Color-Blindness and Commonality

Perceiving Others in Terms of Group Membership

Whereas psychologists earlier considered racism to be a form of abnormal human functioning (e.g., as consequence of an excessively strict parental influence; Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950), researchers currently recognize that, at the individual level, racism is rooted in normal psychological processes. One fundamental process involves humans’ propensities to categorize objects and people into groups. This type of categorization forms an essential basis for the ways people view, think about, and respond to others; it is a critical process in the way that people actively derive meaning from complex environments. The instant an object is categorized, it is assumed to possess properties shared by other category members. As a consequence, people feel they have sufficient information to make reliable decisions quickly about the qualities others possesses and how they will likely behave. Of course, overgeneralized assumptions about another person can produce erroneous decisions. Thus, people generally sacrifice some accuracy in judging others for efficiency in dealing with the often overwhelming complexity of their social world.

Social categorization, the psychological classification of people into discrete groups, often occurs automatically on the basis of physical similarity or shared fate (Campbell, 1958). Social categorization has immediate and profound effects on perceptions of groups and their members. When people (or objects) are categorized into groups, actual differences within groups (i.e., among members of the same category) tend to be minimized, while differences between groups tend to become exaggerated and overgeneralized (Tajfel, 1969). Moreover, social categorization typically involves the identification of the individual with one of the groups, the in-group, often to the exclusion of other groups (out-groups).

Belonging to a group, particularly one that is high in status, provides enormous material and psychological benefits. Disparities in economic security, political power, and opportunities for social advancement (Sidanius & Pratto, 1999) produce different social realities, which substantially shape the everyday lives of members of advantaged and disadvantaged groups (Demoulin, Leyens, & Dovidio, 2009). Psychologically, groups provide a sense of security and connection, reducing feelings of uncertainty and buffering threats to one’s well-being (Correll & Park, 2005).

Perceiving a distinction between in-group and out-group members as a consequence of social categorization significantly shapes how people feel about, think about, and act toward others (Tajfel & Turner, 1979; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987; see also Abrams & Hogg, 2010). Emotionally, people spontaneously experience more positive affect toward other members of the in-group than toward members of the out-group (Otten & Moskowitz, 2000). Cognitively, people process information more deeply for in-group than for out-group members (Van Bavel, Packer, & Cunningham, 2008), tend to discount negative behaviors of in-group members (Hewstone, 1990), and encode negative actions of out-group members in abstract ways (e.g., dishonest) that make these attributions more difficult to disconfirm in the future when encountering specific inconsistent instances (Maass, Salvi, Arcuri, & Semin, 1989). Behaviorally, people are more trusting of in-group than out-group members (Foddy, Platow, & Yamagishi, 2009) and more helpful and cooperative with in-group members (Dovidio et al., 1997), and they tend to behave in a more greedy and less trustworthy way toward members of other groups than if they were reacting to each other as individuals (Insko et al., 2001).

Because of socialization in traditional American culture and the circumstances of contemporary life (e.g., residential and occupational segregation), people automatically perceive others in terms of at least three fundamental forms of group membership—gender, race, and age (Brewer, 1988). Social categorization, in turn, activates category-based expectations (stereotypes) and evaluations and feelings (prejudice) that play critical roles in the lives of members of social groups, particularly for members of low-status social groups because of the power and position of high status or dominant groups. In the next section, we illustrate how automatic and often unconscious psychological processes shape the dynamics of modern forms of bias at the individual level, contributing to color-blind racism at the collective level.

Contemporary Racial Bias

Traditionally, intergroup biases have been measured in explicit ways, such as self-report responses on surveys. As noted earlier, the expressed racial attitudes of White Americans toward Black Americans have become dramatically more positive over time, and traditional stereotypes appear to be rapidly fading (Bonilla-Silva & Dietrich, 2011; Dovidio & Gaertner, 2004; Pearson et al., 2009; Schuman, Steeh, Bobo, & Krysan, 1997). The vast majority of White Americans perceive that the United States is currently characterized by racial and gender equality, deny any personal bias, and, in fact, assert their ability to remain color- and gender-blind in their interactions and decisions.

However, considerable recent attention has been devoted to implicit biases. In contrast to explicit processes, which are conscious and deliberative expressions, implicit processes involve a lack of awareness and intention. Because of the automaticity of social categorization based on race, the mere presence of a Black person is sufficient to activate the associated racial stereotype and attitude. Implicit biases arise through repeated exposure and association of certain groups with specific characteristics (i.e., mental habits; Kawakami, Dovidio, Moll, Hermsen, & Russin, 2000), which may be rooted in repeated personal experience, widespread media exposure, or cultural representations of different groups. Investigations of implicit bias, which often rely on response-latency techniques such as the Implicit Association Test (IAT; Greenwald, Poehlman, Uhlmann, & Banaji, 2009), demonstrate that negative racial attitudes and stereotypes are automatically activated for a majority of White Americans (Blair, 2001), regardless of age, socioeconomic status, and political orientation. In the IAT, for example, White people generally respond more quickly when a positive word is paired with an image of a White person than a Black person, while they respond more quickly when a negative word is paired with a Black person than a White person. Self-report and implicit measures of stereotyping and prejudice are largely uncorrelated, suggesting that, among a large proportion of Americans, implicit biases may operate unconsciously to influence behavior (Nosek, Banaji, & Greenwald, 2002).

The dissociation between explicit attitudes and beliefs, which are inclusive and egalitarian, and pervasive implicit biases and stereotypes leads to contemporary forms of discrimination. Contemporary biases are elusive but powerful phenomena. In situations in which right or wrong is clearly defined or the appropriate course of behavior is obvious, people are unlikely to behave in a sexist or racist manner; to discriminate in these situations would be obvious and would violate personal egalitarian principles. However, in situations in which right and wrong is not clearly defined, appropriate behavior is not obvious, or a negative response could be justified on the basis of some factor other than race or sex, bias will be expressed in a subtle manner that insulates the perpetrator from being recognized—by others or even oneself—as representing unfair treatment (Dovidio & Gaertner, 2004).

For example, Dovidio and Gaertner (2000) experimentally examined White college students’ support for hiring Black and White applicants for a selective campus position within the same college in the years 1989 and 1999. When the candidates’ credentials clearly qualified or disqualified them for the position (very strong and weak qualification conditions), there was no discrimination against the Black candidate (i.e., the highly qualified Black candidate was just as likely to be hired as the highly qualified White candidate). However, when candidates’ qualifications for the position were less obvious and the appropriate decision was more ambiguous (moderate qualifications), White participants recommended the Black candidate significantly less often than the White candidate with exactly the same credentials. Whereas overt expressions of prejudice (measured by items on a self-report scale for each sample) declined over this 10-year period, the pattern of subtle discrimination in selection decisions remained essentially unchanged. This pattern was conceptually replicated in research with human resource professionals in Puerto Rico (see Dovidio & Gaertner, 2007).

Additional research offers further insight into processes that underlie these effects. When ambiguous or mixed credentials are involved, people systematically weigh credentials differently based on their unconscious biases. For example, an experiment systematically varied the race of an applicant and the applicant’s credentials for college admission and asked White participants how strongly they supported admitting the student to their college (Hodson, Dovidio, & Gaertner, 2002). When candidates for admission had mixed credentials (strong high school grades but modest standardized scores, or vice-versa), participants perceived the credential that White candidates were stronger in relative to Black candidates as being the more valid predictor of success in college. This differential weighting of the credentials, in turn, justified students’ stronger recommendations of White than Black candidates for admission (see also Brief, Dietz, Cohen, Pugh, & Vaslow, 2000, for employment bias against Blacks; Rooth, 2007, for hiring biases against Muslims).

Son Hing, Chung-Yan, Hamilton, and Zanna (2008) investigated another form of racial bias, discrimination against Asian job applicants in Canada, using a paradigm related to research on aversive racism and selection decisions (Dovidio & Gaertner, 2000). Specifically, participants in the Son Hing et al. study reviewed the materials for a White or Asian job applicant, who was strongly or only moderately qualified for the position, and offered a recommendation for hiring the person for the position. Paralleling the findings of subtle bias against Blacks in the United States, these researchers found that when assessing candidates with more moderate qualifications, evaluators recommended White candidates more strongly for the position than Asian candidates with identical credentials. However, when evaluating candidates with exceptionally strong qualifications, no such selection bias emerged. Moreover, the researchers found that implicit bias against Asians (as measured by an IAT), but not explicit prejudice, predicted weaker support for hiring Asian candidates who had moderate qualifications. However, when the Asian candidate had distinctively strong qualifications (and a failure to hire the applicant could not be justified on the basis of factors other than race) neither implicit nor explicit prejudice predicted the hiring decision, which generally supported the hiring of the Asian applicant.

Although there is some controversy about the predictive validity of the IAT (Greenwald et al., 2009; cf. Oswald, Mitchell, Blanton, Jaccard, & Tetlock, 2013), key factors in the relative validity of implicit and explicit measures for predicting behavior are the context in which the behavior occurs and the type of behavior being examined (Dovidio, Kawakami, Smoak, & Dovidio, 2009). Explicit attitudes generally shape deliberative, well-considered responses in which the costs and benefits of various courses of action are weighed. Implicit attitudes typically influence actions that people do not view as an overt expression of bias (e.g., decisions that can be justified on the basis of some factor other than race, as in Son Hing et al., 2008) or less controllable responses (e.g., nonverbal behaviors).

Even subtle expressions of bias such as through nonverbal behavior, however, can have significant social impact. For example, Dovidio, Kawakami, and Gaertner (2002) studied initial interactions between Whites and Blacks and found that whereas Whites’ explicit (self-reported) racial attitudes predicted their relatively controllable verbal expressions in their interactions with Blacks, Whites’ implicit attitudes, which were generally negative, predicted their nonverbal behaviors. Moreover, Black interaction partners weighed the nonverbal behavior more heavily than the verbal behavior in their impressions of the White partner and the interaction. Thus, Whites and Blacks had divergent perspectives in their interactions, and Blacks’ awareness of conflicting positive verbal and negative nonverbal behavior undermined how trustworthy they saw the White interaction partner.

These dynamics associated with implicit biases in certain settings, such as in health care, can ultimately have substantial impact on the well-being of Blacks. White physicians generally perceive themselves not only as nonprejudiced and color-blind (Epstein, 2005; Sabin, Rivara, & Greenwald, 2008) but also, based on their responses to the IAT, harbor negative implicit racial biases toward Blacks (Sabin et al., 2008; Sabin, Nosek, Greenwald, & Rivara, 2009). Moreover, physician implicit biases predict medical recommendations representing lower quality of coronary care for Black patients (Green et al., 2007) and less willingness to prescribe narcotics to ease the pain of Black patients (Sabin & Greenwald, 2012) in ways independent of explicit racial bias. Also, consistent with research on the influence of implicit racial bias in social interactions (Dovidio et al., 2002), in actual medical interactions doctors higher in implicit bias speak faster to and have shorter visits with Black patients (Cooper et al., 2012), and they display less warmth in their medical interactions (Penner et al., 2010). Overall, physicians higher in implicit bias are less patient-centered in their care of Black patients (Blair et al., 2013; Cooper et al., 2012).

Black patients’ responses to these interactions directly relate to the doctor’s implicit bias. Following their interactions, Black patients have less respect for, confidence in, and trust in the advice of medical professionals higher in implicit bias (Cooper et al., 2012; Dovidio et al., 2008). This distrust predicts lower levels of adherence to the doctor’s prescriptions 16 weeks after the medical interaction (Penner, Gaertner, et al., 2013). Thus, among highly educated individuals in helping professions, who appear nonprejudiced and deeply committed to the welfare of their patients, implicit biases produce systematic disparities in healthcare and, ultimately, contribute to racial disparities in health (Penner, Hagiwara, et al., 2013).

Color-Blind Inclusiveness

The challenge of combating unconscious biases is that people are often not aware that they possess these biases, and when they consciously monitor their behavior, their actions reinforce their conscious egalitarian self-image. How, then, can the effects of unconscious biases be addressed? To the extent that socially categorizing people into different groups automatically activates stereotypes and prejudice toward members of those groups and increases affinity to members of one’s in-group (Dovidio & Gartner, 2010), one recent technique has targeted a psychological root of the problem: social categorization. The principle behind the common in-group identity model (Gaertner & Dovidio, 2000, 2012) is that inducing people to recategorize in-group and out-group members within a common, superordinate (overarching) identity redirects those motivational and cognitive processes that produce in-group-favoring biases to increase positive feelings, beliefs, and behaviors toward others who were previously regarded primarily in terms of their out-group membership. A superordinate identity may be, for example, a common school, organization, or national identity. The common in-group identity model for improving intergroup attitudes has received considerable empirical support internationally (see Gaertner & Dovidio, 2012).

Whereas much of the research to date has focused on the benefits of creating a single, common in-group identity, the development of a common in-group identity does not necessarily require each group to forsake its original, less inclusive group identity. Depending on their degree of identification with the different categories and contextual factors that make particular identities salient, individuals may activate two or more of their multiple social identities simultaneously (Roccas & Brewer, 2002) or sequentially (Turner et al., 1987). For example, people can conceive of two groups (e.g., science and art majors) as distinct units within the context of a superordinate social entity (e.g., university students). Whereas a common group identity is related to color-blindness and assimilation acculturation ideologies (in which immigrant or majority group members are expected to conform to the norms and values of the dominant culture), a dual identity reflects perceptions integral to a multicultural ideology (in which the norms and values of different groups are respected within an overarching cultural identity).

The mutual intergroup differentiation model (Brown & Hewstone, 2005) relatedly proposes that positive social contact between members of different groups produces particularly robust improvements in intergroup attitudes when different groups maintain their separate identities but have cooperatively interdependent relations. As noted earlier, cooperative interdependence improves intergroup attitudes in part because it helps create a superordinate group identity (Gaertner, Mann, Dovidio, Murrell, & Pomare, 1990). Different from the mutual intergroup model, however, the common in-group identity model further posits that the superordinate identity component of a dual identity can be established in other ways, such as emphasizing identities that groups have in common (e.g., a national or university identity), even in the absence of actual cooperative activity. From the perspective of the common in-group identity model it is the simultaneous salience of separate and superordinate group identities, not the particular mechanism that achieves this, that is important for reducing intergroup bias.

Although recategorization, both in terms of substituting separate group identities with a common in-group identity or creating dual identities, can produce more positive intergroup attitudes (see Gaertner & Dovidio, 2009, for a review), in the remainder of this article we illustrate how there can be a “darker side” of intergroup harmony achieved solely by emphasizing common identity (and related cultural ideologies of color-blindness and assimilation). Recategorization in terms of a dual identity (and the related cultural ideology of multiculturalism), by contrast, can promote positive attitudes and action by members of both high-status and low-status groups. Moreover, we propose that recategorization in the form of a single common in-group identity (vs. a dual identity), relates directly to the concept of color-blind racism (Bonilla-Silva, 2001, 2003), and may be a strategy employed by members of majority group members to reinforce the status quo that benefits their group. Creating a sense of common identity can deflect attention away from group-based disparities, reducing the likelihood that members of high-status groups perceive social injustice, and promote a feeling of harmony and optimism that undermines collective action by members of low-status groups.

In the next two sections, we examine (a) differences in preferences of members of high-status and low-status groups for different representations (one group or dual identity) and associated cultural ideologies (color-blind or multicultural) and (b) how a one-group representation and dual identity can both improve attitudes but have different effects on action to achieve social equality.

Group Status and Representation Preferences

In his classic acculturation framework, Berry (1997; see also Sam & Berry, 2010) presents four forms of cultural relations in pluralistic societies that represent the intersection of “yes–no” responses to two relevant questions: (a) Are cultural identity and customs of value to be retained? (b) Are positive relations with the larger society of value, and to be sought? These combinations reflect four adaptation strategies for intergroup relations: (a) integration, when cultural identities are retained and positive relations with the larger society are sought; (b) separatism, when cultural identities are retained but positive relations with the larger society are not sought; (c) assimilation, when cultural identities are abandoned and positive relations with the larger society are desired; and (d) marginalization, when cultural identities are abandoned and are not replaced by positive identification with the larger society.

Although this framework has been applied primarily to the ways in which immigrants acclimate to a new society, it can be adapted to apply to intergroup relations between high-status and low-status groups generally (see Dovidio, Gaertner, & Saguy, 2007). Substituting the separate strengths of the subgroup and subordinate group identities for the answers to Berry’s (1997) two questions, the combinations map onto the four main representations considered in the common in-group identity model: (a) dual identity (subgroup and superordinate group identities are high, like integration); (b) different groups (subgroup identity is high and superordinate identity is low, like separatism); (c) one group (subgroup identity is low and superordinate group identity is high, like assimilation; and (d) separate individuals (subgroup and superordinate group identities are low, like marginalization).

Two of the ideologies that have received the most attention in the study of intergroup relations are assimilation, which involves a form of common identity, and integration in terms of multiculturalism, which reflects a dual identity. Assimilation and multiculturalism have often been considered oppositional ideologies. As noted earlier, assimilation requires minority-group members to conform to dominant values and ideals, often requiring the abandonment of racial or ethnic group values, to achieve full citizenship and be accepted in society. Multicultural integration, by contrast, strives to be inclusive by recognizing, and often celebrating, intergroup differences and their contributions to a common society.

Research in the area of immigration suggests that members of the host society (the high-status group) and immigrant groups (low-status groups) have different preferences for assimilation and multicultural integration. For example, Verkuyten (2006) summarized the results of eight studies of adolescents and young adults in Europe, consistently finding that minority group members supported multiculturalism (integration) more than did majority group members. These preferences also apply to the preferences of Whites and racial and ethnic minorities: In the United States, Whites prefer assimilation, whereas racial and ethnic minorities favor multiculturalism (Ryan, Hunt, Weible, Peterson, & Casas, 2007).

Moreover, in intergroup interactions, members of high-status and low-status groups are motivated to shape the discourse in ways that emphasize their preferred representation. In two studies, one with laboratory groups varying in control over a valued resource (extra credit for experimental participation) and the other with ethnic groups varying in status in Israel (Ashkenazim, high status; Mizrahim, low status), participants were asked to indicate their preference for topics, identified based on pilot work as focusing on group commonalities or differences, to be discussed in an upcoming interaction with members of the other group (Saguy, Dovidio, & Pratto, 2008). Whereas members of high-status groups preferred discourse that focused virtually exclusively on commonality rather than on group differences, members of low-status groups showed equivalently strong preferences for talking about commonality and difference (the two critical elements of a dual identity). Moreover, the effect of group status on desire to talk about differences between the groups was based on a motivation for changing group positions toward equality. That is, low-status group members’ greater preference to discuss points of difference, relative to that of high-status group members, occurred because they had a greater motivation for a change in the power structure.

The strategic nature of these different preferences for members of high-status and low-status groups is suggested by three other sets of findings that illustrate the dynamic nature of preferred group representations and intergroup ideologies. First, the results from a pair of studied by Saguy and Dovidio (2013) revealed that members of high-status and ethnic majority-group members had a much stronger preference to avoid talking about group differences relative to commonalities (suggesting stronger endorsement of assimilation or color-blindness) when their advantaged status was less secure. Members of a low-status group maintained a strong desire to talk about commonalities across conditions but demonstrated a greater preference to talk about group differences, increasing the salience of subgroup identities while simultaneously emphasizing commonality, when hierarchical relations were insecure (i.e., unstable and illegitimate) compared with secure. That is, greater instability of intergroup status relations affected members of high-status and low-status groups in ways that created even greater divergence in their topic preferences.

Second, Hehman et al. (2012) studied the preferences of Whites and Blacks at two public universities in the United States, one a state college in which Whites represent the majority (85%) of the student body and one a historically Black college in which Blacks are the majority (76%). This contextual status significantly affected identity preferences. White students showed a much stronger preference for multiculturalism when they were in the institutional minority than in the majority; Black students exhibited stronger of endorsement of assimilation when they were the institutional majority than when they were the minority.

Third, if color-blindness is strategic, even if not consciously so, when the status quo becomes unstable, Whites preferences of color-blind policies (and Blacks’ preference for multicultural policies) would be expected to intensify. Exploring this possibility, we examined students’ preferences for university policies three times over an academic year (Dovidio et al., 2007). The first time was at the beginning of the semester, when race relations were perceived to be relatively positive and stable. The second time was near the end of the first semester after a series of racial incidents threatening Blacks (e.g., racial graffiti on campus, several alleged episodes of verbal harassment of Black students). The third time was in the middle of the second semester, when race relations were perceived to be less tense and volatile.

Across these three time periods, we assessed, longitudinally, White and Black students’ support for policies that reflected efforts associated with one group (assimilationist), dual identity (multicultural integrationalist), and separate groups (separatist) initiatives. These policies, which were developed through pilot testing, were not directly related to the racial incidents that occurred. Example of one group policies were, “The university should devote more funds to common activities for all students” and “Students in their first year should be assigned roommates on a random basis.” Examples of dual identity policies were, “The university should devote more funds to multicultural activities on campus” and “Minority students may choose to have a roommate of their same race or ethnicity in their first year, but there should not be separate minority dormitories.” Separatist policies were, “The university should devote more money to activities to groups to support their different racial or ethnic identities” and “Minority students should be allowed to have their own dormitory.”

Even before the racial incidents occurred, majority and minority students showed differential support for one group, dual identity, and separate-group policies. Consistent with the findings reported earlier, Whites exhibited a stronger preference for one group policies than did minority-group students, whereas minority students showed a stronger preference for dual identity policies. For both Whites, and minorities, separatist policies were least supported, and there were no differences between the groups in level of support. During the second measurement period, when racial tensions were high, these racial differences in support for one group and dual identity policies were significantly magnified, and Blacks more strongly supported separate-group policies than did Whites and at a level somewhat higher than their support for one group initiatives. By the third assessment period, when racial tensions had substantially subsided, the pattern of policy preferences approached what it was at the beginning of the year. Whether the different policy-related responses of majority and minority students were conscious strategies or an unconscious reaction to the events is unclear from these data. Nevertheless, the overall pattern is consistent with our hypothesized difference in the goals of the groups.

In summary, these findings converge to reveal that members of high-status groups, who are motivated to maintain the status quo, show a preference for focusing on commonalities to the exclusion of differences and greater support for assimilation over multiculturalism. Members of low-status groups, who desire to alter the status quo to improve their group’s hierarchical position, exhibit a greater desire to talk about differences between the groups but, at the same time, to discuss commonalities between the groups and show greater endorsement of multiculturalism compared with assimilation. These different perspectives systematically relate to the status each group occupies in a particular context and the stability of status relations in that context.

The Irony of Harmony

Thus far we have discussed the different preferences of members of high-status (or majority) and low-status (or minority) groups for different group representations and ideologies, as well as the conditions and goals that moderate these preferences. In this section, we examine the possibility that, to the degree to that factors that promote common identity (e.g., positive intergroup contact), assimilation, or color-blindness reduce attention to structural inequality as they promote positive attitudes toward members of the out-group, they may undermine the motivation of both majority- and minority-group members to engage in action for social change to achieve true equality. This is a potential “irony of harmony.”

As we noted earlier, the subtle nature of contemporary bias typically limits recognition of unfair treatment. Subtle bias is more difficult to detect and respond to than blatant bias even for low-status group members (Ellemers & Barreto, 2009), and high-status group members are less attuned to cues of subtle discrimination than are low-status group members (Salvatore & Shelton, 2007). Thus, focusing only on commonalities between groups may decrease the likelihood that high-status group members will recognize and respond to injustice, particularly in the form of subtle bias, against minority-group members (see Saguy & Chernyak-Hai, 2012).

We (Banfield & Dovidio, 2013) examined this issue in a study in which White participants in the United States were exposed to a manipulation that emphasized common-group (American) identity of Blacks and Whites, separate racial-group memberships, or a control condition that did not emphasize identities. Participants then read a hiring scenario that involved either subtle or blatant discrimination, in which a Black candidate was not offered a job (Salvatore & Shelton, 2007). The outcomes of interest were perceptions of discrimination and expressions of willingness to protest on behalf of the applicant who was denied the job.

As expected, when the bias witnessed was subtle, White participants for whom common identity was emphasized perceived lower levels of bias than those for whom separate identities were emphasized or those in a control condition, and these perceptions mediated less willingness to protest the negative outcome for Black person who was disadvantaged. By contrast, when discrimination was blatant, emphasizing common identity produced somewhat greater perceptions of bias and somewhat more willingness to engage in collective action.

We further investigated the irony of harmony on the perceptions and responses of members of low-status, socially disadvantaged groups. One study experimentally examined the causal effect of a commonality-focused encounter, relative to a difference-focused interaction, on low-status group members’ out-group attitudes, attention to inequality, and expectations of out-group fairness, as well as on high-status group members’ intergroup orientations and resource allocation. The second and third studies generalized and extended the findings, specifically with respect to minority groups, by examining the relation of positive intergroup contact to attitudes, perceptions of inequality and out-group fairness, and support for social change in two naturalistic intergroup contexts.

The laboratory study (Saguy, Tausch, Dovidio, & Pratto, 2009; Study 1) manipulated power between two randomly assigned groups by giving the high-status group the position of assigning extra course credits to the two groups (see also Saguy et al., 2008). Before the members of the high-status group allocated the credits, members of both the high-status and low-status groups interacted with instructions to focus on either intergroup commonalities or differences.

As expected, interaction focusing exclusively on things the groups had in common produced more positive intergroup attitudes for high-status and low-status group members than did discussions about differences between the groups. In addition, for both groups, attention to inequality between groups was lower when the interaction focused on commonalities. Moreover, in part because they had more positive feelings about the other group, members of the low-status group expected the high-status group to be fairer in allocating the resources and to distribute the credits in a more equitable fashion following discussions about commonalities, rather than about differences.

However, when the low-status group members’ expectations were compared with the high status group’s actual allocation, there was a significant discrepancy. As the members of the low-status groups anticipated, high-status groups were substantially biased against the low-status groups in the allocation of credits after they talked about differences between the group but, unexpectedly from the perspective of low-status group members, high-status groups were just as biased in allocating the credits after discussions about what the groups had in common. The more positive intergroup attitudes of high-status group members when they talked about commonalities than about differences with members of the other group did not translate into more material support to achieve equality, and the high-status groups’ allocation fell significantly below what low-status groups anticipated.

The other two studies in this set investigated these processes in two different cultural contexts. A study of Arabs in Israel (Saguy et al., 2009; Study 2) examined the statistical associations among friendships with Jews (a type of positive contact that is particularly likely to involve a focus on commonalities), attitudes toward Jews, awareness of inequality, and perceptions of Jews as fair. It further measured Arabs’ support for social change toward equality. Saguy et al. hypothesized that, because less attention to illegitimate aspects in the inequality and beliefs that progress will be made through out-group fairness can reduce personal motivations for action, such factors would relate to weaker support for social action for change among Arabs.

Consistent with the results of the laboratory experiment, more positive contact with Jews was associated with more positive attitudes toward Jews and with reduced awareness of inequality between Jews and Arabs. In addition, improved attitudes were associated with increased perceptions of Jews as fair. Moreover, and consistent with our theorizing, both perceptions of Jews as fair and reduced awareness of inequality were associated with reduced support for social change. Thus, through its effects on the way low-status group members viewed social inequality and members of the other group, positive contact was associated with a decrease in support for social change.

Results of the study of Muslims in India (see Saguy, Tausch, Dovidio, Pratto, & Singh, 2011) replicated these findings. Having more Hindu friends was related to the improved attitudes of Muslims toward Hindus, but it also reduced awareness of inequality between Muslims and Hindus. In addition, these outcomes predicted stronger perceptions of Hindus as fair, which in turn were related to weaker collective action tendencies (measured as intentions to participate in various actions that could improve the position of Muslims in India).

As Bonilla-Silva (2001) suggests, because of cultural frames and racial stories that often justify disparities, structural discrimination is more difficult to recognize than specific incidents of discrimination (see also Van Zomeren, Postmes, & Spears, 2008). Structural disadvantage refers to a status quo of inequality. When disadvantage is structural, embedded in the status quo, members of low-status groups are susceptible to the effects of system-justifying ideologies (for example, which rationalize the different status of the groups based on the different efforts or competencies of the groups) and come to believe that their group deserves less favorable outcomes than high-status groups (Jost, Pelham, Sheldon, & Ni Sullivan, 2003). Thus, structural discrimination represents another form of ambiguous bias.

Accordingly, we (Ufkes, Calcagno, Glasford, & Dovidio, 2013) conducted a study focusing on the structural disadvantage of racial/ethnic minorities in the United States, in which we varied the emphasis on their common (U.S.) identity with Whites, their separate subgroup (racial/ethnic identity), or their dual identity (Black-American or Latino-American identity). We also assessed two key pathways to collective action identified by Van Zomeren et al. (2008): feelings of anger (associated with perceived injustice) and collective efficacy (reflecting beliefs that collective action will produce change). Emphasizing common identity uniquely led to low levels of anger and lower perceptions that collective action by their minority group would effectively accomplish change. Both of these perceptions, in turn, predicted lower motivation to engage in collective action to address structural inequality.

To summarize this line of research, the fact that emphasis on commonality or color-blind and assimilation perspectives leads members of minority and majority to attend less to group-based disparities may be particularly problematic for responding to these inequities given the nature of contemporary bias. The subtlety of contemporary bias makes it particularly likely that unfair disadvantage will be overlooked or dismissed when there is a focus on common identity (or, relatedly, assimilation or color-blindness). Focusing on only common identity distracts attention against group-based disparities, even when these disparities are detected by members of low-status groups, feelings of common identity can reduce motivation to take action because of greater trust in the system (e.g., Kay et al., 2009). Thus, common identity may particularly undermine collective action by low-status group members and interventions by high-status group members on behalf of low-status groups when the operation of bias is ambiguous.

Social change toward equality depends not only on the actions of members of low-status groups but also involves the support and potential initiative of members of high-status groups. Although collective action by low-status group members can achieve significant social change, such action by high-status group members may be even more effective because of the greater resources they possess and the potential of facing less resistance from other majority-group members. Indeed, when members of high-status groups recognize that the disadvantage of low-status groups is unfair, they are genuinely motivated to restore equity (Saguy et al., 2008; Study 2).

Implications of Individual-Level Processes for Society

The focus of the present article has been on the processes that perpetuate racial bias in its contemporary form among Whites, in terms of their assessments, decisions, and interactions. This bias is typically exhibited in subtle ways, often cloaked by a commitment to color-blindness at a personal and societal level. We note that these individual-level forms of bias, such as aversive racism, are related to, but not synonymous with, Bonilla-Silva’s (2001, 2003) conception of color-blind racism as societal-level “frames, stylistic components, and racial stories” (see Bonilla-Silva & Dietrich, 2011, p. 191). In addition, we acknowledge that the research we have described suggests how micro-level action frames and dynamics can contribute to systemic racism, but direct empirical evidence of these effects is currently limited. However, a psychological perspective on these individual-level processes can help illuminate how individual actions can both reflect and reinforce racial frames and stories that influence social structures.

We also note that the social frame represented by color-blind racism represents a cultural influence that affects the thoughts and actions of minority-group members, as well as majority-group members, in ways that reinforce the dominance of advantaged groups socially. Bonilla-Silva and Dietrich (2011) observed that “the ideology of color-blindness is increasingly affecting even those who are at or near the bottom of the economic and social hierarchies in the United States: blacks and Latinos” (p. 195). The research we reviewed further demonstrates that adopting a color-blind frame can distract members of both disadvantaged and advantaged groups from attending to group-based inequalities that promote action to achieve a truly fair and equitable society.

A social psychological approach offers a complementary perspective to the structural and societal-level approach represented by Bonilla-Silva (2001, 2003; see also Bonilla-Silva & Dietrich, 2011) by identifying the general individual-level mechanisms and propensities, which may operate cross-culturally, that underlie the transmission of these cultural frames and the concrete operation of these frames in social interaction. Social psychological research has revealed a tendency to perceive prevailing social structures as what “should” be (Kay et al., 2009) and a general motivation to preserve the status quo, even at a high cost one’s group (see system justification theory; Jost, Banaji, & Nosek, 2004).

These processes can help explain why members of disadvantaged groups are likely to endorse color-blind racism and may offer new insights. For example, people who feel more dependent on the current social system, including those disadvantaged by it, are more likely to be susceptible to these effects. In one study (Kay et al., 2009; Study 3), for instance, female Canadian read a brief description of the responsibilities of Canadian members of Parliament, accompanied by a graph showing that currently only 20% of the members of Parliament were women. As hypothesized, women who were led to believe that they were more highly dependent on the government were more likely to defend the status quo. These women were less likely than those who believed that they were low in dependency to endorse statements that there should be more women in politics and in Parliament. From their perspective, having women account for only 20% of the members of Parliament was fair and acceptable. Women were highly motivated to preserve the status quo, even at a high cost to themselves and fellow group members. In a subsequent study (Kay et al., 2009; Study 4), women who felt more dependent on the current social system, and thus more motivated to justify the current social system, actively derogated a woman whose ambitions in business threatened the status quo of gender relations.

The research reviewed in this article also indicates that although inclusive feelings of superordinate identity promote harmonious relations between members of different groups, this inclusiveness itself does not guarantee fair treatment and may, in fact, contribute to the perpetuation of unfair treatment of racial and ethnic minority-group members. To the extent that the superordinate identity is defined by the standards and attributes of the dominant group, which it typically the case (Mummendey & Wenzel, 1999), members of nondominant groups included within that identity are vulnerable to being perceived as deviant (Waldzus, Mummendey, Wenzel, & Boettcher, 2004) or allowing their subgroup identity to become invisible, not only to others but for themselves. Thus, they may experience immediate benefits of being able to “pass,” increasing their personal chances of success but at the sacrifice of the psychological buffering effects of subgroup identity (Branscombe, Schmitt, & Harvey, 1999; see also Schmitt, Branscombe, Postmes, & Garcia, 2014), social support, and special support to other members of their subgroup. At the same time, society may fail to reap the creative benefits that diversity has to offer (Antonio et al., 2004).

Conclusion

In conclusion, color-blindness may have broad appeal in principle because it seems to represent, at least superficially, the core American value of equal treatment and is consistent with an acculturation tradition emphasizing assimilation and being a “melting pot.” In practice, understanding the nature of subtle bias at the individual level and the automatic processes that underlie it can help illuminate how well-meaning interventions can create a veneer of tolerance while reinforcing structures that perpetuate traditional racial hierarchy in the United States. At an individual level, emphasizing color-blindness through commonality may represent a valuable step toward reducing intergroup tensions and developing trust and intimacy between members of different groups. However, at a fundamental level, adopting this perspective permits the perpetuation of systems that reinforce racial dominance that appear to be fair on the surface but which are actually unjust. Addressing racial inequalities at both the societal- and individual-levels requires being race conscious, not color-blind, to recognize racial disparities and to understand their basis in unfair treatment. Good intentions, if not appropriately informed, are not enough to achieve fairness socially, institutionally, and interpersonally (Kalev, Dobbin, & Kelly, 2006).