Perceptions of women of color who claim compound discrimination: Interpersonal judgments and perceived credibility

Abstract

People of color who attribute rejection to racism and women who attribute rejection to sexism are perceived as troublemakers. Women of color may encounter racism and sexism simultaneously; however, it is unclear how compound discrimination claims are perceived. We examined interpersonal judgments of claimants and perceptions of the credibility of compound discrimination claims. In contrast to the double jeopardy perspective, which predicts that the consequences of multiple stigmas are cumulative, a Black woman (Study 1) was not perceived as a bigger troublemaker when she attributed rejection to compound discrimination versus either racism or sexism. Instead, racism and compound discrimination claims incurred similarly high interpersonal costs. Likewise, an Asian woman (Study 2) was not perceived as less credible when she attributed rejection to compound discrimination versus either racism or sexism. Instead, compound discrimination was the only discrimination attribution reliably judged as more credible and appropriate than baseline.

Keywords

attribution discrimination intersectionality multiple identities

Employment discrimination is prohibited by Title VII, a federal statute that protects the rights of employees who are members of stigmatized groups. In the five decades that have passed since the inception of Title VII, courts have debated whether race and gender discrimination claims represent distinct charges that should be filed independently (Carbado, 2000). For example, in Degraffenreid v. General Motors Assembly Division, Black female employees sued General Motors because they claimed that the seniority-based promotion system disadvantaged Black women, who tended to be hired more recently than other employees. At the time, the court ruled that Black women are not members of a unique marginalized group entitled to protections not given to Black men and White women. However, recent court rulings acknowledge that treating racism and sexism claims as independent charges privileges people with a single stigmatized identity in antidiscrimination law. Women of color who have been overlooked for promotions given to men of color and White women will have a difficult time proving that employers were racist or sexist because both people of color and women were promoted within those organizations (Carbado, 2000). Thus, compound discrimination—in which an individual encounters two or more forms of discrimination simultaneously—may best describe some stigmatizing experiences of women of color (Carbado, 2000; Purdie-Vaughns & Eibach, 2008). Here, we use the term compound discrimination to refer to marginalization due to race and gender. However, other forms of discrimination are experienced concurrently as well.

As the law shifts to accommodate claims of compound discrimination, it is important to understand how members of multiple stigmatized groups (or multiply stigmatized targets) are perceived when they speak out about such experiences. Interpersonal judgments—by friends and colleagues, for example—directly affect the willingness of stigmatized individuals to report experiences of discrimination. Stigmatized people are often reluctant to report prejudice because they worry about being seen as complainers (Kaiser & Miller, 2003; Shelton & Stewart, 2004). As a result, victims of prejudice miss out on opportunities to seek compensation, and they may also feel as if they are disguising their true selves (Shelton, Richeson, Salvatore, & Hill, 2006). In addition, when prejudiced actions go unreported, perpetrators may assume that their behaviors were appropriate or justified (Czopp & Monteith, 2003). Perceptions of claim credibility are also important. Some perceivers, such as human resource officers and jurors, are in positions to undermine fair scrutiny of a discrimination claim if they perceive it as lacking credibility. For example, jurors may regard a victim’s decision to report discrimination as inappropriate if they view the accused individual as a generally egalitarian person (Kaiser et al., 2013). Thus, studying perceptions of compound discrimination claims will inform our understanding of the amount of support that women of color may receive from others when they attribute negative outcomes to racism and sexism simultaneously.

Perceptions of Discrimination Claimants: An Intersectional Analysis

Research to date has examined consequences for people of color who attribute rejection to racism (Kaiser & Miller, 2003) and women who attribute rejection to sexism (Stangor, Swim, van Allen, & Sechrist, 2002). Treating racism and sexism as if they occur separately may seem to allow researchers to isolate the consequences of being a person of color or a woman. However, by viewing race and gender as separate identities, we fail to consider intersectionality: the contextualization of one social identity by the other social identities an individual possesses (Warner, 2008). Individuals experience their social identities in combination and in ways that are unique to the intersections at which they find themselves (Bowleg, 2008). Intersectionality describes how identities interact, such that experiences related to one identity (e.g., race) are shaped by experiences related to another identity (e.g., gender; Warner, 2008; Woods-Giscombé & Lobel, 2008). Some scholars suggest that marginalization due to race and gender accumulates in an additive fashion (Beale, 1979). Others suggest that, when stigmatized identities are combined, new elements of experience emerge that represent neither the addition, nor the subtraction, of those identities (Kunda, Miller, & Claire, 1990).

Consistent with an additive account of multiple stigmatization, double jeopardy theory predicts that the negative outcomes of belonging to disadvantaged race and gender groups are compounded for women of color (Beale, 1979). Research on discrimination experiences supports this hypothesis. In one study, White, Asian, Black, and Latino male and female employees completed measures of sexual and ethnic harassment in the workplace. Responses to these measures were averaged to yield reports of general discrimination. The results revealed that women of color reported experiencing more overall discrimination than White men, men of color, and White women (Berdahl & Moore, 2006).

Other lines of research, however, suggest that the properties of intersectional categories cannot be understood by studying the relevant component categories separately, or by adding together information about these categories (Kunda et al., 1990). Research on stereotype content illustrates this point: for example, Black women are judged according to stereotypes, such as “promiscuous,” not attributed to Black individuals or to women separately (Ghavami & Peplau, 2012). Therefore, by studying the categories “Black” and “women” separately, we may learn very little about the intersectional category, “Black women.” In addition, the model of intersectional invisibility proposes that prototypical members of stereotyped groups possess only one stigmatized identity. Men of color are therefore considered prototypical people of their race, and White women are considered prototypical of their gender. Consequently, women of color are rendered socially invisible by their nonprototypical status (Purdie-Vaughns & Eibach, 2008). In some cases, invisibility leads to less recognition of the contributions and achievements of women of color (Sesko & Biernat, 2010). In other cases, however, invisibility reduces the extent to which women of color are judged according to negative stereotypes more often applied to prototypical group members (Livingston, Rosette, & Washington, 2012). A woman of color who attributes rejection to her multiple subordinate identities may highlight her nonprototypical status, reminding perceivers that she does not represent the prototypical racism claimant (a man of color), nor the prototypical sexism claimant (a White woman). As a result, perceivers may not penalize a compound discrimination claimant more than a racism or a sexism claimant. Broadly, such findings would support the perspective that the properties of intersectional categories are qualitatively different from the sum of the properties of the relevant component categories (Bowleg, 2008).

Given that some negative consequences of multiple stigmatization are cumulative, and that other negative consequences are not cumulative, it is important to study compound discrimination claims directly to learn how such claims are perceived. In two studies we investigated whether the negative consequences of reporting discrimination are compounded for women of color who report racism and sexism simultaneously. Women of color who attribute rejection to compound discrimination may incur greater interpersonal costs, and be judged as behaving less appropriately because their claims are less credible, than women of color who attribute rejection to either racism or sexism independently. In contrast, reporting compound discrimination may not be associated with more negative consequences than reporting either racism or sexism independently.

Interpersonal Judgments and Perceptions of Credibility

Underreports of discrimination perpetuate the myth that prejudice is no longer a problem in our society (Czopp & Monteith, 2003; Kaiser, Drury, Spalding, Cheryan, & O’Brien, 2009). However, stigmatized people who experience discrimination often remain silent about their experiences in order to avoid seeming like troublemakers (Kaiser & Miller, 2003; Sechrist, Swim, & Stangor, 2004). Shelton and Stewart (2004) studied this reluctance to confront discrimination by examining whether women are less likely to confront sexism in high- versus low-cost situations. Female participants interviewed for a prestigious, competitive, high-paying position (high cost) or a less prestigious position that did not pay well and that the experimenter described as easy to obtain (low cost). Women in this study were more likely to confront a sexist interviewer when they faced low versus high costs of confronting. The authors concluded that women who expected to miss out on opportunities if interviewers disliked them strategically refrained from reporting sexism.

Research suggests that majority group members label discrimination claimants as troublemakers, in part, because high-status group members are motivated to believe that individuals are responsible for their standing in society. To maintain this belief in the face of discrimination reports, discrimination claimants are perceived as routinely excusing a personal lack of effort with grievances of discrimination. Thinking of stigmatized people in this way recasts inequalities as legitimate group differences as opposed to injustices (Kaiser, Dyrenforth, & Hagiwara, 2006). Regardless, then, of whether stigmatized individuals report one type or a combination of types of discrimination, they may incur similar interpersonal costs (operationalized as “troublemaker” judgments) that punish their unwillingness to take personal responsibility for failure (Kaiser et al., 2006). Finding that women of color incur similar interpersonal costs for attributing rejection to single and compound forms of discrimination would rebuff assumptions that the negative consequences of multiple stigmas are cumulative (Beale, 1979).

Discrimination claims that appear to lack credibility may be ignored or undermined once victims do gather the courage to file reports, and discrimination claims may be perceived as more or less credible depending on the social context. For example, employees of color who report racism are perceived as having less credible claims if the organizations they work for otherwise promote diversity. Within such organizations, colleagues are unlikely to support racism claimants because they assume that their employers value equality (Kaiser et al., 2013). Here, we examined perceptions of credibility, in addition to interpersonal judgments, because of the potential for each construct to predict different real-world outcomes. Interpersonal judgments may relate to relational responses to discrimination claims, such as not wanting to work with a claimant (Shelton & Stewart, 2004). Perceptions of credibility may relate to justice-serving responses, such as believing that a candidate responded appropriately to a legitimate threat and, as a result, recommending fair financial remedies (Blodorn, OʼBrien, & Kordys, 2011; Kaiser & Major, 2006). Thus, recognizing that, in a particular case, a stigmatized person was justified in speaking out against mistreatment may not preclude people from viewing that person as, generally, a troublemaker.

Research Overview

In two studies, we examined perceptions of women of color who attributed rejection to racism and sexism simultaneously. Participants read about Black and East Asian women who attributed rejection in an interview setting to racism, sexism, compound discrimination, or to their own interview skills. We investigated whether the negative consequences of reporting discrimination accumulate for women of color who report compound discrimination. In Study 1, we examined whether a Black woman was judged as more of a troublemaker when she reported compound discrimination than when she reported a single form of discrimination. In Study 2, we examined whether an Asian woman was perceived as behaving less appropriately because she faced a less credible discrimination threat when she attributed rejection to compound discrimination versus a single form of discrimination.

Study 1: Interpersonal Judgments

Method

Participants and design

Participants were 112 undergraduate students (66 women) who completed the study for course credit. On average, participants were 18.58 years old (SD = 2.50). Seventy-two of the participants were White, 20 were Asian, 8 were Black or African American, 5 were multiracial, 2 were Native American, 2 described themselves as “other,” and 3 declined to answer this question. The study involved one between-subjects factor (candidate attribution) with four conditions: racism, sexism, compound discrimination, and interview skills.

Procedure

The experimenter informed participants that the purpose of the study was to examine perceptions of job interview situations. Participants were instructed to read a portfolio about a job interview, which they were told was randomly selected from a large database of similar portfolios, and contained materials from an actual interview. In fact, the portfolio was contrived for the purpose of the study and modeled after materials used in past research (Kaiser & Miller, 2003; Remedios, 2015). The experimenter instructed participants to read through the portfolio carefully and to remember as many details as possible. The portfolio included information about the interviewer, who was described as a professionally dressed man with blond hair and fair skin (indicating that he was a White), as well as the interview room, which was described as bright and simple, with a large desk and comfortable chairs. The ad was for a retail sales management position. The candidate’s application for the position was described as above average, and participants learned that she was one of eight candidates interviewing for the position. Participants also saw a photo of a young, professionally dressed Black woman who was revealed to be the candidate.

Next, participants saw the hiring decision, which simply revealed that the candidate did not get the job. No justification for the hiring decision was provided. However, the decision was accompanied by the following comments from an independent third party:

I have been the interviewer’s coworker for the last 9 years. We have worked on several projects together. To be honest, I find him to be somewhat prejudiced. I think his prejudice could affect his judgment regarding which candidate to hire.

We presented all participants with the independent suggestion that the interviewer was prejudiced in order to conduct a test of whether discrimination attributions negatively affect how targets are perceived, even when those attributions are likely to be legitimate. Kaiser and Miller (2003) showed that people disregard the prejudice level of a perpetrator when evaluating discrimination claimants and disparage claimants similarly regardless of whether claimants interacted with a person known or unknown to be prejudiced. Thus, even when provided with a suggestion that the interviewer was prejudiced, we expected participants to dislike discrimination claimants.

At this point in the study, we introduced the only manipulation. Participants viewed an exit survey ostensibly completed by the candidate, and were led to believe that the candidate had indicated the factors she felt had contributed to her rejection. This questionnaire was presented as part of a larger postinterview survey given by the company, and all items were rated on a 9-point scale (1 = not at all responsible for rejection, 9 = completely responsible for rejection). The survey listed a number of factors, including her anxiety during the interview and the quality of her academic credentials. Participants viewed a completed survey in which all measures were rated as “1” except for the factor to which the candidate attributed her rejection, which received an “8.” Critically, the survey listed racism, sexism, and poor interview skills as factors. Participants were randomly assigned to one of four conditions (racism, sexism, compound discrimination, or poor interview skills) and were presented with a different version of the exit survey depending on condition. In the racism attribution condition, for example, participants viewed a survey in which racism was rated as “8,” and all other factors were rated as “1.” In the compound discrimination condition, both racism and sexism were rated an “8” and all other factors were rated as “1.” Lastly, participants evaluated the candidate, the interview setting, and the position. We were primarily interested in evaluations of the candidate; however additional ratings were included to bolster the cover story.

Measures

As in Kaiser and Miller (2003), participants rated the extent to which the target seemed hypersensitive, like a complainer, irritating, like a troublemaker, and argumentative. An average of these ratings was used to create a “troublemaker” index (M = 2.55, SD = 1.13, α = .84). Participants also rated whether the candidate seemed likable, to have a good personality, nice to have a conversation with, easy to get along with, considerate, and good to have as a friend. An average of these ratings comprised a “niceness” index (M = 4.87, SD = 0.98, α = .91). Items were scored on a 7-point scale (1 = not at all, 7 = extremely).

Results

We conducted one-way (Candidate Attribution: racism, sexism, compound discrimination, or interview skills) ANOVAs, followed by post hoc tests, to test our hypotheses. We evaluated post hoc tests according to an alpha criterion corrected for multiple comparisons (α/6 comparisons = .008). Initial analyses indicated that participant gender did not interact with candidate attribution to significantly affect the dependent measures; as a result, we do not discuss this factor further.¹

The results revealed that candidate attribution significantly affected troublemaker ratings, F(3, 108) = 5.49, p = .002, ɳ_p² = .13. We were primarily interested in the following question: Compared to a single discrimination claim, did making a compound discrimination claim make the candidate seem like more of a troublemaker? The results suggest that the answer to this question is no. The candidate was not penalized more for reporting compound discrimination (M = 2.79, SD = 1.23) compared to racism (M = 3.05, SD = 1.04), t(108) = 0.94, p = .35, r = .11, or sexism (M = 2.48, SD = 1.11), t(108) = 1.12, p = .27, r = .13. Consistent with other work from our lab (Remedios, 2015), however, the candidate was rated as more of a troublemaker when she attributed rejection to racism versus sexism, t(108) = 1.99, p = .049, r = .26. Although this comparison did not survive an alpha corrected for multiple comparisons (p = .008), we are somewhat confident in its reliability given that it replicates our previous findings.

Compared to an interview skills attribution (M = 1.96, SD = 0.84), a racism claim, t(108) = 3.92, p < .001, r = .50, and a compound discrimination claim, t(108) = 3.10, p = .003, r = .37, earned the candidate significantly higher troublemaker ratings. These findings are consistent with past research showing that discrimination claimants incur interpersonal costs (Kaiser & Miller, 2003). However, the results also showed that a sexism claim did not earn the candidate significantly higher troublemaker ratings in comparison to when she made an interview skills attribution, t(108) = 1.86, p = .07, r = .25. Candidate attribution also did not significantly affect niceness ratings, F(3, 108) = 2.05, p = .11, ɳ_p² = .05. Regardless of her attribution, the candidate was rated as similarly nice (M_Racism = 4.52, SD = 0.94; M_Sexism = 5.02, SD = 0.97; M_Compound = 4.79, SD = 1.09; M_Skills = 5.12, SD = 0.86).

Lastly, as in past research (Kaiser & Miller, 2003), average troublemaker ratings of the candidate were below the midpoint of the scale across all conditions. Overall, therefore, participants did not feel strongly that the candidate was a troublemaker. Nevertheless, the candidate was perceived most positively when she took personal responsibility for failing to get the job, even when all participants had reason to believe that the interviewer was prejudiced.

Discussion

The results of Study 1 do not support a double jeopardy perspective that the interpersonal costs of reporting racism and sexism are cumulative. The Black female job candidate was perceived as more of a troublemaker when she attributed rejection to racism and compound discrimination than when she took personal responsibility for the rejection. In addition, replicating recent work from our lab (Remedios, 2015), the job candidate was perceived as more of a troublemaker when she claimed to have experienced racism versus sexism. Perceivers may view racism attributions as more objectionable than sexism attributions because the greater perceived severity of racism may make racism seem like a less likely explanation for why the candidate was rejected (Czopp & Monteith, 2003). Thus, perceivers may view a racism claim (more so than a sexism claim) as an excuse, and this attitude may explain why participants disparage a racism claimant more than a sexism claimant (Remedios, 2015). That sexism is often tolerated in our society (Czopp & Monteith, 2003) may also help to explain why the target in the present study did not (statistically speaking) incur interpersonal costs for attributing rejection to sexism. Still, viewing a sexism claimant as less of a troublemaker than a racism claimant may represent a paternalistic dismissal of the sexism claimant’s report rather than true support of the claim (Remedios, 2015).

Returning to the main focus of this work: The results of Study 1 showed that women of color do not incur greater interpersonal costs when they attribute outcomes to multiple, versus single, forms of discrimination. Instead, race-based and compound discrimination attributions were viewed by participants as equally undesirable behaviors. In Study 2, we went beyond interpersonal judgments to examine the perceived credibility of compound discrimination claims. Discrimination claims that are perceived as lacking credibility may be viewed as inappropriate responses to benign situations. In addition, in Study 2, participants evaluated an Asian woman. We were interested in whether the interpersonal judgments observed in Study 1 extended to a woman of a different race.

Study 2: Perceptions of Credibility

Method

Participants and design

Participants were 123 undergraduates and community members (66 women) who completed the study for course credit or $10.00. Participants’ average age was 20 years old (SD = 3.47). Most participants (79) identified as White. Of the remaining participants, 22 were Asian, 12 were multiracial, 5 were Black or African American, 1 was Mexican, and 4 declined to respond. The present study involved one between-subjects factor (candidate attribution) with four conditions: racism, sexism, compound discrimination, and interview skills.

Procedure and materials

Participants completed the same procedure as in Study 1, with two exceptions. First, participants in the current study evaluated an Asian female job candidate. Second, participants completed measures examining how credible and appropriate the candidate’s attribution seemed. As in Study 1, all participants learned from an independent source that the interviewer was prejudiced. Participants then rated the extent to which the candidate seemed like a troublemaker (M = 2.94, SD = 1.03, α = .83), and seemed nice (M = 4.45, SD = 0.75, α = .88). To measure the perceived credibility of the candidate’s claim, we assessed the extent to which participants believed that the interviewer seemed prejudiced, biased, and fair (reverse-coded; M = 4.81, SD = 1.12, α = .86). Our measure of the appropriateness of the candidate’s claim consisted of the following items: “I am surprised by the candidate’s reaction to the hiring decision” (reverse-coded), “I expected the candidate to react to the hiring decision this way,” “The candidate’s reaction to the hiring decision seemed appropriate,” “The candidate’s reaction to the hiring decision seemed fitting,” and “I think most people in this situation would react the way the candidate reacted” (M = 4.18, SD = 1.09, α = .84). All ratings were on 7-point scales (1 = not at all, 7 = extremely; for appropriateness: 1 = strongly disagree, 7 = strongly agree).

Results

We conducted one-way (Candidate Attribution: racism, sexism, compound discrimination, or interview skills) ANOVAs, followed by post hoc tests, to examine our hypotheses. Once again, we evaluated post hoc tests according to an alpha criterion corrected for multiple comparisons (α/6 = .008). Initial analyses indicated that participant gender did not interact with candidate attribution to significantly affect the dependent measures; as a result, we do not discuss this factor further.

Troublemaker and Niceness Ratings

Candidate attribution significantly affected troublemaker ratings, F(3, 119) = 9.91, p < .001, ɳ_p² = .20. The candidate was rated as less of a troublemaker when she attributed rejection to her interview skills (M = 2.11, SD = 0.66) than when she attributed rejection to racism (M = 3.12, SD = 1.10), t(119) = 4.07, p < .001, r = .48, sexism (M = 3.24, SD = 0.96), t(119) = 4.50, p < .001, r = .57, and compound discrimination (M = 3.22, SD = 0.95), t(119) = 4.40, p < .001, r = .49. As in Study 1, troublemaker ratings in the compound discrimination attribution condition did not differ significantly from the racism, t(119) = 0.43, p = .67, r = .05, or sexism, t(119) = 0.08, p = .94, r = .05, attribution conditions. In contrast to Study 1, the candidate was not rated as more of a troublemaker when she attributed rejection to racism versus sexism, t(119) = 0.51, p = .61, r = .06.

Analyses also revealed that candidate attribution significantly affected niceness ratings, F(3, 119) = 2.99, p = .03, ɳ_p² = .07. However, even though the candidate was perceived as nicer when she blamed rejection on her interview skills (M = 4.79, SD = 0.82) than when she reported racism (M = 4.28, SD = 0.82), t(119) = 2.30, p = .02, r = .30, or sexism (M = 4.31, SD = 0.68), t(119) = 2.11, p = .04, r = .30, no conventionally significant post hoc comparisons survived a corrected alpha criterion (p = .008). Niceness ratings also did not differ significantly between the compound discrimination (M = 4.45, SD = 0.60) and interview skills attribution conditions, t(119) = 1.48, p = .14, r = .23. The candidate was rated as similarly nice when she reported compound discrimination versus racism, t(119) = 0.78, p = .44, r = .12, and sexism, t(119) = 0.61, p = .54, r = .11. Niceness ratings did not differ significantly between the racism and sexism attribution conditions, t(119) = 0.15, p = .88, r = .02.

Perceived Credibility

Our measure of perceived credibility examined the extent to which participants believed that the interviewer was prejudiced. In the discrimination attribution conditions, agreement that the interviewer was prejudiced signaled that participants believed the candidate’s claim. In the interview skills attribution condition, agreement revealed baseline beliefs that the interviewer was prejudiced. The results revealed that candidate attribution significantly affected perceptions of interviewer prejudice, F(3, 119) = 3.47, p = .02, ɳ_p² = .08. If the negative consequences of multiple stigmas are compounded, skepticism toward multiple discrimination claims may cause perceived credibility to be lowest in the compound discrimination attribution condition. The results did not support this prediction. In fact, perceived credibility was highest in the compound discrimination attribution condition (M = 5.20, SD = 1.03), although ratings in this condition did not differ significantly from when the candidate claimed racism (M = 4.96, SD = 1.36), t(119) = 0.91, p = .36, r = .10, or sexism (M = 4.71, SD = 0.90), t(119) = 1.84, p = .07, r = .25. Similarly, ratings of interviewer prejudice did not differ significantly when the candidate attributed rejection to racism versus sexism, t(119) = 0.96, p = .34, r = .11.

We did observe that perceived credibility was significantly higher in the compound discrimination condition than in the interview skills condition (M = 4.33, SD = 0.96), t(119) = 3.21, p = .002, r = .40. The comparison between the racism and interview skills attribution conditions did not survive the correction for multiple comparisons, t(119) = 2.38, p = .02, r = .26. Perceived credibility did not differ significantly when the candidate reported sexism compared to when she blamed rejection on her interview skills, t(119) = 1.41, p = .16, r = .20.

Thus far, the results do not support the hypothesis that compound discrimination claims are perceived as less credible than individual racism or sexism claims. In contrast, only in the compound discrimination attribution condition did credibility ratings differ significantly and reliably from ratings in the interview skills attribution condition. Put differently, participants rated only a compound discrimination claim (and neither a racism nor a sexism claim) as significantly more credible than baseline. To better understand how compound discrimination claims are evaluated, we further explored the consequences of perceived credibility for the perceived appropriateness of the candidate’s response.

Mediation: Credibility of Compound Discrimination Claims and Appropriateness of the Candidate’s Response

To further explore the finding that a compound discrimination claim was perceived as more credible than baseline, and the consequences of this result for perceptions of the candidate, we tested whether a compound discrimination claim increased participants’ beliefs that the candidate’s behavior was appropriate. We tested a mediation model in which beliefs that the interviewer was prejudiced when the candidate claimed compound discrimination (relative to the interview skills condition) explained increased agreement that the candidate’s response to the rejection was appropriate. Candidate attribution was coded using indicator coding (Hayes & Preacher, 2014). We created three dummy variables in which the reference category, the interview skills condition, was always coded as 0. Candidate attribution (coded as racism = 0, sexism = 0, and compound discrimination = 1) was entered as the independent variable. We entered alternative candidate attribution codes (racism = 1, sexism = 0, and compound discrimination = 0; racism = 0, sexism = 1, and compound discrimination = 0) as control variables. Interviewer prejudice (standardized) was entered as a mediator.

Next, we conducted the mediation analysis with 5,000 bootstrapped resamples (Preacher & Hayes, 2008). The analysis showed that candidate attribution condition significantly affected perceived credibility, b = .78, t(118) = 3.07, p = .003. As well, we regressed appropriateness on perceived credibility and observed a significant relationship, b = .27, t(118) = 2.72, p = .008. The total effect of candidate attribution on appropriateness was significant, b = .63, t(118) = 2.23, p = .03; however, the direct effect (accounting for perceived credibility), was nonsignificant b = .42, t(118) = 1.47, p = .15. The bias corrected 95% CI did not include zero [.04, .50], indicating that perceived credibility was a significant mediator of the relationship between candidate attribution and appropriateness of the candidate’s response.

Discussion

Participants in Study 2 rated the interviewer as the most prejudiced when the Asian female job candidate attributed rejection to compound discrimination. Although perceptions of credibility did not differ significantly between the compound discrimination, racism, and sexism attribution conditions, a compound discrimination attribution elicited the strongest agreement that the interviewer was prejudiced, and ratings in this condition were the only ratings that differed significantly and reliably from ratings in the control condition. In sum, the results do not show that compound discrimination claims are judged as less credible and appropriate than individual racism and sexism claims. Instead, this pattern of results leaves open the possibility that perceivers actually evaluate compound discrimination claims as believable. Compared to baseline perceptions of interviewer prejudice (in a condition in which the candidate invoked a nondiscrimination attribution to explain rejection), participants agreed more strongly that an Asian woman who claimed compound discrimination faced a credible discrimination threat. In turn, participants viewed the candidate’s compound discrimination attribution as an appropriate response to an unfair situation. Future research is needed to examine the persuasive potential of compound discrimination claims and to explore why citing multiple sources of discrimination may convince perceivers that a claimant experienced discrimination. We explore possible explanations for this finding in the General Discussion section.

We also observed that the Asian woman in Study 2, like the Black woman in Study 1, did not incur greater interpersonal costs when she attributed rejection to compound discrimination than when she attributed rejection to racism or sexism independently. Overall, the present study extends the results of Study 1 to women of a different race by showing that Asian women who report racism, sexism, and compound discrimination are regarded as troublemakers. However, a difference did emerge between the two studies: an Asian, but not a Black, woman was evaluated as more of a troublemaker when she made a sexism attribution than when she accepted personal responsibility for a rejection. We did not necessarily expect Black and Asian women to be perceived differently in these studies. Even so, one explanation for this difference across Studies 1 and 2 may relate to intersectional stereotypes. Asian, but not Black, women are stereotyped as feminine (Galinsky, Hall, & Cuddy, 2013; Ghavami & Peplau, 2012). Perceivers may be more likely to expect feminine than masculine women to report sexism and may assume that the more often women report sexism, the more likely it is that they use this type of attribution as an excuse for personal failure.

Relatedly, to the extent that intersectional invisibility can be conceptualized as a continuum, Black women may be more socially invisible than Asian women and less prototypical of women who report sexism (Purdie-Vaughns & Eibach, 2008). As a result, Black women may be targeted less frequently than Asian and White women by negative judgments of sexism claimants. Lastly, it is also possible that attributions to racism and sexism are interpreted differently as a function of the race of the claimant. If perceivers conjure different offenses when thinking about the types of sexism and racism expressions that Asian and Black women face, they may evaluate claims of such offenses differently. For example, in line with intersectional stereotypes, people may assume that Black women who report discrimination were evaluated as loud or unrefined. In contrast, people may assume that Asian women who report discrimination were evaluated as shy or submissive (Ghavami & Peplau, 2012). Such assumptions could affect judgments of the claimant. Thus, it will be interesting for researchers to explore these hypotheses in the future.

General Discussion

We examined perceptions of women of color who attributed rejection to compound discrimination. We considered two types of responses: interpersonal judgments (relevant to the candidate’s personality) and perceived credibility (relevant to the believability of the candidate’s claim). Past research examining interpersonal judgments of discrimination claimants suggests that people of color who report racism (Kaiser & Miller, 2003) and women who report sexism (Shelton & Stewart, 2004) are evaluated as troublemakers. Broadly, we investigated a hypothesis inspired by double jeopardy theory, which states that the negative consequences of belonging to multiple stigmatized groups are cumulative for women of color (Beale, 1970). Thus, one might expect women of color who attribute rejection to compound discrimination to incur greater interpersonal costs, and to be perceived as less credible, than those who attribute rejection to single forms of discrimination. In contrast, we observed that Black and Asian women were not judged to be bigger troublemakers when they reported compound discrimination than when they reported racism or sexism independently (Studies 1–2). The results suggest that the interpersonal costs of attributing outcomes to compound discrimination do not equal the sum of the costs of attributing outcomes to racism and sexism individually. Thus, we cannot draw conclusions about compound discrimination claims by studying perceptions of racism and sexism claims separately and expecting that the outcomes can easily be added together.

Second, we examined perceptions of the credibility of different types of discrimination claims. Once again, we did not observe an accumulation of negative consequences for compound discrimination claimants. The results of Study 2 revealed that participants were similarly likely to believe that the interviewer was prejudiced when the candidate reported compound discrimination, racism, and sexism. Moreover, we observed that participants were more likely to believe that a compound discrimination claimant (vs. a nondiscrimination claimant) faced a prejudiced interviewer. Indeed, a compound discrimination claim was the only discrimination attribution that elicited agreement that the candidate faced prejudice that was significantly stronger than such beliefs at baseline. Although the evidence is indirect, this finding hints that participants viewed a compound discrimination claim as credible. Additional analyses revealed that participants judged the candidate’s response as appropriate when she made a compound discrimination attribution because they believed that she was unfairly treated.

To understand this outcome, we considered that discrimination attributions can be thought of as a type of persuasive message. That is, one goal of a discrimination attribution is to convince observers that the claimant faced a credible discrimination threat. When audience members are not motivated or able to carefully scrutinize persuasive messages, their attitudes may be changed because the message is associated with positive or negative peripheral cues (Petty & Cacioppo, 1981). Petty and Cacioppo (1984) showed that argument number is a peripheral cue that persuades individuals to agree with a message because individuals implicitly link more arguments with greater argument validity. The authors asked college students to consider the benefits of introducing comprehensive exams into their university curriculum. Participants who read that the exams would be introduced in 10 years—and, thus, did not feel that the message was personally relevant—were more likely to agree with the idea when nine versus three arguments for exams were presented. Argument number predicted agreement with the message regardless of whether the arguments were high or low quality. Similarly, observers who are not personally affected by a discrimination claim may rely on peripheral cues to evaluate the validity of the discrimination attribution. Observers may be more inclined to believe that a job candidate who cites multiple ways in which an interviewer was discriminatory actually experienced prejudice. Thus, the results leave open an interesting possibility that perceivers use claim number as a peripheral cue when evaluating the strength of a discrimination-based explanation for rejection. It will be important for future research to investigate this possibility.

Overall, attributing rejection to racism and sexism simultaneously, versus attributing rejection to racism or sexism independently, does not appear to come at greater costs to the reputations of women of color. In addition, reports of compound discrimination do not appear to be dismissed as less credible than reports of individual forms of discrimination. It is important to investigate both types of responses, as the consequences of interpersonal judgments and perceptions of credibility may diverge. Troublemaker perceptions may relate to interpersonal judgments, like not wanting to work closely with targets (Shelton & Stewart, 2004). Credibility perceptions, in contrast, may relate to justice-serving outcomes, like recommended financial remedies (Blodorn et al., 2011). In sum, then, all discrimination claimants risk earning negative reputations; however, compound discrimination claimants may, at least, take solace in knowing that they do not fare even worse than racism or sexism claimants.

Implications for Intersectionality Research in Psychology

Women of color may be disadvantaged by their race and gender simultaneously; however, past research has neglected to consider experiences of compound discrimination (Bowleg, 2008; Purdie-Vaughns & Eibach, 2008; Woods-Giscombé & Lobel, 2008). Instead, race and gender are often studied as if they are separable stigmas and, consequently, many situations faced by women of color (for whom race and gender stigmas are intertwined) are overlooked (Cole, 2009; Remedios, Chasteen, & Paek, 2012; Warner, 2008). The present research contributes to a growing literature examining the intersectionality of multiple stigmas. Evidence is accumulating to show that biases faced by women of color cannot be understood by summing across the experiences of men of color and White women (Purdie-Vaughns & Eibach, 2008). For example, rather than facing negative race and gender stereotypes in all situations, Black women may escape negative gender stereotypes that are applied to White women in leadership domains. Black women who behave assertively are not subject to the same backlash endured by White women. In fact, assertive Black women are perceived as more hirable than assertive White women (Livingston et al., 2012). Likewise, our findings suggest that negative consequences of attributing rejection to compound discrimination do not equal the sum of the negative consequences of attributing rejection to racism and sexism independently. Although all discrimination claims (racism, sexism, and compound discrimination) were penalized by participants, perceivers were not the most punitive toward compound discrimination claimants.

Legal Implications

Victims of discrimination may take legal action against institutions or individuals in order to seek compensation and to enforce or improve antidiscrimination policies. Our studies examine the conditions under which discrimination claimants are perceived as making trouble and as being credible, therefore clarifying how people who file discrimination claims may be perceived by employers, jurors, judges, and the general public (Kaiser & Major, 2006). Perceptions of compound discrimination claimants are particularly interesting in light of the historical legal invisibility of multiply stigmatized people. Traditionally, antidiscrimination laws have been organized around the experiences of people with a single disadvantaged identity. The notion that race and gender discrimination claims should be filed independently implies that racism and sexism are experienced separately, an implication that is most accurate for White women and men of color (Purdie-Vaughns & Eibach, 2008). Consequently, it is difficult for women of color who experience compound discrimination to show that actions taken against them were due to either racism or sexism separately. To illustrate why, Carbado (2000) described the hypothetical case of an Asian woman who was denied promotions given to less qualified and less experienced Asian men and White women. To show that she was mistreated due to race and gender, the plaintiff would have to provide evidence of her employer’s bias against employees of color and women. However, such evidence would be difficult for the plaintiff to gather because both Asian men and women of other racial groups were promoted within this hypothetical firm.

Thus, similar cases in which women of color experience compound discrimination may not be legally viable because of restrictions on the types of claims that individuals can make in court (Purdie-Vaughns & Eibach, 2008). Our results show, however, that compound discrimination claims are not more interpersonally costly, nor are they perceived as less valid than independent racism and sexism claims. Indeed, compound discrimination claims may convince others (above and beyond baseline agreement) that discrimination occurred. Allowing women of color to cite multiple experiences of racism and sexism may make them seem more persuasive, increasing the likelihood that jurors and others in positions to judge the legitimacy of discrimination claims will support their cases.

Limitations and Future Directions

Our conclusion that compound discrimination claims and individual racism and sexism claims are, in many respects, perceived similarly, is limited given that the present studies are not well-powered. Although the number of participants per condition in the present studies is larger than cell sizes recommended by some experts (Simmons, Nelson, & Simonsohn, 2011), it will be important for future research to examine perceptions of compound discrimination claims in larger, more representative samples. Furthermore, greater racial diversity in future samples will enable researchers to examine effects of participant race on how compound discrimination claims are perceived.² An additional limitation of the current studies is our operationalization of compound discrimination. Women of color who report compound discrimination in the real world may elaborate on their experiences, contextualizing their mistreatment by providing specific examples of discrimination and/or describing mistreatment as racialized sexism or gendered racism rather than as equal parts racism and sexism. We operationalized an attribution to compound discrimination as a simple combination of attributions to individual forms of discrimination to augment internal validity; however, more externally valid approaches should be explored in the future.

In terms of future directions, we have described perceptions of credibility as potentially relating to justice-serving reactions to discrimination claims, such as recommending fair financial remedies (Blodorn et al., 2011; Kaiser & Major, 2006). However, given that we did not directly study financial remedies, additional research is needed to test the hypothesis that people recommend more compensation to claimants whom they believe faced credible discrimination threats. Likewise, we focused only on reports of different types of discrimination; however, perhaps stigmatized individuals who cite multiple examples of racism or sexism are perceived as more persuasive than individuals who cite only one example of racist or sexist treatment. In other words, people may be more likely to believe individuals who cite multiple examples of discrimination, in general, including (but not limited to) multiply stigmatized targets who report different forms of prejudice simultaneously. Future research can investigate this possibility.

Conclusion

The present work revealed that Black and Asian women do not incur greater interpersonal costs when they report compound discrimination versus when they report racism or sexism independently. Participants in two studies evaluated compound discrimination, racism, and sexism claimants as similarly likely to be troublemakers, perhaps because discrimination claims in general lead to accusations of hypersensitivity (Kaiser & Miller, 2003). Importantly, in comparison to a control condition in which a woman of color made a nondiscrimination attribution to explain rejection, a woman of color was perceived as credible, and her response was perceived as appropriate, when she reported compound discrimination. Broadly, the present results show that reactions to compound discrimination claims are not equal to the sum of reactions to separate racism and sexism claims. It is therefore important to consider the intersectionality of race and gender stigmas in order to better capture how women of color are perceived by others.

Footnotes

Acknowledgements

The authors thank Andrew Page, Nii-Ofei Dodoo, and the members of the Social Identity and Stigma Lab at Tufts University for their assistance with data collection.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Beale

F. M.

(1970). Double jeopardy: To be Black and female. Detroit, MI: Radical Education Project.

Beale

(1979). Double jeopardy: To be Black and female. In Cade

(Ed.), The Black woman (pp. 90–100). New York, NY: New American Library.

Berdahl

J. L.

Moore

(2006). Workplace harassment: Double jeopardy for minority women. The Journal of Applied Psychology, 91, 426–436. doi:10.1037/0021–9010.91.2.426

Blodorn

O’Brien

L. T.

Kordys

(2011). Responding to sex-based discrimination: Gender differences in perceived discrimination and implications for legal decision making. Group Processes & Intergroup Relations, 15, 409–424. doi:10.1177/1368430211427172

Bowleg

(2008). When Black + Lesbian + Woman ≠ Black Lesbian Woman: The methodological challenges of qualitative and quantitative intersectionality research. Sex Roles, 59, 312–325. doi:10.1007/s11199–008–9400-z

Carbado

D. W.

(2000). Race and sex in antidiscrimination law. In Levy

L. W.

Karst

K. L.

Mahoney

D. J.

West

J. G.

(Eds.), Encyclopedia of the American Constitution (2nd ed., pp. 2089–2093). New York, NY: MacMillan.

Cole

E. R.

(2009). Intersectionality and research in psychology. The American Psychologist, 64, 170–180. doi:10.1037/a0014564

Czopp

A. M.

Monteith

M. J.

(2003). Confronting prejudice (literally): Reactions to confrontations of racial and gender bias. Personality & Social Psychology Bulletin, 29, 532–544. doi:10.1177/0146167202250923

Degraffenreid v. General Motors Assembly Division. (1976). 558 F.2d 480, 484.

10.

Galinsky

A. D.

Hall

E. V.

Cuddy

A. J. C.

(2013). Gendered races: Implications for interracial marriage, leadership selection, and athletic participation. Psychological Science, 24, 498–506. doi:10.1177/0956797612457783

11.

Ghavami

Peplau

L. A.

(2012). An intersectional analysis of gender and ethnic stereotypes: Testing three hypotheses. Psychology of Women Quarterly, 37, 113–127. doi:10.1177/0361684312464203

12.

Hayes

A. F.

Preacher

K. J.

(2014). Statistical mediation analysis with a multicategorical independent variable. British Journal of Mathematical and Statistical Psychology, 67, 451–470. doi:10.1111/bmsp.12028

13.

Kaiser

C. R.

Drury

B. J.

Spalding

K. E.

Cheryan

O’Brien

L. T.

(2009). The ironic consequences of Obama’s election: Decreased support for social justice. Journal of Experimental Social Psychology, 45, 556–559. doi:10.1016/j.jesp.2009.01.006

14.

Kaiser

C. R.

Dyrenforth

P. S.

Hagiwara

(2006). Why are attributions to discrimination interpersonally costly? A test of system- and group-justifying motivations. Personality & Social Psychology Bulletin, 32, 1523–536. doi:10.1177/0146167206291475

15.

Kaiser

C. R.

Major

(2006). A social psychological perspective on perceiving and reporting discrimination. Law and Social Inquiry, 31, 801–830. Retrieved from http://www.jstor.org/stable/4490537

16.

Kaiser

C. R.

Major

Jurcevic

Dover

T. L.

Brady

L. M.

Shapiro

J. R.

(2013). Presumed fair: Ironic effects of organizational diversity structures. Journal of Personality and Social Psychology, 104, 504–519. doi:10.1037/a0030838

17.

Kaiser

C. R.

Miller

C. T.

(2003). Derogating the victim: The interpersonal consequences of blaming events on discrimination. Group Processes & Intergroup Relations, 6, 227–237. doi:10.1177/13684302030063001

18.

Kunda

Miller

D. T.

Claire

(1990). Combining social concepts: The role of causal reasoning. Cognitive Science, 14, 551–577. doi:10.1207/s15516709cog1404_3

19.

Livingston

R. W.

Rosette

A. S.

Washington

E. F.

(2012). Can an agentic Black woman get ahead? The impact of race and interpersonal dominance on perceptions of female leaders. Psychological Science, 23, 354–358. doi:10.1177/0956797611428079

20.

Petty

R. E.

Cacioppo

J. T.

(1981). Attitudes and persuasion: Classic and contemporary approaches. Dubuque, IA: William C. Brown.

21.

Petty

R. E.

Cacioppo

J. T.

(1984). The effects of involvement on responses to argument quantity and quality: Central and peripheral routes to persuasion. Journal of Personality and Social Psychology, 46, 69–81. doi:10.1037//0022–3514.46.1.69

22.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. doi:10.3758/BRM.40.3.879

23.

Purdie-Vaughns

Eibach

R. P.

(2008). Intersectional invisibility: The distinctive advantages and disadvantages of multiple subordinate-group identities. Sex Roles, 59, 377–391. doi:10.1007/s11199–008–9424–4

24.

Remedios

J. D.

(2015). Discrimination claims by multiply-stigmatized targets: Reactions to women of color who claim racism versus sexism. Unpublished manuscript, Tufts University.

25.

Remedios

J. D.

Chasteen

A. L.

Paek

J. D.

(2012). Not all prejudices are experienced equally: Comparing experiences of racism and sexism in female minorities. Group Processes & Intergroup Relations, 15, 273–287. doi:10.1177/1368430211411594

26.

Sechrist

G. B.

Swim

J. K.

Stangor

(2004). When do the stigmatized make attributions to discrimination occurring to the self and others? The roles of self-presentation and need for control. Journal of Personality and Social Psychology, 87, 111–122. doi:10.1037/0022–3514.87.1.111

27.

Sesko

A. K.

Biernat

(2010). Prototypes of race and gender: The invisibility of Black women. Journal of Experimental Social Psychology, 46, 356–360. doi:10.1016/j.jesp.2009.10.016

28.

Shelton

J. N.

Richeson

J. A.

Salvatore

Hill

D. M.

(2006). Silence is not golden: The intrapersonal consequences of not confronting prejudice. In Levin

van Laar

(Eds.), Stigma and group inequality: Social psychological perspectives (pp. 65–82). Mahwah, NJ: Lawrence Erlbaum Associates.

29.

Shelton

J. N.

Stewart

R. E.

(2004). Confronting perpetrators of prejudice: The inhibitory effects of social costs. Psychology of Women Quarterly, 28, 215–223. doi:10.1111/j.1471–6402.2004.00138.x

30.

Simmons

J. P.

Nelson

L. D.

Simonsohn

(2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22, 1359–1366. doi:10.1177/0956797611417632

31.

Stangor

Swim

J. K.

van Allen

K. L.

Sechrist

G. B.

(2002). Reporting discrimination in public and private contexts. Journal of Personality and Social Psychology, 82, 69–74. doi:10.1037//0022–3514.82.1.69

32.

Warner

L. R.

(2008). A best practices guide to intersectional approaches in psychological research. Sex Roles, 59, 454–463. doi:10.1007/s11199–008–9504–5

33.

Woods-Giscombé

C. L.

Lobel

(2008). Race and gender matter: A multidimensional approach to conceptualizing and measuring stress in African American women. Cultural Diversity and Ethnic Minority Psychology, 14, 173–182. doi:10.1037/1099–9809.14.3.173