.3) | 4 | (8.2) | 33 | (67.4) | |
| Multiracial/ethnic | 50 | (5.0) | 3 | (6.0) | 5 | (10.0) | 7 | (14.0) | 35 | (70.0) | |
|
| |||||||||||
| Gender | |||||||||||
| Female | 729 | (73.5) | 43 | (5.9) | 227 | (31.1) | 132 | (18.1) | 327 | (44.9) | 2.29 (0.022) |
| Male | 263 | (26.5) | 26 | (9.9) | 47 | (17.9) | 48 | (18.3) | 142 | (54.0) | |
|
| |||||||||||
| Education | |||||||||||
| High school or less | 65 | (6.5) | 15 | (23.1) | 15 | (23.1) | 11 | (16.9) | 24 | (36.9) | 2.91 (0.004) |
| Some college | 149 | (14.9) | 16 | (10.7) | 40 | (26.9) | 26 | (17.5) | 67 | (45.0) | |
| ≥ Bachelor’s degree | 788 | (78.6) | 38 | (4.8) | 225 | (28.6) | 145 | (18.4) | 380 | (48.2) | |
|
| |||||||||||
| Annual household income | |||||||||||
| ≤ $24,000 | 121 | (12.7) | 18 | (14.9) | 32 | (26.5) | 17 | (14.1) | 54 | (44.6) | 2.05 (0.040) |
| $25,000 – $74,000 | 329 | (34.5) | 20 | (6.1) | 90 | (27.4) | 70 | (21.3) | 149 | (45.3) | |
| ≥$75,000 | 504 | (52.8) | 22 | (4.4) | 145 | (28.7) | 84 | (16.7) | 253 | (50.2) | |
|
| |||||||||||
| Region | |||||||||||
| South | 181 | (18.2) | 14 | (7.7) | 44 | (24.3) | 33 | (18.2) | 90 | (49.7) | 0.83 (0.409) |
| East | 428 | (43.0) | 37 | (8.6) | 129 | (30.4) | 70 | (16.4) | 192 | (44.9) | |
| Midwest | 320 | (32.2) | 14 | (4.4) | 87 | (27.2) | 69 | (21.6) | 150 | (46.9) | |
| Southwest | 40 | (4.0) | 1 | (2.5) | 7 | (17.5) | 5 | (12.5) | 27 | (67.5) | |
| Other | 26 | (2.6) | 3 | (11.5) | 8 | (30.8) | 4 | (15.4) | 11 | (42.3) | |
Friendship Survey
This survey comprised 20 questions in four sections. For questions 1–6, respondents described the characteristic of a friend who is a different race than their own. Respondents with only friends of the same race were instructed to follow a skip pattern, skipping questions 2–6 because they focused on specific characteristics of the primary cross-racial friend. Respondents with more than one cross-racial friend were asked to focus on the primary or most important friend when responding to questions 2–6. Question 6 focused on the friendship depth or level of intimacy by asking respondents whether they had engaged in any of the following activities with their primary cross-racial friend: (1) socializing at home during such events as dinners, parties, and birthday celebrations, (2) vacationing together, (3) calling in times of emotional distress late at night or early in the morning, (4) borrowing money, or (5) witnessing a family argument. Questions 7–11 assessed the respondents’ attitude on contemporary race relations. Respondents were asked to endorse or refute statements about attitudes on contemporary race relationships, including (1) “Societal messages about race relations influence my choice to have friends that cross racial lines,” (2) “America’s history of race relations prevents people from crossing racial lines in friendship,” (3) “I am uncomfortable in social settings where I am the only one present of my racial group,” and (4) “I do not have a problem with interracial marriage.” For those reporting at least one cross-racial friendship, the survey asked about shared interests and background, including (1) educational level, (2) socioeconomic status, (3) hobbies, (4) geographic location, (5) social group of friends, (6) political affiliation, (7) place of employment, and (8) faith community.
Questions 12–17 requested respondents’ demographic information. Questions 18, 19, and 20 were optional open-ended questions that solicited examples from respondents of their experiences with a cross-racial friend that were stressful or upsetting, inspiring or motivating, and humorous or funny.
Sampling Procedures
The national survey cohort was constructed by using two response methods — online and paper distribution. In both cases, participation was voluntary and conducted in compliance with the university review board. Participants were asked to respond to a survey describing their friendship patterns. An opportunity to win a $100 gift card was offered as an incentive to online participants and a $5 gift card was provided for completion of the paper survey. The online survey (83%, n=874) was administered through Survey Monkey. Respondents to the online survey were solicited through participation in colleges, universities, and churches and through social media. Respondents to the paper survey (17%, n=181) were a national sample solicited through church communities and a community agency.
Statistical Analysis
We began by examining univariate distributions for all variables. For the first bivariate analysis, the response variable (number of cross-racial friends) was left in its native ordinal format with four categories. Participant characteristics and opinions about cross-racial friendships were compared across categories of the ordinal response variable. For the second bivariate analysis, a Friendship Depth Index was constructed based on the total number of items endorsed as part of question 6 (range 0–5). Here, we compared mean scores on the Friendship Depth Index across categories of respondent sociodemographic characteristics and opinions about cross-racial friendships. For the final bivariate analysis, we compared mean scores on the Friendship Depth Index across number of shared interests. Statistical significance for the bivariate comparisons was determined by analysis of variance (ANOVA), the Bonferroni post-hoc test for multiple comparisons, the chi-squared test for independence, or the non-parametric trend tests using a method described by Cuzick (1985) as an extension of the Wilcoxon rank-sum test, as appropriate.
The first multivariable analysis took the number of cross-racial friendships as the outcome variable. To preserve all information contained in this ordinal variable, we used partial generalized ordinal logistic regression employing an approach described by Williams (2006). The outcome variable had four levels (Level 1: 0 cross-racial friends, Level 2: 1–4 cross-racial friends, Level 3: 5–8 cross-racial friends, and Level 4: > 8 cross-racial friends). Therefore, this procedure produced odds ratios for three cumulative comparisons: (1) Levels 2, 3, and 4 versus Level 1, (2) Levels 3 and 4 versus Levels 1 and 2, and (3) Level 4 versus Levels 1, 2, and 3. For each covariate, we tested the null hypothesis that the odds ratios were identical for all three comparisons, taking an alpha threshold of 0.003 to account for multiple comparisons. Failure to reject the null hypothesis led us to accept the proportional odds ratio assumption (i.e., odds ratios were not significantly different across comparisons) and report a single summary odds ratios for all three comparisons. When the proportional odds ratio assumption did not hold, separate odds ratios were reported for each comparison. Based on this model, we reported odds ratios and 95% confidence intervals for factors associated with the three cumulative comparisons described above and separately for having no cross-racial friends (versus having one or more friends).
The multivariable analysis for friendship depth used linear regression with robust standard errors based on the Huber-White sandwich variance estimation procedure. Beta coefficients were interpreted as adjusted differences in the Friendship Depth Index, and precision was ascertained from the 95% confidence intervals.
Results
Among all respondents, 47% (95% CI: 41%–50%) reported that they had eight or more cross-racial friends. Of those reporting at least one cross-racial friendship, 86% (95% CI: 84%–88%) indicated that these friendships were formed from a routine, predictable interaction that might occur at work, at school, or from living in the same neighborhood. Approximately 64% (95% CI: 61%–67%) of all friendships was formed more than five years earlier. Of all friendship pairs, about half was reported to share the same socioeconomic level, live in the same geographic location, or work at the same job. About one-third shared hobbies, but a quarter or less shared political or religious affiliation. Most (84%; 95% CI: 82%–87%) socialized at each other’s home, but fewer (32%; 95% CI: 29%–35%) vacationed together. More than half (54%; 95% CI: 51%–57%) felt comfortable calling their friend in times of emotional distress late at night or early in the morning, but fewer felt comfortable borrowing money (14%; 95% CI: 12%–16%). Likewise, only 26% (95% CI: 24%–29%) ever witnessed a family argument. Among those reporting any cross-racial friendships, the mean score on the Friendship Depth Index was 2.11 (standard deviation 1.36, range 0–5).
Table 1 presents results from the bivariate analyses which examined the association of sociodemographic characteristics with the number of cross-racial friends. From this analysis, all non-white participants had more cross-racial friends than those who were white. Females were less likely than males to have eight or more cross-racial friends. Larger income was associated with having more cross-racial friends. There were no associations with number of friends and age or geographic region. From Table 2, being uncomfortable as the only one of the respondent’s race in a social setting was associated with having fewer cross-racial friends. No other significant associations were observed between respondent opinion about societal messages and the number of cross-racial friends. Tables 3 and and44 describe the variation in mean scores on the Friendship Depth Index. As age increased, friendship depth decreased. Significant differences in friendship depth were observed by race. Hispanics and those with multiracial backgrounds had higher mean scores than white participants, and those with a multiracial background had higher scores than blacks. Friendship depth increased as the number of cross-racial friends increased and as the number of shared interests increased. Viewing societal messages as important and being comfortable as the only one of the respondent’s race were associated with greater friendship depth.
Table 2
Participant Opinions about Cross-Racial Friendships, Overall and by Number of Cross-Racial Friends, National Cross-Racial Friendship Cohort, 2013–14 (n=1008).
| Number of Cross-Racial Friendships | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||
| Overall | No Friends | 1–4 Friends | 4–8 Friends | > 8 Friends | z-score1 (p-value) | ||||||
| n | (%) | n | (%) | n | n | (%) | n | (%) | |||
| Societal messages as important2 | |||||||||||
| No | 844 | (84.5) | 57 | (6.8) | 232 | (27.5) | 144 | (17.1) | 411 | (48.7) | 1.38 (0.166) |
| Yes | 155 | (15.2) | 11 | (7.1) | 46 | (29.7) | 37 | (23.9) | 61 | (39.4) | |
|
| |||||||||||
| Racial history as important3 | |||||||||||
| No | 440 | (44.6) | 41 | (9.3) | 116 | (26.4) | 67 | (15.2) | 216 | (49.1) | 0.36 (0.721) |
| Yes | 547 | (55.4) | 26 | (4.8) | 160 | (29.3) | 114 | (20.8) | 247 | (45.2) | |
|
| |||||||||||
| Uncomfortable as the only one | |||||||||||
| No | 731 | (73.1) | 47 | (6.4) | 164 | (22.4) | 133 | (18.2) | 387 | (52.9) | 6.30 (<0.001) |
| Yes | 269 | (26.9) | 23 | (8.6) | 114 | (42.4) | 47 | (17.5) | 85 | (31.6) | |
|
| |||||||||||
| Approve of interracial marriage5 | |||||||||||
| No | 83 | (8.3) | 6 | (7.2) | 31 | (37.4) | 15 | (18.1) | 31 | (37.4) | 1.87 (0.062) |
| Yes | 912 | (91.7) | 64 | (7.0) | 245 | (26.9) | 164 | (18.0) | 439 | (48.1) | |
Table 3
Mean (Standard Deviation) for Friendship Depth Index by Participant Sociodemographic Characteristics, National Cross-Racial Friendship Cohort, 2013–14 (n=938).1
| mean | sd | Statistic2 (p-value) | |
|---|---|---|---|
| Age | |||
| 18 – 35 years | 2.46 | 1.44 | 5.14 (<0.001) |
| 36 – 55 years | 2.10 | 1.34 | |
| ≥ 56 years | 1.80 | 1.23 | |
|
| |||
| Race/ethnicity | |||
| White/European American | 2.00 | 1.30 | 11.88 (4, 923) (<0.001)3 |
| Black/African American | 2.02 | 1.31 | |
| Asian American/Pacific Islander | 2.47 | 1.44 | |
| Hispanic/Latino/Latina | 2.60 | 1.53 | |
| Multiracial/ethnic | 3.23 | 1.32 | |
|
| |||
| Gender | |||
| Female | 2.12 | 1.35 | 0.02 (1,921) (0.889) |
| Male | 2.11 | 1.36 | |
|
| |||
| Education | |||
| High school or less | 1.88 | 1.48 | 0.01 (0.991) |
| Some college | 2.37 | 1.51 | |
| ≥ Bachelor’s degree | 2.08 | 1.32 | |
|
| |||
| Annual household income | |||
| ≤ $24,000 | 2.27 | 1.47 | 0.75 (0.453) |
| $25,000 – $74,000 | 2.13 | 1.36 | |
| ≥$75,000 | 2.10 | 1.34 | |
|
| |||
| Region | |||
| South | 1.99 | 1.31 | 2.10 (4,921) (0.079) |
| East | 2.34 | 1.44 | |
| Midwest | 1.99 | 1.25 | |
| Southwest | 2.36 | 1.44 | |
| Other | 2.21 | 1.27 | |
|
| |||
| No. Cross-Racial Friends | |||
| 1–4 | 1.49 | 1.21 | 11.09 (<0.001) |
| 5–8 | 1.88 | 1.25 | |
| ≥8 | 2.56 | 1.32 | |
|
| |||
| No. Shared Interests | |||
| 1–2 | 1.52 | 1.14 | 8.75 (<0.001) |
| 3–6 | 2.26 | 1.34 | |
| 7–8 | 2.88 | 1.41 | |
Table 4
Mean (Standard Deviation) for Friendship Depth Index by Participant Opinions about Cross-Racial Friendships, National Cross-Racial Friendship Cohort, 2013–14 (n=1008).1
| mean | sd | F (df) (p-value)2 | |
|---|---|---|---|
| Societal messages as important3 | |||
| No | 2.16 | 1.33 | 5.90 (1, 929) (0.0153) |
| Yes | 1.86 | 1.43 | |
|
| |||
| Race history as important4 | |||
| No | 2.13 | 1.36 | 0.20 (1,918) (0.652) |
| Yes | 2.09 | 1.35 | |
|
| |||
| Uncomfortable as the only one 5 | |||
| No | 2.20 | 1.33 | 11.21 (1, 928) (<0.001) |
| Yes | 1.87 | 1.38 | |
|
| |||
| Approve of inter-racial marriage6 | |||
| No | 1.99 | 1.33 | 0.85 (1, 923) (0.375) |
| Yes | 2.13 | 1.36 | |
Figure 1 presents the results from the multivariable model with number of cross-racial friendships as the outcome. As described above, this model produced three cumulative comparisons for each covariate. Compared to white participants, those who were African American, Hispanic, or multiracial had greater odds of being in a higher cumulative category for number of cross-racial friends across all comparisons. Likewise, the odds of being in a higher cumulative category for number of friends were greater across all comparisons for those who had attained at least a bachelor’s degree than for those with a high school education or less. Conversely, the odds of being in a higher cumulative category were lower across all comparisons for those respondents who were uncomfortable being the only one of their race in a social setting. For women, the odds of being in a higher cumulative category for number of friends were less for comparisons 2 and 3, but not for comparison 1. In other words, our analysis isolated these observed differences by gender to the upper categories for number of friends, revealing no difference in the likelihood of having cross-racial friends.
Adjusted Odds Ratios and 95% Confidence Intervals for Being in a Higher Cumulative Category for Number of Cross-Racial Friends, National Cross-Racial Friendship Cohort, 2013–2014.
Approving of interracial marriage, viewing our nation’s racial history as important, educational attainment of at least a bachelor’s degree, or being African American, Hispanic, or multiracial were associated with having more cross-racial friends. Being female and being uncomfortable as the only one of the respondent’s race were associated with having fewer cross-racial friends.
The response variable had four levels: Level 1 - No cross-racial friends, Level 2 - 1–4 cross-racial friends, Level 3 - 5–8 cross-racial friends, and Level 4 - >8 cross racial friends. Reference categories: Age 18–23 years, white race, college degree or greater, and annual household income of $75,000 or greater. Significant confidence intervals exclude 1. *Indicates factors with odds ratios that differed across comparisons. More detailed explanation is provided in the text.
Figure 2 recasts the findings from comparison 1 (Figure 1) to present the odds ratios for having no cross-racial friends. From Figure 2, it is apparent that African Americans, Hispanics, and those with a multiracial background are less likely to have no cross-racial friends. Likewise, those with more education had lower odds of having no cross-racial friends. In contrast, being uncomfortable as the only one of the respondent’s race in a social setting was associated with a substantial increase in the odds of having no cross-racial friends.
Adjusted Odds Ratios and 95% Confidence Intervals for Having No Cross-Racial Friends, National Cross-Racial Friendship Cohort, 2014–2014.
Those who approved of interracial marriage, who attained at least a bachelor’s degree, and who were African American, Hispanic, or multiracial had lower odds of having no cross-racial friends. Those who were uncomfortable being the only one of their race in a social setting were more likely to have no cross-racial friends.
Reference categories: Age 18–23 years, white race, college degree or greater, and annual household income of $75,000 or greater. Confidence intervals for significant comparison do not include 1.
Figure 3 presents results from the multivariable model with depth of friendship as the outcome. Increasing age was associated with less friendship depth. In contrast, having a multiracial background, having more cross-racial friends, and having more shared experiences were associated with greater friendship depth. Figure 4 shows an almost linear trend of increasing friendship depth with number of shared experiences, and Table 5 presents the mean scores on the Depth of Friendship Index by shared experiences.
Adjusted Beta Coefficients and 95% Confidence Intervals for Friendship Depth Scale, National Cross-Racial Friendship Cohort, 2013–2014.
Increasing age was associated with less friendship depth. These factors were associated with greater friendship depth: being multiracial, being female, having a greater number of cross-racial friends, and having a greater number of shared experiences.
Reference categories: Age 18–23 years, white race, high school education or less, and annual household income of $24,000 or less. Confidence intervals for significant comparisons do not include 0.
Mean Scores on the Friendship Depth Index by Number of Shared Experiences, National Cross-Racial Friendship Cohort, 2013–2014.
There was an increasing relationship between number of shared experiences and the mean score on the Friendship Depth Index (potential range 0–5).
Potentially shared interests (n=8) included educational level, socioeconomic status, hobbies, geographic location, social groups of friends, political affiliation, place of employment, and faith community.
Non-parametric trend, p < 0.001.
Table 5
Mean (Standard Deviation) for Friendship Depth Index by Individual Shared Experiences, National Cross-Racial Friendship Cohort, 2013–14 (n=1008).1
| Potentially Shared Experience | mean | sd | F (df) (p-value)2 |
|---|---|---|---|
| Educational Attainment | |||
| No | 1.84 | 1.29 | 17.06 (1, 936) |
| Yes | 2.23 | 1.37 | <0.001 |
|
| |||
| Socioeconomic Status | |||
| No | 1.95 | 1.36 | 11.13 (1,936) |
| Yes | 2.24 | 1.34 | <0.001 |
|
| |||
| Hobbies | |||
| No | 1.81 | 1.28 | 73.59(1,936) |
| Yes | 2.56 | 1.35 | <0.001 |
|
| |||
| Geographic Region | |||
| No | 2.01 | 1.40 | 4.63 (1,936) |
| Yes | 2.20 | 1.31 | 0.031 |
|
| |||
| Social Circles | |||
| No | 1.82 | 1.30 | 51.58 (1,936) |
| Yes | 2.44 | 1.35 | <0.001 |
|
| |||
| Political Affiliation | |||
| No | 1.94 | 1.31 | 37.15 (1,936) |
| Yes | 2.53 | 1.39 | <0.001 |
|
| |||
| Shared Job | |||
| No | 2.20 | 1.35 | 3.74 (1,936) |
| Yes | 2.02 | 1.36 | 0.536 |
|
| |||
| Faith Community | |||
| No | 2.03 | 1.33 | 9.51 (1,936) |
| Yes | 2.36 | 1.40 | 0.002 |
Qualitative Analysis
Grounded theory methods were employed within the mixed methods research design to provide the framework for the qualitative analysis. We began using inductive data drawn from the literature outlining the benefits of cross-racial friendships and then invoked iterative strategies of going back and forth between the data and the analysis, using comparative methods as we examined the data and the emerging analysis (Charmaz, 2014). Responses to three survey questions (18, 19, and 20) were read and categorized independently by two reviewers using a color coding method, assigning different colors for each emerging theme. Phrases and concepts were treated as segments of data and assigned a short label that simultaneously summarized and accounted for each piece of data. Categories were not predetermined but emerged from repeated readings of the responses. A text analysis of the most frequently used words also revealed themes present in the responses. Tables 6–8 detail the themes and categories and report consistency of ratings (unadjusted agreement and Kappa statistics).
Table 6
Question 18 Stressful Situations: 6 Themes Derived from 31 Categories with Inter Rater Reliability Scoring.
| Theme | Categories | Agreement | Kappa* |
|---|---|---|---|
| Race Used as Identifier | African American, Arab, Asian, Black, Hispanic, Latino, Native American, Multiracial, White, Identity | .81 | .53 |
| Discussion of Current Events or Situation | Healthcare, King/Zimmerman Trials, Religion, Politics, General Discussion | .89 | .43 |
| Environment | Education, Family, School, Neighborhood, Work | .80 | .42 |
| Racial Profiling | Racial Profiling | .81 | .32 |
| Racially Motivated Behavior | Racial Comments, Slurs, Derogatory Comments (backhanded or slight) | .93 | .45 |
| Other/Miscellaneous | Body Language, Female, Male, Privacy, Respect, Friendship, Social Interaction, Not Applicable | .78 | .51 |
Table 8
Question 20 Humorous: 4 Themes Derived from 14 Categories with Inter Rater Reliability Scoring.
| Theme/Group | Categories | Agreement | Kappa* |
|---|---|---|---|
| Universal Elements of Friendship | Natural Aspects, Sharing Stories, Work, Hanging Out | .62 | .24 |
| Cultural Meaning | Cultural Meaning, Dance, Food, Slang, Joke, Laughter | .81 | .30 |
| Relationships | Dating and Relationships, Family, Children | .87 | .48 |
| Other | N/A | .67 | .55 |
Question 18 solicited examples of an experience with a cross-racial friend that was stressful or upsetting. Of the 387 examples provided, racially charged comments (23%, n=89), racial profiling (14%, n=53), and racial slurs (6%, n=25) were most frequent. Representative samples include the following:
I find it stressful when other friends make racial jokes because they think that shows their “friendship comfort level,” but really I feel embarrassed because it reinforces the ignorance of white people and how they don’t understand what racism really means.
One of my co-workers told a joke that I thought was offensive. When I told her how I felt about it, the friendship ended abruptly. I tried to revive the friendship and share my achievements but she apparently decided to move on. I knew then the friendship was not genuine.
Question 19 solicited examples of an experience with a cross-racial friend that was inspiring or motivational. Of the 396 examples provided, mutual affection, caring, concern, and trust (36%, n=142) were the highest category. Unsolicited support in times of illness or emotional stress (26%, n=103) and overcoming adversity together (18%, n=70) were also examples:
One of my friends from high school said that I was the first black person that he had ever known. He said that I was so sweet and nice that I really wasn’t at all what he had expected. At first I thought it was kind of insulting but then I realized if I can change one person’s mind by just being myself then maybe I shouldn’t feel bad about that.
I have been inspired by my circle of friends during the presidential elections for Obama. To see a group of inspired individuals of all races working hard for his election and celebrating his victory, our country’s victory, gave me hope for our future as a nation.
Question 20 solicited examples that were humorous or funny. Of the 341 examples provided, the majority of the responses described laughter or a joke between the friends (44%, n=151) and examples of taking delight in their friendship (35%, n=118):
From my husband (half black, half white), I learned the term “minute” to be defined as a long period of time. For example, “I haven’t done that in a minute” would mean it has been a long time since I have done that, whereas normally this particular word would mean a short period of time. I used this term at work with a white co-worker and a Hispanic co-worker. They were both confused and one asked, “I thought you said you hadn’t done that in a while?” This situation happened in front of different races and was picked up by a person of a completely different race. It was a humorous situation.
Study 1 Summary
The study has created a profile for an individual most likely to have cross-racial friends and the conditions upon which cross-racial friendships are established. Men and those with higher education are more likely to have more friends of a different race. Women are likely to have fewer cross-racial friends but those friendships are characterized by more depth. Cross-racial friendships are formed across all ages in adulthood. Younger men, women, and those with higher education are more likely to have high-depth friendships. Consistent with the literature on social distance, blacks and whites show the least probability of forming cross-racial friends, although blacks are still twice as likely as whites to have cross-racial friends.
Living in diverse neighborhoods, working in diverse organizations, and attending schools with diverse populations form the basis for establishing cross-racial friendships (Aboud, Mendelson, & Purdy, 2003; Page-Gould, Mendoza-Denton, & Troop, 2008; Page-Gould, 2010). The literature documents that the likelihood of cross-race and same-race friendship formation depends on structural conditions, including whether or not the groups live near or work with one another (Blau, 1977; Hallinan & Smith, 1985; Hallinan & Williams, 1989). Our study adds that level of education and shared interests greatly support or enhance the depth of the friendship. Table 5 further depicts these findings.
Cross-racial friendships mirror the characteristics of same-race friendships in what is deemed as inspiring or motivational and in what is experienced as humorous or funny. A critical mass of the respondents reporting stressful or upsetting incidents referred to a racial slur or racially charged comments or being racially profiled by their cross-racial friend at some point in their relationship. This finding supports the assertion that race influences the nature of interracial interactions in ways that contribute to the miscommunications and mistrust that have historically characterized race relations in the United States and that cross-racial friendships are rooted in a social context (Dovidio & Gaertner; 2004; Taylor, 2011).
Study 2
To provide additional context for the findings from study 1, three focus groups were conducted. These focus groups were designed to enhance the understanding of the nature of cross-racial friendships in a social context and examine specific aspects of cross-racial friendships that were similar to and different from same-race friendships. As in study 1, participants identified those aspects of their life — leisure activities, political affiliations, religious affiliations, neighborhood gatherings, civic activities, and consumer spending — that were shared with cross-racial friends. In addition, the influence of social context, such as police shootings of unarmed black men, on forming and sustaining cross-racial friendships was discussed.
Methods
Participants
Three focus groups were conducted. One focus group (12 participants: 4 male, 8 female) was held in Cleveland, Ohio, one was held in Atlanta, Georgia (10 participants: 2 male, 8 female), and a third was held San Francisco, California (9 participants; 5 male, 4 female). Participants were adults aged 20–66 years and were racially diverse (white n= 9; Hispanic n =2; Asian n=6; black n=10; Native American n=1; multiracial, n=3) and socioeconomically diverse (income range less than $9,000 annually to $150,000–$199,000).
Sampling Procedures
Focus group respondents were solicited by the facilitators through their professional networks using email lists and event postings on Facebook. The only screening criterion was that the participant must be at least 18 years of age. The Cleveland focus group was videotaped and facilitated by one of the authors. The Atlanta and California focus groups were audiotaped and facilitated by paid moderators. Informed consent was met through a participant data form explaining the project and its minimal risks, signed by each participant. The study was conducted in compliance with the university review board and all information was de-identified for analysis and reporting. All three moderators used the same moderator’s guide and process for facilitation. Twelve interview questions were organized around key themes for understanding the nature of cross-racial friendships and identifying shared experiences with participants’ cross-racial friends.
The 12 questions that comprised the moderator’s guide were divided into three sections: (1) activities with participants’ friends in general and activities with cross-racial friends, (2) frequency of contact and quality of the cross-racial friendships, and (3) attitudes on contemporary race relations that may or may not have affected their relationships with friends of a different race.
Results
A transcript-based analysis of the three focus groups was conducted using a color coding method similar to that used for the survey open-ended questions (Krueger & Casey, 2000; Krueger, 1997). Following grounded theory principles, categories and themes were not predetermined but emerged from the repeated readings of the responses (Charmaz, 2014).
Three themes emerged from the data: (1) the level of intimacy/depth and trust, (2) the educational experience of the friendship, and (3) the social value derived from the friendship.
Level of Intimacy/Depth and Trust
The majority of the focus group participants reported that although they engaged in the same kind of leisure activities with their cross-racial friends as they did with same-race friends, they experienced lower levels of trust and intimacy with cross-racial friends than with same-race friends. They attributed this difference to historical influences of American race relations. Illustrative quotations are as follows:
White folks don’t “get it” and some things would be different if my friend was black. Whites have a fear factor from segregating themselves based on old stereotypes. 55-year-old black female
I told my children not to marry a white person because history would come into the marriage. I am going to teach them the beauty of the Hindu culture. It is not racism. I will promote having more Indian friends because you can’t control who you fall in love with. If his circle of friends were more Indian/Hindu, he would more likely to marry a Hindu woman. 40-year-old Asian Indian female
I have white friends, Asian friends, Indian friends, you know, and kind of looking at the evolution of my friendships, starting out in the friendship there might have been a conversation that I would have with my African American friends that I wouldn’t necessarily have with my white friends, but over the course of time as that trust started to build, the conversations became pretty much the same. 27-year-old black female
I am listening to this conversation [about friends of different races] and I have nothing to draw from. I don’t really have any close friends of a different race. 46-year-old white female
I have an African American friend at work and we talk about everything and we are friends at work and away from work and we live in the same neighborhood and he’s friends with my husband. We talk about the struggles among races and we get down and dirty and it’s so refreshing to just be able to talk about how we get past this. 30-year-old white female
Educational Experiences
Among their cross-racial friends, participants cited situations of cultural misunderstanding as well as excitement from learning new things when they achieved understanding. Illustrative quotations are as follows:
We have some Japanese friends now. However, when we first moved to this community, respectively our kids would talk about a new friend who was also Japanese to her parents, or our daughter would talk about a new friend, who was black, to us. We, as parents, were both eager to meet. When we did, we both laughed as each of our children, about how each child described their new “Japanese/black” friend through their eyes. 38-year-old black female
There’s a group of people that are older than us that thought one way. We grew up during the civil rights era, then you see our kids and what you see in our kids is different than what you see in us. It’s different than what you see in our parents. It’s different than what you saw before that. I travel around the country quite a bit and even when you get into the Deep South, you see a change there. And it’s almost like they were always friends, but they weren’t allowed to admit it. And now you’re starting to see more friendships… 52-year-old white male
I grew up in a predominantly white neighborhood, so when I went to college it was really my first time that I was around other Asians. I was scared to death of other Asians to tell you the truth. My ultimate goal has always been (more so when I was younger) assimilation. I wanted to assimilate in society. I wanted to be like the rest of the neighborhood which was white. I resisted anything that would look Asian at all. It was only when I went to college that I had more experiences with people of more diverse backgrounds. That’s when I met friends of Asian backgrounds and different backgrounds. 32-year-old Asian male
It’s like that pendulum swinging, we go all the way over here and hope we get a little bit of movement and then we go all the way back over here and somewhere in the middle it [understanding] will show up. And that’s where we need to be to have the discussion. That seems to be, so far, my theme for myself. The important thing right now is to at least be able to have the discussion. It might end up in a blowup which doesn’t intimidate me as much as it used to. I’ll still continue to have the discussion. 46-year-old white female
I grew up in a very diverse city; I had a lot of diverse fiends and then I went on to a historically black college and that kind of ended. I lost touch with some of my friends of different races. Then coming back into corporate America (where you see there’s quite a lot of diversity) I am back in the situation where I have friends from different races. But I think even that it’s just on the surface, it never goes deeper; you never have that conversation with people because you may not be comfortable digging too deep into that relationship. 33-year-old black female
Social Value
Participants reported both enhanced and decreased ability to navigate a diverse society due to having cross-racial friends. Quotations that illustrate social value are as follows:
When I look at the diverse friendships that I have, a lot of them come from kind of like my faith activities and people I’ve met by being Buddhist. Connecting with my friend who is Indian, I’m motivated to learn or to be open to something new. I feel like my life becomes enhanced by interacting with her. 28-year-old black female
You want to tell me with what I know about the history of how blacks have been treated by whites in this country, that now I should forgive them and turn around and be friends with them? 45-year-old black male
I have a black friend at work and we talk about everything and we are friends at work and away from work and we live in the same neighborhood and he’s friends with my husband. We talk about the struggles among races and we get down and dirty and it’s so refreshing to just be able to talk about how we get past this. 30-year-old white female
Why do I even have to have black friends? 47-year-old white male
In addition to enhancing our understanding of the characteristics of cross-racial friendships, we sought to elicit examples of how life was shared among friends who crossed racial lines and to note any differences in these shared experiences from those of same-race friendships. As an initial exercise, participants first listed all of the activities they engaged in with their friends in general. They were then asked to review the list in comparison to those activities that they did with cross-racial friends. Participants identified leisure activities, attending neighborhood gatherings, engaging in civic activities, and consumer spending at the same establishments as mutual activities across all their friendships. Political affiliation and participation in faith communities were aspects of life shared with same-race friends that were not shared with cross-racial friends. Participants described the activities they did not share with cross-racial friends as “done out of comfort” or because they were “tired of waiting for people to catch up with understanding my reality.”
Younger participants (> 35 years of age) strongly believed there is a generational disconnect in the ease of making friends across racial lines and noted that although there was an appreciation for the sacrifices made to achieve racial equality, civil rights was “not my story. It’s my parent’s story. It does not get in the way of my friendships.”
Although the majority of participants reported that they believed their cross-racial friends had similar views about racially charged events such as police shootings of unarmed black men, they also were convinced that such cases were too controversial to talk about with cross-racial friends. They reported not wanting to “risk being misunderstood” and possibly destroying the relationship by bringing these issues into the friendship.
Study 2 Summary
In sum, focus group participants underscored social context as a factor that distinguished the nature of their cross-racial friendships from their same-race friendships. Participants also identified more challenges in establishing, maintaining, and nurturing cross-racial friendships than same-race friendships. However, participants also acknowledged the accompanying benefits of cross-racial friendships that outweighed the challenges:
I think the important thing is to realize that we need to be open minded and once we realize that every white person’s not trying to hold us back or every person of color is not angry and out to get a white person, then we’ll realize OK, I’m approaching this friendship and I’m being sincere and I take the time to understand this race and I know, oh this person likes certain things and it’s just easier to get to know people in that way instead of coming to them with those prejudices that are blocking and stifling friendships in the first place. 38-year-old Hispanic female
General Discussion
Both studies advanced efforts in cross-racial relations by demonstrating the nature and context in which adult cross-racial friendships are formed and sustained. Devine and Vasquez (1998) determined that lack of clear social norms for behavior, combined with a fear of appearing prejudiced in social settings, creates difficulties for most people who espouse egalitarian values and desire to act in non-prejudiced ways. Of importance to this success is enhancing our cultural competence, the ability to successfully navigate multicultural social and community settings.
We assert that having cross-racial friends provides a living laboratory for enhancing individual and collective cultural competence. We discuss our findings and the impact of societal influences on the shared life experiences of cross-racial friends in the following sections.
Race Differences in Cross-Racial Friendships
Results from this study suggest that Asians, Hispanics, and multiracial individuals are more likely than their white and black counterparts to have friends that cross racial lines. Unlike whites and blacks, Asians and multiracial individuals had more cross-racial friends and cross-racial friendships of greater depth (see Figure 3). Although significant progress has been made, a racial divide exists in having cross-racial friendships, most notably among blacks and whites. Two responses from focus group participants illustrate the reluctance to cross racial lines in friendship:
Why do I even have to have black friends? 47-year-old white male
You want to tell me with what I know about the history of how blacks have been treated by whites in this country, that now I should forgive them and turn around and be friends with them? 45-year-old black male
As depicted in these two quotations, cross-racial friendships frequently represent more than the relationship between the two individuals. How cross-racial friendships are viewed and interpreted is symbolic of and influenced by broader social nuances and themes related to race relations (Taylor, 2011). The following quotations depict the interplay of race and assimilation and its impact on having cross-racial friendships:
I grew up in a predominantly white neighborhood, so when I went to college it was really my first time that I was around other Asians. I was scared to death of other Asians to tell you the truth. My ultimate goal has always been (more so when I was younger) assimilation. I wanted to assimilate in society. I wanted to be like the rest of the neighborhood which was white. I resisted anything that would look Asian at all. It was only when I went to college that I had more experiences with people of more diverse backgrounds. That’s when I met friends of Asian backgrounds and different backgrounds. 32-year-old Asian male
I grew up in a very diverse city; I had a lot of diverse fiends and then I went on to a historically black college and that kind of ended. I lost touch with some of my friends of different races. Then coming back into corporate America (where you see there’s quite a lot of diversity) I am back in the situation where I have friends from different races. But I think even that it’s just on the surface, it never goes deeper; you never have that conversation with people because you may not be comfortable digging too deep into that relationship. 33-year-old black female
Taylor (2011) specified historical, social, and political-making systems as differentiating factors within cross-racial friendships. In our study, 5% (n=19) of the 387 examples of stressful or upsetting experiences with a cross-racial friend referred to a specific social context, such as when George Zimmerman fatally shot 17-year-old Trayvon Martin, setting off a national debate on race, or when discussing the election of President Obama. However, when one goes deeper, as evidenced by the focus group data, social context becomes an important factor in almost all cross-racial friendships. Focus group participants acknowledged avoiding discussions of racially charged events with cross-racial friends. They realized that race among cross-racial friends was both a unifying and polarizing force: unifying in that the friendship heightened their awareness of the universal nature of humanity and polarizing in that race also evoked disparities in how one is treated in America. Properly designed and executed training in having difficult conversations among cross-racial friends may hold important promise for changing the current narrative and improving race relations.
In this post-civil rights era, where overt forms of racism are outlawed, contemporary forms of racism, such as unconscious bias, are hard to identify and difficult to measure. An example of this is what Bonilla-Silva (2003) called “racism without racists.” “Racism without racists” is the interplay of structural racism (Lawrence & Keleher, 2004) and invisible white privilege (McIntosh, 2003). It is one of the hidden realities that make friendships between blacks and Hispanics with whites less probable. More education on recognizing and managing unconscious bias is needed to reduce tensions among blacks, Hispanics, and whites.
The higher probability of substantial numbers of cross-racial friendships with more depth for Asians and multiracial individuals speaks to the correlation between how contemporary forms of racism are experienced by these racial groups and how these experiences are partially mediated by their personal perceptions of racial identity (Alvarez, Juang, & Liang, 2006; Remedios & Chasteen, 2013). Further understanding of the connections among racial identity development, the racial socialization process, and perceptions of racism for Asians and multiracial individuals would help to unravel the quality of their cross-racial friendships and how these friendships support positive race relations.
Gender Differences in Cross-Racial Friendships
Studies have suggested that women have more favorable racial attitudes than men (Galupo & Gonzalez, 2013; Golebiowska, 2006; Hughes & Tuch, 2003). Our finding that women’s cross-racial friendships were characterized by more depth was consistent with the literature on gender differences in friendship patterns. These studies have argued that women share greater levels of emotional expressiveness and self-disclosure in their friendships than men (Caldwell & Peplau, 1982; Gillespie, Lever, Frederick, & Royce, 2014; Ryle, 2012). As a result, women are assessed as having more depth in their relationships. Following are illustrative quotations from women describing their cross-racial friendships:
When I look at the diverse friendships that I have, a lot of them come from kind of like my faith activities and people I’ve met by being Buddhist. Connecting with my friend who is Indian, I’m motivated to learn or to be open to something new. I feel like my life becomes enhanced by interacting with her. 28-year old black female
I have an African American friend at work and we talk about everything and we are friends at work and away from work and we live in the same neighborhood and he’s friends with my husband. We talk about the struggles among races and we get down and dirty and it’s so refreshing to just be able to talk about how we get past this. 30-year-old white female
We note here that assessments of depth in friendships are generally measured by feminine styles of intimacy, characterized by verbal self-disclosure and emotional expressiveness versus masculine styles that are characterized by engaging in activities together and commonality of interests (Cancian, 1987; Wright, 1988). Our Depth of Friendship Index included both measures of emotional connectedness and engaging in activities together, but did not include examples of shared physical activities that are more aligned with masculine styles of depth or intimacy. Thus, we bring the same caution for interpreting gender differences in friendships noted in the research to our study on cross-racial friendships.
Age and Education Differences in Cross-Racial Friendships
No association with the number of cross-racial friendships was found for age or education. Thus, cross-racial friendships are realized across generations and across educational levels. However, less education was an indicator of having fewer or no cross-racial friends. Younger people had greater depth of friendship. Of promise for improving race relations, younger generation adults with more education are likely to have more cross-racial friendships and those friendships are more likely to have greater depth. These generational differences were emphasized in the focus groups.
Shared Life Experiences in Cross-Racial Friendships
Consistent with Taylor’s (2011) assertion of social context as differentiating cross-racial friendships, those with no cross-racial friends reported higher endorsement of America’s history of race relations as a barrier. Although focus group participants reported sharing many leisure and civic activities with cross-racial friends, they reported that they did not share the same political or religious affiliation. Similarly, less than 25% of survey respondents shared the same political or religious affiliation as their friend of a different race. Political ideology is shaped by one’s relationship with society and its government. Because our political beliefs are guided by culture, it is reasonable to assume that different racial cultures hold differing views on government and what is or should happen in society and that political affiliation is not the same for friends of different races. Conversely, the majority of religious affiliations share a principle for racial justice, but the majority of cross-racial friends reported not sharing the same religious affiliation with cross-racial friends. Aversive racism, characterized by conscious (explicit) egalitarian attitudes and negative unconscious (implicit) attitudes and beliefs, may be a possible explanation. Despite a strong conscious commitment to racial equality, religious affiliations remain racially segregated (Pew Research Center, 2015).
In contrast to polling data reported by Reuters (Dunsmuir, 2013) and the study conducted by the PRRC (2013), the majority of the respondents to our survey and the focus group participants reported having cross-racial friends. A possible explanation is the difference in the methodology and the context for solicitation. Reuters is an international news agency and its polling methodology gathered yes or no responses over a five-day period to the statement: “I have a close friend of a different race.” Asking people if they have a friend of a different race would “lead people to search their memory more broadly for any cross-race tie, oversampling cross-race ties relative to same-race ties in memory and possibly creating interviewer demand effects” (McPherson, Smith-Lovin & Cook, 2001, page 420). In contrast, we were interested in any and all cross-racial friendships, not just close friendships. Our survey targeted adults aged 18 and older to assess both the number and the depth of their friendship patterns. The average score on the Depth of Friendship Index was the endorsement of only two shared experiences with cross-racial friends (Table 1). Thus, substantial room exists to increase shared experiences within cross-racial friendships and establish the foundation for a shared worldview.
Eliminating spatial racism by developing mixed-raced neighborhoods, establishing more multiracial faith communities, and creating political ideologies that are broader and less provincial are societal challenges that if translated into opportunities will increase the number of cross-racial friendships and provide benefits to all of us as a society.
Limitations
Our findings should be viewed within the context of certain limitations. The most important one follows from potential bias that may be introduced because respondents were volunteers and perhaps predisposed to addressing questions about their friendship patterns positively. In addition, we ascertained the number of cross-racial friendships in ordinal categories, thus preventing a more granular analysis of this important outcome. In particular, our highest category was ≥ 8 cross-racial friends, preventing a more detailed analysis among those participants who had a large number of friends. This is the classification approach that we used in our previous work, and we wanted this most recent update to be historically compatible (Plummer, 2004, 2009). The intent for the research design was to replicate the same study 10 years later and compare findings. Those findings will be reported in a separate study. However, our categories and cutoffs were aligned with other reported average numbers of friends (Cohen, 2013; Carroll, 2004). This suggests that our ascertainment categories were centered close to the average number of reported close friendships and that there would not be much additional meaningful variation beyond our top category of friendship number. Despite these limitations in how we collected the number of cross-racial friendships, several significant and important findings emerged.
Finally, Native Americans and non-U.S. respondents to the survey were too few to report. Although the focus groups were racially diverse, we were limited in the number of Hispanic participants and no reported Native Americans participated. We hope to broaden this representation in future research.
Our overall sample did contain a wide range of respondent sociodemographic characteristics and provided salient characteristics of individuals most likely to have cross-racial friends. It further provided detailed data about facets of contemporary race relationships that are not present in other available studies. In doing so, this study highlighted the influence of social, historical, and political context in shaping cross-racial friendships, especially among friendships between blacks and whites.
Future Considerations
Our study presents a profile of individuals who are most likely to have cross-racial friends, underscoring that whites are least likely to have cross-racial friends, cross-racial friendships among women are characterized by more depth, and cross-racial friendships are realized across all ages and educational levels. It also highlights the prominence of social context in cross-racial friendships. Earlier studies (Blau, 1997; Hallinan & Smith, 1985; Hallinan & Williams, 1989) emphasized social structures and the intersectionality of those contacts, whether geographic location or group status and group size, in fostering or limiting cross-racial interactions. What has not been considered is the intersectionality of interpersonal factors and identity group preferences that influence crossing racial lines in friendship. Our study examines some of these points of intersectionality. Further understanding of these points of intersectionality would serve as a baseline for an increased understanding of the nature of interracial marriage, the most intimate level of friendship, and its potential power for advancing societal race relations.
In addition, neuropsychology offers promise for shedding light on individual and group preferences that extend beyond social structures, interpersonal factors, and identity group preferences and might point us toward individual differences that exist. Recent studies have found that our nervous system reacts differently to stressful situations that we perceive as challenges versus those we perceive as threats. Although they both incite a physiological response, challenges incite a sequence that sends more blood to our system, whereas threats restrict our blood flow and, thus, “facing challenges is good for you; facing threats in not” (Page-Gould, 2010, p. 42). Cross-racial friendships hold the potential for turning a perceived threat into a challenge, thereby reducing implicit stereotypes and unconscious bias. Further study is required to determine whether cross-racial friendships can indeed reduce bias and change cultural beliefs.
Conclusion
Our survey respondents and focus group participants reported greater ease in navigating a global society and enhanced understanding of other cultures due to their experiences with cross-racial friends. However, most of the cross-racial friendships that were formed had low depth, suggesting much room for enhancing the quality of our cross-racial friendships, particularly to create the necessary foundation to challenge assumptions and expand worldviews. Because of the possibilities that cross-racial friendships hold for individual and societal benefit, it is imperative that further studies examine how barriers to establishing such friendships can be removed and how cross-racial friendships can be nurtured.
Table 7
Question 19 Inspiring or Motivational: 4 Themes Derived from 174 Categories with Inter Rater Reliability Scoring.
| Theme/Group | Categories | Agreement | Kappa* |
|---|---|---|---|
| Support | Mutual Affection, Caring, Concern, Acceptance, Friendship | .87 | .66 |
| Relationships | Children, Family, Interracial Relationship | .85 | .38 |
| Overcoming Adversity | Overcoming Adversity | .81 | .43 |
| Celebrating Together | Celebrating Together | .90 | .42 |
| Other/Miscellaneous | Political Choices, Prayer, Religion, N/A | .80 | .39 |
Contributor Information
Deborah L. Plummer, University of Massachusetts Medical School.
Jeroan Allison, University of Massachusetts Medical School.
Rosalie Torres Stone, Clark University.
Lauren Powell, University of Massachusetts Medical School.
References
- Aboud F, Mendelson M, Purdy K. Cross-race peer relations and friendship quality. International Journal of Behavioral Development. 2003;27(2):165–173. [Google Scholar]
- Alvarez A, Juang L, Liang C. Asian Americans and racism: When bad things happen to “model minorities. Cultural Diversity and Ethnic Minority Psychology. 2006;12(3):477–492. [PubMed] [Google Scholar]
- Banaji M, Greenwald A. Blindspot: Hidden biases of good people. New York, NY: Delacorte Press; 2013. [Google Scholar]
- Blau PM. A macrosociological theory of social structure. American Journal of Sociology. 1977:26–54. [Google Scholar]
- Bonilla-Silva E. Race and Racism. Lanham, MD: Rowman & Littlefield; 2003. Racism without “racists” pp. 91–97. [Google Scholar]
- Brokaw T. Dateline NBC. New York, NY: Scripps Howard Foundation; 1997. Why We Can’t Live Together [Radio series episode] [Google Scholar]
- Burns C, Barton K, Kerby S. The state of diversity in today’s workforce. 2012. Jul 1, [Google Scholar]
- Caldwell M, Peplau L. Sex differences in same sex friendships. Sex Roles. 1982;8:721–732. [Google Scholar]
- Cancian F. Love in America: Gender and self-development. Cambridge: Cambridge University Press; 1987. [Google Scholar]
- Carroll J. Americans Satisfied with Number of Friends, Closeness of Friendship. 2004 Retrieved from http://www.gallup.com/poll/10891/americans-satisfied-number-friends-closeness-friendships.aspx.
- Charmaz K. Constructing grounded theory. 2. Thousand Oaks, CA: Sage Publications; 2014. [Google Scholar]
- Cohen H. Number of Real Friends Declining. 2011 Retrieved from http://jonbarron.org/article/number-real-friends-declining#.VhK24vlVhBc.
- Cuzick J. A Wilcoxon-type test for trend. Statistics in Medicine. 1985;4:87–90. [PubMed] [Google Scholar]
- Devine P, Vasquez K. Confronting racism: The problem and the response. Newbury Park, CA: Sage Publications; 1998. The rocky road to positive intergroup relations; pp. 234–262. [Google Scholar]
- DiPrete T, Gelman A, McCormick T, Teitler J, Zheng T. Segregation in social networks based on acquaintanceship and trust. American Journal of Sociology. 2011;116(4):1234–1283. [PubMed] [Google Scholar]
- Dovidio J, Kawakami K, Gaertner S. Implicit and explicit prejudice and interracial interaction. Journal of Personality and Social Psychology. 2002;82(1):62–68. [PubMed] [Google Scholar]
- Dovidio J, Gaertner S. Aversive racism. Advances in Experimental Social Psychology. 2004;36:1–52. [Google Scholar]
- Dunsmuir L. Many Americans have no friends of another race: Poll. Reuters. 2013 Aug 8; Retrieved from: http://www.reuters.com/article/2013/08/08/us-usa-poll-race-idUSBRE97704320130808.
- Galupo M, Gonzalez K. Friendship values and cross-category friendships: Understanding adult friendship patterns across gender, sexual orientation and race. Sex Roles: A Journal of Research. 2013;68:779–790. [Google Scholar]
- George F. Dwell in My Love. 2001 Apr 4; Retrieved from http://www.catholic.org/featured/headline.php?ID=390.
- Gillespie B, Lever J, Frederick D, Royce T. Close adult friendships, gender, and the life cycle. Journal of Social and Personal Relationships. 2014:1–28. [Google Scholar]
- Glaeser E, Vigdor J. The end of the segregated century: Racial separation in America’s neighborhoods, 1890–2010. 2012. Jan 1, [Google Scholar]
- Golebiowska E. Tolerance in the twenty-first century. Lanham, MD: Rowman & Littlefield Publishing Group; 2006. Gender and tolerance; pp. 113–137. [Google Scholar]
- Hallinan MT, Smith SS. The effects of classroom racial composition on students’ interracial friendliness. Social Psychology Quarterly. 1985:3–16. [Google Scholar]
- Hallinan MT, Williams RA. Interracial friendship choices in secondary schools. American Sociological Review. 1989:67–78. [Google Scholar]
- Huber PJ. The behavior of maximum likelihood estimates under nonstandard conditions. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability; Berkeley, CA: University of California Press; 1967. pp. 221–223. [Google Scholar]
- Hoenig C. What does segregation look like today? Diversity Inc. 2013 Retrieved from http://www.diversityinc.com/news/what-does-segregation-look-like-today/
- Hughes M, Tuch S. Gender differences in whites’ racial attitudes: Are women’s attitudes really more favorable? Social Psychology Quarterly. 2003;66(4):384–401. [Google Scholar]
- Krueger R. Analyzing and reporting focus group results (Focus Group Kit) Thousand Oaks, CA: Sage Publications; 1997. [Google Scholar]
- Krueger R, Casey M. Focus groups 3rd edition: A practical guide for applied research. Thousand Oaks, CA: Sage Publications; 2000. [Google Scholar]
- Lawrence K, Keleher T. Structural racism. Chronic Disparity: Strong and Pervasive Evidence of Racial Inequalities Poverty Outcomes 2004 [Google Scholar]
- Lieberman M. Social: Why our brains are wired to connect. New York, NY: Crown Publisher; 2013. [Google Scholar]
- McCubbin L, Bennett S. Encyclopedia of counseling. Vol. 3. Thousand Oaks, CA: Sage Publications; 2008. Cultural encapsulation; pp. 1091–1092. [Google Scholar]
- McIntosh P. White privilege: Unpacking the invisible knapsack. In: Plous S, editor. Understanding prejudice and discrimination. New York, NY, US: McGraw-Hill; 2003. pp. 191–196. [Google Scholar]
- Mcpherson M, Smith-Lovin L, Brashears M. Social isolation in America: Changes in core discussion networks over two decades. American Sociological Review. 2006;71:353–353. [Google Scholar]
- Mcpherson M, Smith-Lovin L, Cook J. Birds of a feather: Homophiliy in social networks. Annual Review of Sociology. 2001;2:415–444. [Google Scholar]
- Page-Gould E, Mendoza-Denton R, Tropp L. With a little help from my cross-group friend: Reducing anxiety in intergroup contexts through cross-group friendship. Journal of Personality and Social Psychology. 2008;95(5):1080–1094. [PubMed] [Google Scholar]
- Page-Gould E. Are we born racist? New insights from neuroscience and positive psychology. Boston, MA: Beacon Press; 2010. The unhealthy racist; pp. 41–44. [Google Scholar]
- Pettigrew T, Tropp L. A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology. 2006;90(5):751–783. [PubMed] [Google Scholar]
- Pew Research Center. America’s Changing Religious Landscape. 2015 May 12; Retrieved from http://www.pewforum.org/2015/05/12/americas-changing-religious-landscape/
- Phelps E, O’Connor K, Cunningham W, Funayama E, Gatenby J, Gore J, Banaji M. Performance on indirect measures of race evaluation predicts amygdala activation. Journal of Cognitive Neuroscience. 2000;12(5):729–738. [PubMed] [Google Scholar]
- Phelps E, O’Connor K, Cunningham W, Funayama E, Gatenby J, Gore J, Banaji M. Performance on indirect measures of race evaluation predicts amygdala activation. Journal of Cognitive Neuroscience. 2000:729–738. [PubMed] [Google Scholar]
- Plummer D. Race talk in the workplace: Unraveling intention and impact. OD Practitioner. 2004;36(2):35–39. [Google Scholar]
- Plummer D. Racing across the lines: Changing race relations through friendship. Cleveland, OH: Pilgrim Press; 2009. Rev. ed. [Google Scholar]
- Public Religion Research Institute. American Value Survey. 2013 Retrieved from http://publicreligion.org/research/2014/08/analysis-social-network/
- Remedios J, Chasteen A. Finally, someone who “gets” me! Multiracial people value others’ accuracy about their race. Cultural Diversity and Ethnic Minority Psychology. 2013;19(4):453–460. [PubMed] [Google Scholar]
- Ryle R. Questioning gender: A sociological exploration. Thousand Oaks, CA: Sage/Pine Forge Press; 2012. [Google Scholar]
- Taylor M. Encyclopedia of Child Behavior and Development. Springer US; 2011. Cross-race friendships; pp. 441–444. [Google Scholar]
- Toossi M. Labor force projections to 2020: A more slowly growing workforce. Monthly Labor Review. 2012:43–64. [Google Scholar]
- Vedantam S. How to Fight Racial Bias When It’s Silent and Subtle. 2013 Jul 19; Retrieved from http://www.npr.org/blogs/codeswitch/2013/07/19/203306999/How-To-Fight-Racial-Bias-When-Its-Silent-And-Subtle.
- Wheeler M, Fiske S. Controlling racial prejudice; Social cognitive goals affect amygdala and stereotype activation. Psychological Science. 2005:56–63. [PubMed] [Google Scholar]
- White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica. 1980;48:817–830. [Google Scholar]
- Williams D, Collins C. Racial residential segregation: A fundamental cause of racial disparities in health. Public Health Reports. n.d;116(September–October):404–416. [PMC free article] [PubMed] [Google Scholar]
- Williams D, Collins C. Racial residential segregation: A fundamental cause of racial disparities in health. Public Health Reports. 2001;116:404–416. [PMC free article] [PubMed] [Google Scholar]
- Williams R. Generalized order logit partial proportional odds model for ordinal dependent variables. Stata Journal. 2006;6(1):58–82. [Google Scholar]
- Wrenn C. The culturally encapsulated counselor. Harvard Educational Review. 1964;32(4):444–449. [Google Scholar]
- Wright P. Interpreting research on gender differences in friendship: A case for moderation and a plea for caution. Journal of Social and Personal Relationships. 1988;5:367–373. [Google Scholar]
- Xu J. Confidence intervals for predicted outcomes in regression models for categorical outcomes. Stata Journal. 2005;5(4):537–559. [Google Scholar]
Facebook
Twitter
Google+