|Home | About | Journals | Submit | Contact Us | Français|
This mixed-methods exploratory study examined the diverse content and situated context of White adolescents' racial-ethnic identities. The sample consisted of 781 9th–12th grade White adolescents from three New England schools, which varied in racial and economic make-up. Open-ended responses provided a range of thematic categories regarding the importance of race-ethnicity to the adolescents' identities, representing the diverse ideologies of White adolescents' explanations, ranging from colorblind claims to ethnic pride. This study also found significant relationships between racial-ethnic identity importance (centrality) and parents' education for White adolescents. These findings highlight the diversity of White adolescents' understanding of their racial-ethnic identities and the importance of context in shaping racial-ethnic centrality.
Whereas exploration of ethnic and racial identity for adolescents of color in psychology has burgeoned over the past decade, “Whiteness” remains an under-investigated construct, often unseen and unacknowledged (Knowles and Peng 2005). However, there is increasing awareness that acknowledgement and exploration of White racial identity are important steps toward recognizing and understanding racism and White privilege (e.g., Doane 1997; McIntosh 2003). A recent trend in scholarship on White racial identity highlights the complexity of White identity (Frankenberg 2001), emphasizing the diversity within the White population. McDermott and Samson (2005) pointed out that while theorists have taken steps to understand experiences of Whiteness and White racial identity, little empirical work has investigated these theories in real-life settings, or explored the diverse meanings of White identity.
The current mixed-methods study aimed to reduce the invisibility of Whiteness by qualitatively exploring the meanings White adolescents ascribe to their racial-ethnic identities. We use the constructs of racial centrality and ideology to capture dual aspects of White adolescents' racial-ethnic identity; namely, the importance of their race-ethnicity and the meanings they attribute to their group membership. To address the diverse contexts that shape adolescents' experiences and identities, we also quantitatively explore racial-ethnic identity against the backdrop of adolescents' social class and school racial demographics.
Though sociologists and other social scientists have had a relatively long history of studying members of dominant social groups, psychologists have been slow to include empirical studies of White identity (Knowles and Peng 2005). This stems in part from a less salient role of racial and ethnic identity for White adolescents as opposed to those of color (Charmaraman and Grossman 2008; French et al. 2000), for whom it constitutes a protective response from the negative effects of systemic racial oppression (Helms and Cook 1999). The invisibility of Whiteness may serve as another research barrier, as Whiteness is often considered the “default,” unexamined racial category (Hyde 1995; McIntosh 2003), marking its higher status on the ladder of social hierarchy (Rosaldo 1989). McIntosh (2003) argued that even those Whites who recognize the disadvantages of racism for people of color are often taught to ignore the implications of White privilege. Whereas many persons of non-White backgrounds, in particular African Americans, confront their race on a regular basis, many Whites do not believe they have a race at all, as evidenced in studies with college students (Jackson and Heckman 2002). These combined factors can lead to limited interest in and engagement with Whiteness in both personal and academic contexts, emphasizing the need to further investigate Whiteness.
While discussion within psychology is ongoing regarding intersections of racial and ethnic identity, recent research documents many parallels between these constructs, such as developing a sense of belonging, learning about one's group memberships, and responding to discrimination (Phinney and Ong 2007). Cross and Cross (2007) also noted conceptual and statistical overlap of these terms, and proposed classifying groups into “racial-ethnic” descriptors in a hybrid approach, which acknowledges the socially constructed and interlaced nature of both terms. The relationship of racial and ethnic identity may have unique complexities among White people. Waters (1990) points out that ethnic identity can provide a more comfortable means of identification by avoiding discomfort and distress that often accompany White racial identity. Fear of negative external perceptions of White racial identification may also discourage Whites from actively engaging with or acknowledging White racial identity (Frankenberg 1993). Further, ethnic identity can provide more tangible cultural beliefs and practices in contrast to the elusive domain of White racial identity and its concomitant cultural attitudes and values (Phinney 1990). These complexities suggest that the roles of racial and ethnic identity among Whites are highly intertwined, and may be particularly difficult to disentangle.
One of the understudied aspects of racial-ethnic identity research is the construct of importance or centrality of race-ethnicity in one's identity. Given the focus on race in the centrality research, the majority of the studies we review for the current study are from the racial identity literature. For the sake of consistency, we used the racial-ethnic terms that the original researchers used in their studies. Amid the range of proposed components of racial and ethnic identity (e.g., Phinney and Ong 2007; Yip and Fuligni 2002), Sellers and his colleagues' multidimensional model of African American racial identity is unique in its emphasis on individuals' subjective perceptions of their race (Sellers et al. 1998). These authors posited that the importance and explanations that individuals attribute to their race, rather than externally imposed meanings, provide critical insight into one's racial identity. Thus, it is critical to separate the constructs of racial centrality (the extent to which people identify race as an important part of their identities) and ideology (the meanings people attribute to their racial identity).
While there has been preliminary exploration of centrality and ideology with adolescents of color, they represent under-explored aspects of White racial identity (Knowles and Peng 2005). Existing studies indicate lower levels of ethnic importance for White adolescents than for Asian Americans, Blacks, and Latinos (Herman 2004; Phinney and Alipuria 1990), which may relate to discomfort with White privilege and acceptance of Whiteness as the “norm” (Frankenberg 1993; McIntosh 2003). In a focus group study of White college students, only one in fifteen participants identified race as an important aspect of their identities, implying Whiteness as the standard for comparison (Jackson and Heckman 2002). Knowles and Peng (2005) also demonstrated that the higher an individual's level of White identity centrality, the more likely one feels somewhat responsible and possibly guilty for the historical transgressions of the in-group. Thus, initial research demonstrates low levels of racial importance among Whites and explores potential disincentives for racial engagement.
In attributing meanings to their White identities, people vary in the extent to which they recognize racial inequities and the implications of White privilege, with ideologies such as colorblind humanism at one end of the spectrum, and multicultural ideologies at the other. Many White people do not make connections between their structural advantages and opportunities in life and their racial background (Lipsitz 1998). McDermott and Samson (2005) argue that the denial of White privilege is the foundation of an ideology characterized by racism that purports to be “color blind,” taking the stance that we are all only human beings and our racial/ethnic differences should be deemphasized or forgotten. This colorblind “humanism” spares dominant group members the discomfort of considering unearned advantages merely by being members of a White race, and may link to findings that White adolescents and young adults are often unable to articulate what it means to be White (Jackson and Heckman 2002; Perry 2002). At the other end of the spectrum, a multicultural ideology that is “race-cognizant” entails a belief that race and ethnicity are critical cultural differences between groups that powerfully shape their life experiences (Frankenberg 1993). Multicultural theorists encourage people of all colors to actively confront race and racism, by reflecting on such topics as the evolution of negative racial stereotypes and the subtle sources of perpetuating prejudice and discrimination in our society (Akintunde 2006). While preliminary findings suggest possible explanations for Whites' racial ideologies, further empirical investigation of racial centrality for Whites would help to gain deeper insight into the meaning and implications of Whites' racial and ethnic identities.
Despite overall low levels of racial centrality among White adolescents, studies suggest some variation in the salience of White racial identity (e.g., Hartigan 1999; Phinney et al. 1997). Variations in White adolescents' racial centrality and ideology can be understood in part through an exploration of the multiple contexts that shape adolescents' experiences, attitudes, and beliefs. Ecological theory emphasizes the importance of context in contributing to youths' development and outcomes (Bronfenbrenner 1979), namely that there are many layers of proximal and distal environments, such as microsystems (family, peers) and mesosystems (school, neighborhood), that together impact individuals' experiences and outcomes. This theoretical framework is consistent with a situational understanding of White racial identity that has become a focus of recent research, in which Whiteness is understood as fluid across contexts and over time (McDermott and Samson 2005). Two constructs through which situated Whiteness has been initially explored are exposure and marginalization. While exposure focuses on opportunities to interact with people different from oneself, marginalization relates to ways in which a person has experienced a non-dominant identity.
Within the context of mesosystems such as school or neighborhood, research has shown that White adolescents become especially conscious of their race-ethnicity in situations when they are in a numerical minority or in an ethnically diverse setting (French et al. 2000; Phinney et al. 1997; Roberts et al. 1999). Such a situation juxtaposes the impact of being a majority race in larger society, but a numerical minority in specific contexts. Social Identity Theory identifies the importance of group membership to self-concept and proposes that contact with others different from oneself increases one's in-group identification. Within the context of White racial identity, this theory suggests that social exposure to non-Whites would increase the centrality of one's White racial identity by more closely connecting self-perceptions with in-group identification (Tajfel 1978). Past research supports the importance of exposure in shaping the centrality of White racial identity across school (French et al. 2000) and neighborhood (Knowles and Peng 2005) contexts. For example, Knowles and Peng (2005) found higher levels of racial centrality among White adolescents from largely non-White regions than from regions that were primarily White. Phinney and colleagues (1997) found that White adolescent ethnic identity greatly affected self-esteem and the need for group solidarity, particularly when adolescents comprised a small proportion of a non-White setting. Therefore, one would expect that being in the numerical minority in a primary life setting might influence the salience of White adolescents' racial identity.
School serves as a critical context of adolescents' racial socialization, and a growing body of research documents relationships between school racial/ethnic composition and adolescents' racial experiences and identities (e.g., Cross 1995; Umaña-Taylor 2004). For example, in an ethnographic study, Perry (2001) compared White adolescents in a predominantly White high school with those in a multicultural high school, which made a conscious effort to expose the student body to different ethnic cultural traditions, such as multilingual murals along the school grounds and the creation of culture-based student groups. While White students at the predominantly Caucasian high school took Whiteness for granted, at the multicultural high school, White students tended to be discontent at the lack of an “ethnic” identity, and felt the need to have a unique cultural identity and pride that the minority groups seemed to share amongst themselves (Perry 2002). Given the paucity of research exploring relationships between school context and racial-ethnic identity, French and her colleagues (2006) recommended further investigations of adolescent racial-ethnic identity across heterogeneous and homogeneous school contexts. Such research may help to elucidate the processes through which proximal environments interact with White adolescents' developing understanding of racial and ethnic identity.
Not all White adolescents experience constant privilege, since certain contextual factors such as rural settings, sexual orientation, or socioeconomic class can place some of these White adolescents in the cultural margins (Buck 2001). McDermott and Samson (2005) suggested that due to ties between Whiteness and privilege, Whites from lower socioeconomic classes may have more awareness and nuanced understandings of what it means to be White. For example, working class Whites living in a predominantly Black community were found to be more aware of their Whiteness than those from more affluent backgrounds, experiencing both shame and pride in their racial identity (Hartigan 1999). In this example, both social class and exposure to people of different backgrounds may have shaped Whites' awareness. The relationship between social class and ethnic identity appears to be more mixed, such that Irish adolescents are significantly more likely to identify themselves as Irish if they are from a lower socioeconomic background, yet other studies have not found any relationship between the stages of ethnic identity and social class among high school students (Phinney 1990). In sum, the application of exposure and marginalization to the study of White racial identity suggests that contexts can play a meaningful role in shaping how White adolescents perceive the importance of their racial-ethnic identities.
The main goals of this mixed-methods study were to examine White adolescents' explanations of the level of importance they placed on their race and/or ethnicity and the role of school composition and parental education in adolescents' reports of racial-ethnic centrality. We investigated ideological orientations of White in-group identification utilizing grounded theory (Glaser and Strauss 1967). Quantitative analyses assessed the influences of parental education, centrality, and minority/majority racial status at school on participants' racial-ethnic centrality, and provided a descriptive context for the open-ended data, where participants made meaning of their racial and ethnic identities. Our first research questions provide descriptive contexts for the study: (a) Do White adolescents, who are a numerical minority at school, rate race-ethnicity as more central to their identity than those attending schools in which they are in the majority?; (b) Do White adolescents from lower socioeconomic backgrounds, as measured by mother's and father's education, report more central racial-ethnic identity than those with higher socioeconomic backgrounds? Our next research questions form the core of this study; (c) What are the phenomenological themes that arise when White adolescents are asked about how important their race-ethnicity is to their identities, including whether they focus on race, ethnicity, both, or neither?; (d) Does school context relate to the themes reported by White adolescents?
This sample was part of a larger research study on adolescent racial and ethnic identity and self-definition among monoracial and mixed-ancestry youth (Tracy et al. 2008). A total of 1793 adolescents in three schools participated in the study. This paper focuses on the 781 White, non-Hispanic participants.
In this White-minority school, 33 White adolescents (52% female) participated in the study and were comprised mainly of Italian, Polish, and Irish ethnicities. Latin High School is a low-income, predominantly Latino urban school, whose student body composition was the following: 68% Latino, 16% White (non-Hispanic), 15% Black, 1% Asian (School Matters website 2008). The median income for this predominantly Latino community was $22,628 (U.S. Census Bureau 2000).
In this White-majority school, 648 White adolescents (52% female) participated in the study and consisted mainly of Irish, Italian, and Jewish ancestry. White High School is an affluent suburban school, in which the vast majority (84%) of the students were White, non-Hispanic, and the rest were 7% Asian, 4% Black, and 3% Latino, and 2% multi-racial (School Matters website 2008). The median income for this predominantly White community was $113,686 (U.S. Census Bureau 2000).
In this White-minority school, 100 White adolescents (49% female) participated in the study, and it was mainly composed of Irish and Italian adolescents, as well as those of multiple European mixes. Multicultural High School is a middle income suburban school, in which 55% were Black, 23% White (non-Hispanic), 15% Asian, and 7% Latino (School Matters website 2008). The median income for this community was $55,255.
Students ranged in age from 14 to 21 years (M = 15.82, SD = 1.19). Approximately half were female (52%).
Several high schools throughout the New England area were contacted and invited to participate. The three schools that agreed to participate recommended using passive informed consent procedures, and administrative staff and school counselors assisted in disseminating parental permission forms. Each student in the high school was given a letter which described the study and an accompanying parent/guardian informed consent form to take home. Students were asked to provide their active assent before completing the in-class survey, which was about “how people form ideas of who they are.” Students were informed that their participation was voluntary and that their answers would be kept confidential. The entire survey took approximately 20 minutes to complete. Data were collected within a two-month period in early 2007. Project personnel were on hand to answer questions and to collect the completed surveys. The survey included questions about each participant's age, gender, mother's and father's educational levels, racial-ethnic identification, and the importance of his or her racial-ethnic background.
Respondents were asked about their race-ethnicity in one of three different randomly assigned ways: (a) multiple choice checkboxes, one or more of which could be checked, taken directly from the 2000 Census form; (b) checkboxes and fill-ins for the major categories selected for focus in this study; (c) open-ended fill-in of race-ethnicity. In addition, participants were given other opportunities to identify their race/ethnic background, such as reporting the race/ethnicity of each biological parent. Only adolescents who identified White as their sole racial identification were included in the current study. Adolescents who identified as White Hispanic or of mixed backgrounds including White were excluded.
We used the inclusive descriptor of “race/ethnicity” in order to widely capture participants' self-concept of what ethnic and racial labels meant to them. The following question was used to assess centrality of race-ethnicity on adolescent identity: “How important to you is race/ethnicity in describing who you are?” This item was scored on a four-point Likert scale, with end points of Not at all important (1) to Among the most important (4). This self-report variable was positively skewed with responses more frequent at the lower end of racial-ethnic importance (M = 1.95, SD = 0.72). Responses were: not at all important (38%), somewhat important (49%), very important (11%) and among the most important (2%). Ten participants were missing data from this variable. Respondents were given the opportunity to comment on the importance of the race-ethnicity rating that they indicated. Five hundred and one (64% out of 781) White adolescents provided answers to the open-ended item.
Two items reporting parents' levels of education (maternal and paternal education) served as proxies for socioeconomic status (Lien et al. 2001). For these items, participants were asked “What is the highest level of your parents' education?” Responses were scored on a five-point scale, ranging from Did not complete high school to Masters or doctorate. Both maternal education and paternal education were negatively skewed, reporting relatively high levels of formal education. Frequencies for maternal and paternal education are reported in Table 1. Twenty-seven participants were missing data for maternal education, and forty-four were missing paternal education. A factor model was used to combine mother's and father's education, reflecting the joint educational attainment relative to the other participants in the study.1 Standardized factor loadings were .84 for each of the education variables.
We created a dichotomous variable to assess whether White students were in the numerical majority at their school. For this variable, students from White high (White majority) were compared with students from Multicultural and Latin high (non-White majority). We also created a contrast variable of Latin vs. Multicultural high, which was included as a covariate in our analysis (Cohen et al. 2003).
White adolescents' reported racial-ethnic centrality levels were largely in the low to mid-level range (M = 1.75, SD = .72), with most participants reporting that their race-ethnicity was either “not at all important” or “somewhat important,” whereas adolescents of color reported moderately higher levels (M = 2.59, SD = .90) (see Charmaraman and Grossman 2008). The intraclass correlation of this measure for current sample was .06. A significant difference was found on racial-ethnic centrality between adolescents who did and did not provide explanations for their centrality ratings (t = −2.90, p < .01). A lack of explanation is consistent with lower racial-ethnic centrality, suggesting a lack of engagement with racial and ethnic identity for these adolescents. No significant sex differences were found for racial-ethnic importance.3
To provide a descriptive context for the thematic analysis presented below, we conducted multiple regression analysis in which parent education and two dummy variables contrasting schools with different racial/ethnic composition were used to predict centrality. The results of this analysis are presented below and in Table 2.4
Only parental education, used as a proxy for socioeconomic status in this study, was significant, negatively predicting self-rated racial-ethnic centrality (B = −0.10, SE = .04, p < .01), indicating that adolescents whose parents had more formal education reported lower levels of racial-ethnic importance. Majority/minority school composition did not significantly predict racial-ethnic importance.
The qualitative data in the current study are a subset from a larger dataset comprised of adolescents from multiple racial/ethnic backgrounds (Charmaraman and Grossman 2008). Only data with responses to the open-ended item were included (N = 501). We qualitatively analyzed the free response portion of the centrality question in which the phenomenological themes emerged from the data rather than relying on pre-designated researcher categories. Specifically, we utilized a grounded theory approach (Glaser and Strauss 1967), in which we developed codes for recurrent themes and categories, in a multi-step process. To bolster coding reliability, we first coded a sample of the data independently, then generated themes collaboratively in an on-going, iterative process. For the first stage of open coding, we randomly selected and assigned preliminary categories to 10 responses from each school, including both White and non-White adolescents, until we reached a total of 200 responses. Since many responses were not one-dimensional, some responses were coded in more than one category. To prevent being inadvertently influenced by contextual elements during the coding process, we were “blind” to the respondents' school, parental education background, and race or ethnicity. As open coding resumed, the remaining qualitative responses were incorporated into the coding process in order to refine the initial categories. We continued to compare our independently rated researcher codes into the axial coding stage, which consists of periodic recalibration of the initial open codes for overarching core themes until all potential categories are saturated. At this stage, we were able to focus our attention on generating sub-themes within the broader categories in order to further describe the different layers of each theme. Once codes were refined, the entire dataset was re-coded.
When there was disagreement, final ratings were negotiated between the raters. Responses that were undecipherable or too vaguely written to be easily interpretable were left out of the analysis. This occurred in 11 of the cases, leaving a total of 490 qualitative responses for analysis from the original 501 participants who provided open-ended responses. Interrater reliability of codes with two coders was calculated during the axial coding stage, using Miles and Huberman's (1994) formula in which reliability equals the number of agreements divided by sum total of agreements and disagreements. Initial intercoder reliability was .89, and eventual consensus was reached on each thematic code via discussion and debate over an area of disagreement. A review of coding results indicated that themes from White students' responses diverged from those of the non-White sample. Therefore, we decided to focus exclusively on White, non-Hispanic adolescents for the current study.
Thematic analyses suggested that among those responses evidencing some engagement with race-ethnicity, the following main themes emerged: (a) Positive Regard, indicating a sense of pride and positive affect; (b) Ambivalent Regard, which signals a sense of ambivalence or somewhat negative affect toward one's cultural background. Responses coded with Positive Regard and Ambivalent Regard identified mid to high levels of centrality, thus we considered both to constitute some engagement with race-ethnicity. Three sub-themes emerged from the overall theme of Positive Regard, including Internal Pride or the desire to become acquainted with your own culture(s) for internal validation, External Pride or the desire to externally represent your unique culture to the outside world so that others can gain a positive perspective and external validation, and Homophily or a preference for or comfort with others of the same racial or ethnic background; (c) Awareness of Inequities which was coded into one of two sub-themes: Stereotypes and Discrimination, which is a general realization that prejudice and discrimination exist or White Privilege, signaling an acknowledgement of the phenomenon; and (d) Acceptance of Diversity, which acknowledges that we are not all the same and that culture may shape who we are. Examples of themes and sub-themes are shown in Table 3.
Among participants' responses reflecting some rejection of race-ethnicity, explanations related to the following main themes: (a) Racial Disengagement or a belief that race does not or should not matter and (b) Unexamined Racial-Ethnic Identity or a lack of reflection on one's racial-ethnic identity. The theme of Racial Disengagement indicates some prior reflection and active decision-making about one's racial identity choices. Though the theme of Unexamined Racial-Ethnic Identity is also marked by some level of racial disengagement, it is further distinguishable by a lack of reflection or a pre-awareness of how race or ethnicity might be expressed or how it might impact their daily lives. The theme of Racial Disengagement broke down into the following sub-themes: (a) Colorblind Ideology refers to the belief that we are all human and shouldn't be categorized by race or ethnicity; (b) Individuality refers to the belief that one's personality or individual qualities/talents are equally (or more) central in self-characterization; (c) Non-homophily reflects the idea that race is not an important factor in connecting with others. Participants reported that they had friends of all racial backgrounds, and were open to relationships outside of their own racial group; (d) American Identity is a belief that American nationality was a primary descriptor of their background, rather than race or ethnicity; and (e) Low Definition corresponds to the rejection of the idea that race or ethnicity defines oneself.
The most common response over-all (42%) was the major theme of Racial Disengagement, with Low Definition (16%) and Colorblind Ideology (11%) being the most popular sub-themes, followed by Individuality (9%), as shown in Table 3. Responses such as “race does not define me,” appeared repeatedly during thematic analysis, embodying the sub-theme of Low Definition. Other responses coded under Low Definition included personal reflections surrounding race and skin color being far from central to their identities: “My race/ethnicity does not make me who I am or what I'm like. It is only a color” or “I am me, I could be purple with pink polka-dots and I would still be me. Race is not important.” The sub-theme of Colorblind emphasized a generalized belief that people shouldn't be categorized by race or ethnicity: “I see people of all races as equal in mind, body, and soul. This way it doesn't make me different in any significant way, except physical features which aren't as important.” Other types of colorblind responses discussed the desire to distance themselves from acknowledging any racial differences in society: “In this school, race doesn't matter. I have friends of all races and I don't notice it. I am not deep in my culture at all, and it just doesn't affect me at all.” Another sub-theme that focused on personalized aspects of racial disengagement was Individuality, in which individuals believe that characteristics such as personality or abilities (individualism) are more central to their identities than race or ethnicity, as indicated in the following response: “It shouldn't even matter what race you are. Your own self controls who you are in life and no other factors.” Other responses forwarded the notion that individual efforts count more than other factors which might give an advantage: “It shouldn't matter-my race/ethnicity-because that has nothing to do with who I am and what I can accomplish.”
Unexamined responses (18%) ranged from lack of engagement surrounding issues of race or ethnicity (e.g., “I never think about my race during my daily life”) to statements that race and ethnicity are irrelevant to understanding their personal identities (e.g., “It does not affect my life in any way”). Some participants referred to their majority or White status as a reason for this lack of engagement or exploration, such as “Since I have been labeled White all my life, it has never occurred to (me to) specify what that means.”
In contrast to the high percentage of racially disengaged responses, there were also relatively high numbers of some engaged responses, particularly in the Positive Regard theme, which contained the second highest number (37%) of coded responses from the White adolescents. Within that category, Internal Pride was the most frequently coded sub-theme (29%), with External Pride (5%) and Homophily (3%) trailing far behind. The types of responses within Internal Pride ranged from being invested in connecting with one's heritage (e.g., “I'm Italian and it's an important part of me because I have a large history of Italian ancestors and I'm proud to be who I am,”) to more ambivalent statements that reflect both engagement with heritage and emphasis on it being one of many aspects of identity (e.g., “I feel that my race and ethnicity is a part of me and makes me who I am, but it is not everything I am”). White adolescents less frequently demonstrated White racial pride (48% of internal pride responses referred to ethnicity only, while 11% referred only to race), in contrast with the greater emphasis on race in responses expressing disengagement from race/ethnicity, in which 46% referred to race, while only 14% referred to ethnicity.
While participants were not directly asked to report their ethnic background in this study, half of the sample (50%) specifically identified an ethnic background. From those who did report their ethnic background, ethnic trends emerged relative to some thematic categories. Within Internal Pride, the ethnic groups that demonstrated particularly high levels relative to their representation in the sample were Italian, Jewish, and Polish. In contrast, participants showed relatively low levels of External Pride, of which over half were of Irish ancestry. For example, “I want people to know that I'm Irish/Scottish.” Other groups which were overrepresented within External Pride relative to the entire sample included French, German, Portuguese, Scandinavian, and Middle Eastern participants. Participants also reported low levels of Homophily, or a preference for those of the same racial or ethnic background, such as “Most people I encounter everyday are the same ethnicity. However, I feel more comfortable and less worried about not offending other cultures when I'm with people of my same race.” While Positive Regard emerged more frequently, a minority of participants reported responses consistent with Ambivalent Regard (7%), showing negative or ambivalent feelings toward their race-ethnicity. Some of these responses reflected feelings of shame or distress at one's Whiteness: “It defines me—but sometimes I feel shame about it,” while other responses showed reluctance to publicly recognize one's racial-ethnic background: “I don't tell people that I'm Caucasian/white. It's not a good conversation starter.”
Awareness of Inequities was the next most common type of response (16%), showing awareness of the negative effects of prejudice in society. Most respondents discussed these issues externally, referring to stereotyping and discrimination experienced by others, in particular, minorities: “It is not that important to me but sometimes I feel uncomfortable talking to minorities who are being obviously discriminated against.” Less common responses showed awareness of the sub-theme of White Privilege, alluding to the fact that Whites are often sheltered from hardships faced by minorities: “There is still racism, so being White prevents me from being persecuted.” The least common type of theme addressed by White adolescents was Acceptance of Diversity, which only 4% of the sample mentioned in their free response explanations. For example, “Your race is who you are. By showing and doing things from your race people learn new things about you.”
Preliminary descriptive analysis of the open-ended responses demonstrated several qualitative differences between the different schools. In Table 4 the percentage of responses by theme and school are listed. Since the number of White students at Latin High School was far lower than at the other two schools, it was difficult to make any claims about the differences across schools. Thus, we will primarily describe how responses from Multicultural High and White High appeared different with respect to White students' explanations of racial-ethnic centrality. These observations of school differences are preliminary and descriptive, identifying initial trends observed in this data.
The school reflecting the highest number of Positive Regard responses, especially in terms of External Pride, was Multicultural High (44%), where Whites are not in the majority as compared to White High (36%). One student at Multicultural High specifically mentioned how their school composition affected their identity: “It's important because in this school there are very few white kids and I'm proud to be white and act how I act.” Multicultural High students also showed more responses consistent with Individuality (13%) than those at White High (8%). Acceptance of Diversity, while represented in only a small number of participant responses, was found in relatively equal proportions in Multicultural High and White High (3%). The Unexamined theme was most common with adolescents at White High School (20%) indicating that they were more likely than Multicultural High (9%) to not have explored their White race-ethnicity. White High students were the only ones who reported Homophily (4%) and American Identity (4%) out of all the schools. Homophily responses showed students' preference to spend time with others of their own race or ethnicity, such as: “I'm proud of my background and can relate easily to people of my same background.” In the American Identity sub-theme, students reflected on what being American meant to them: “I am very proud to be Irish-American. My family celebrates Irish holidays and I am extremely patriotic towards America.”
Thematic results from this study show wide diversity of experiences of White identity. However, consistent with past research, these findings suggest distancing from White racial identity, at times choosing instead to focus on one's ethnic origins. The degree of importance adolescents placed on race depended on the parental educational level in which they were situated, such that the higher the level of parental educational attainment, the lower the emphasis on race as an important aspect of self-definition. A more situational, less enduring context of school racial composition did not significantly impact the relative importance of race and/or ethnicity to White adolescents' individual identities. These findings highlight the broad range of White adolescents' understanding of racial-ethnic identity, and the potential role of context in shaping racial-ethnic importance.
Adolescents' overall low ratings of racial-ethnic centrality in this study are consistent with a trend among White Americans to downplay racial differences (Bonilla-Silva 2003). The minimizing of Whiteness in participants' quantitative and qualitative responses may relate to Whites' fears that any acknowledgement of their racial identity could be thought of as racist (Frankenberg 1993). This protective strategy is often perceived by those adopting it as a means to ignore racism and White domination. In this sense, colorblindness is an attempt to make the invisibility of race a goal for everyone, including people of color (Sullivan 2006).
This study's findings for parental education support past research demonstrating that people of higher socioeconomic backgrounds are less likely to be aware of their Whiteness (Frankenberg 2001; Hartigan 1999). Adolescents from highly educated families likely experience multiple levels of privilege, and therefore have more to lose from engagement with Whiteness and the privileges it holds. Families with less class privilege, however, may see reduced benefit from White privilege, which is often intertwined with White identity in general. These findings are also consistent with self-perceptions of middle class to affluent White society as “cultureless” (Hartigan 1999). While the parental education variable in this study does not account for all dimensions of social class, it furthers the preliminary link between social class and racial-ethnic identity.
In contrast to parental education, school racial composition did not predict racial-ethnic centrality beyond the effects of parental education in this study. It may be that more targeted school socialization processes, rather than exposure to people of different backgrounds alone, are needed to raise White students' awareness of their racial-ethnic identity. Indeed, Allport (1954) maintained that inter-racial contact itself was not sufficient to encourage greater levels of tolerance. He proposed that institutional factors such as ensuring equal status, common goals, and mutual interdependence between members of different racial groups were critical in generating improved racial dynamics in social contexts. Other factors not assessed in this study, such as multicultural education and awareness, racial-ethnic mixing of peer groups, and the presence of non-White teachers and administrators may better predict White student racial and ethnic engagement.
The generalized lack of engagement with race or ethnicity evident in open-ended responses is consistent with past scholarship on Whiteness. However, in the current study, the nuances of adolescent racial-ethnic engagement and disengagement were dissected and categorized. The pre-awareness type of disengaged response, Unexamined Racial-Ethnic Identity, is consistent with the relatively low centrality of racial-ethnic identity in White adolescents. This unexplored status of White racial identity is similar to Helms and Carter's (1990) least developed racial identity schema, wherein there is a denial of meaningfulness of race for one's personal identity as well as for life in general. Our data suggest that not only is there a status where denial of the importance of racial-ethnic identity occurs (which could comprise elements of Colorblind Ideology), but also a status wherein a White adolescent does not appear to have reflected on these concepts at all. Several of the other disengaged sub-themes, such as Low Definition and Individuality, showed participants' rejection of the notion that race and/or ethnicity capture who they really are, distancing these constructs from their self-definition and identity. Many White adolescents responded with a “postcultural” Colorblindness or refusal to acknowledge the existence of race or ethnicity's influence in their lives, in part by emphasizing the separation of identity from one's physical appearance. Colorblindness represents a well-known ideological strategy for White adolescents that attributes one's accomplishments solely to individual causes, rejecting positive or negative influences of race or ethnicity (DiTomaso et al. 2003), and ameliorates discomfort from complex feelings of guilt or responsibility (Knowles and Peng 2005). The sub-themes within racial-ethnic disengagement begin to detail the multiple ways in which White adolescents understand and explain their distancing from White racial-ethnic identity.
For the theme of Positive Regard, White adolescents emphasized their ethnic pride more frequently than their White racial pride. Ethnic pride may provide a less problematic source of identification for White adolescents. Further, those expressing less engagement with racial-ethnic identity focused more on race than ethnicity. According to Jacobson (1998), White individuals attempt to distance themselves from ownership of White privilege by claiming connection with oppression experienced by past generations of their families. These findings also corroborate Waters' (1990) illustration of White ethnic identity as optional or one that could be highlighted instrumentally in certain situations and contexts, as opposed to their racial identity, which is invisible and mainstream. Participants' low levels of External Pride are unsurprising given that this concept is more common among adolescents of color, who typically must face public scrutiny about their origins and physical features, particularly when growing up in all White neighborhoods (Ramsey 1991). External Pride may be less familiar to White adolescents, who do not feel the need to positively represent their race, since they often blend into the dominant culture. Ambivalent Regard responses, though rare, show both internal ambivalence about one's own historical racial legacy and fears that others may negatively judge expressions of one's White racial identification (Frankenberg 1993). Even as respondents evidenced Positive Regard, their responses revealed more ambiguous feelings toward Whiteness, such as embracing ethnicity rather than racial heritage, which suggests the complexity of White adolescents' relationship with their racial-ethnic identity.
For the theme of Awareness of Inequities and its sub-themes of Stereotypes and Discrimination and White Privilege, participants opted to bring up societal issues regarding race-ethnicity rather than focusing solely on their own personal experiences. They varied in the extent to which they externalized these issues as only applying to people of color or implicated themselves as benefiting from these structural advantages. This range represents two levels of engagement with this issue, as individuals may acknowledge negative effects of discrimination, while rejecting the benefits they reap from it. This overall theme is similar to Helms and Carter's (1990) status of “pseudo-independence” which shows an intellectual awareness of White privilege, and is consistent with prior research in which some White college students identified Whiteness as “easy” and powerful (Jackson and Heckman 2002). These responses suggest that although participants ranged in how closely they identified with inequities, issues of injustice represented a central area of concern within racial-ethnic identity.
Finally, Acceptance of Diversity emerged as a far less common theme, which may reflect White adolescents' focus on more charged aspects of race and ethnicity, such as either acknowledging racism or pushing away from recognition of racial inequities. While if asked specifically about Acceptance for Diversity, adolescents would likely endorse this concept, the theme may provide a less salient aspect of White racial and ethnic identity. In addition, since many White adolescents may feel like outsiders in situations where racial-ethnic diversity is celebrated, it is not surprising that responses reflecting Acceptance of Diversity did not occur frequently.
Preliminary comparisons across schools suggested some differences in explanations of racial-ethnic centrality. Students from Multicultural High, where Whites are not in the majority when compared to White High, seemed to show the most Positive Regard, especially in terms of External Pride, which could be a reflection of their numerical minority status. Individuality showed similar results, which may tie to concerns of White students at Multicultural High that they may be judged solely by their race. In contrast, Caucasian students at White High, surrounded by other students of their race, may feel that their race is not a factor in others' perceptions. These preliminary findings are somewhat consistent with Perry's (2001) qualitative research, which showed that students attending the multicultural high school reflected more on what it meant to be White, especially in terms of defining their interests and tastes in popular culture in contrast to other cultural groups.
As expected, White High students reported more frequent Unexamined responses, suggesting that since they are in the majority culture, race and ethnicity may not be particularly salient for them, consistent with themes of minimization and denial of Whiteness identified by Doane and Bonilla-Silva (2003). This lack of exploration may also relate to the greater privilege associated with the upper middle class community of White High, as identified in descriptive analyses. The sole presence of Homophily at White High may relate to students' assumptions that their homogenous environments are the norm and that it is only natural to spend time in the company of their own race. It may also relate to reduced opportunities to befriend others of a different race or ethnicity. Similarly, relatively high American identification among White High students may relate to their unquestioned sense of belonging in the “American” category, since according to Social Identity Theory, it is easier to align oneself with a majority racial group that is often equated with the in-group in the U.S. (Phinney et al. 1997). Despite these potential explanations, we must be cautious when interpreting these findings, given that the number of responses of the other two schools may have been too small for these themes to be observed.
In interpreting the study results, several limitations should be considered. Since this study focused on adolescents from the Northeast, it is unclear whether these findings are generalizable to other regions of the country with different racial-ethnic composition. The one-item measure of racial-ethnic centrality limits its scope, and the use of parental education as a proxy for social class provides only one aspect of this construct. While this mixed-methods data allowed us to investigate both qualitative and quantitative findings, its cross-sectional design limited our ability to explore processes or directionality of influences. Further, while our open-ended query regarding the importance of race-ethnicity permitted a broad range of explanatory responses and provided an opportunity to explore the extent to which adolescents identified with ethnic and/or racial identity, it sometimes presented ambiguity in whether participants' centrality ratings referred to ethnic or racial identification. In addition, we cannot fully explore their understanding of such broad topics as diversity or White privilege based on the limited response set. While an interview design would have limited the number of respondents, it would also have provided opportunities for participants to elaborate on or explain unclear answers. Differences in study variables between participants who did and did not complete qualitative responses suggest limits to the generalizability of thematic findings within this sample, as they may more accurately represent White students who have at least minimal levels of engagement with racial or ethnic identity. In addition, it may be advantageous to develop meaningful methods of asking White adolescents about their racial-ethnic identity by utilizing techniques such as implicit measures, in order to control for social desirability. Finally, low numbers of White participants in one of the schools curtailed the potential to examine school-level differences.
Despite these limitations, this study's findings may help to shape future practice and research. The relatively low levels of White adolescents' engagement with race and ethnicity in this and other studies suggests the need to educate White adolescents about their own racial and ethnic backgrounds, and to challenge the often invisible default standard of Whiteness. As Perry (2001) noted, the focus of research and evaluations of multicultural-sensitivity programs has been primarily on students of color, with many schools and colleges incorporating multicultural celebrations and offering “ethnic” support centers, often confirming that Whites are the neutral, homogenous standard. They repeatedly and perhaps unwittingly recognize the less politically-charged categories of style, food, and custom differences between groups without confronting historical and current realities of power inequity and oppression (Sullivan 2006). Future research could investigate the unique racial socialization within White families, such that the cultural practices of language, food, customs, and celebrations might be examined in as much detail as the cultural traditions of ethnic “others.”
The existence of an often overlooked “underclass” of Whites presents a challenge to the default notion that being White equals privilege (Hartigan 1997). We hope to generate more research on the diversity and possible sources of marginalization within the White racial experience, e.g. lower socioeconomic status, sexual orientation, White ethnic backgrounds in the numerical minority. Specifically, our findings suggest a need to further explore the processes through which class privilege interacts with White privilege and identity. Future directions should also empirically deconstruct how Whites distinguish race from ethnicity in forming their identities and provide longitudinal research that explores developmental influences on how these identities are formulated.
Practice-related directions can entail multicultural awareness programs that include a joint training effort for school counselors, teachers, administrators, and students, and explicitly include White adolescents and overt discussions of how to deal with race and racism. Such programs can include the role of White teachers' perceptions of their own racial identity, which may affect the inclusion of “non-color blind” lessons in their pedagogical practices (Blumer and Tatum 1999). Another important direction would be to design programs that confront anti-oppression, yet aim to minimize the potential feelings of defensiveness or “White guilt” through promoting positive alternative models of Whiteness and the utilization of White allies in the anti-oppression struggle by people of color (Tatum 1994). By encouraging nonracist White racial and ethnic identity development, teachers and counselors can potentially foster academic and social environments where (a) negative stereotypes regarding both non-Whites and Whites are confronted and deconstructed; and (b) pervasive issues regarding power and privilege are acknowledged in both racially integrated and more segregated school environments.
This study gives empirical support to earlier, more anecdotal research regarding White adolescents' explanations of racial-ethnic identity. It reifies the relative lack of importance of race-ethnicity to White adolescents, but more importantly, it documents the wide range of meanings that White students' attribute to their racial-ethnic centrality. This finding challenges a simplified perception of White racial identity as uniformly disengaged, moving toward a more complex appreciation of the diversity within White racial-ethnic identity. It also suggests that during high school, racial and ethnic identity exploration has already begun for many White students, which may provide opportunities for engagement with White high school students and school staff around racial issues. Our results also indicate that contexts may contribute to White adolescents' racial-ethnic importance. Specifically, we found that higher levels of parents' formal education predicted lower levels of racial-ethnic centrality, suggesting that socioeconomic class may be one of the factors that shape the importance of race and ethnicity to adolescents' identities. Family status and privilege associated with parental education have not yet been widely examined in adolescent identity research. This finding provides an impetus to further pursue this area.
The authors were funded by the National Institute for Child Health and Human Development Postdoctoral Fellowship Program. This study stems from a larger project on Mixed Ancestry also funded by the NICHD. We are grateful to Allison Tracy for her statistical assistance, and to Sumru Erkut, Ineke Ceder and the Wellesley Centers for Women writing group for their editorial contributions. Special thanks to Peggy McIntosh for her support and encouragement.
Jennifer M. Grossman is a Research Scientist at Wellesley Centers for Women. She received her Ph.D. in Counseling Psychology from Boston College. Her major research interests include adolescent development in the context of racial and ethnic experiences, peer relationships, and family socialization processes.
Linda Charmaraman is an NICHD postdoctoral research fellow at the Wellesley Centers for Women. She received her Ph.D. in Human Development and Education from UC Berkeley. Her research interests include adolescent identity, positive urban youth development, and youth media cultures.
1The measurement model for parental education was fit within the context of the full regression model. This strategy yields a fully identified measurement model. Further, embedding the measurement model in the context of the substantive model allows us to appropriately handle missing data in education variables via full information maximum likelihood estimation (FIML). In other words, rather than dropping individuals from the analysis by means of listwise deletion and running the risk of a biased sample, information from the responses from all the variables in the model are used to estimate unbiased regression parameters in the presence of missing data in the full sample.
2For the quantitative data analysis, we used the Mplus statistical software (Version 5.0; Muthén and Muthén, 1998–2008) to fit our multiple regression analysis. With the Mplus software, it is possible to account for missing data in the dependent variable using full-information maximum likelihood estimation, accommodate the clustered sampling design, and appropriately specify and model non-normally distributed variables.
3We explored the possibility that there was a moderating effect between parent education and schools minority/majority composition in predicting racial/ethnic centrality by including a product term in the analysis. This model yielded no significant interaction effect so the product term was dropped from the regression model.
4Because the education variables are specified as censored from above, no standardized fit indices (e.g., χ2 statistic, CFI, RMSEA, etc.), regression parameters, or effect sizes (R2) are available. However, ignoring the censoring of these variables violates the assumption of multivariate normality. Sensitivity analyses were conducted in a normal theory framework and the results were similar but standard errors were larger.