Silva, Fabiana, Irene Bloemraad, and Kim Voss. "Frame Backfire: The Trouble with Civil Rights Appeals in the Contemporary United States." American Sociological Review 90, no. 3 (2025): 349-386.
Abstract: Many scholars and activists consider civil rights to be a powerful, effective way to frame diverse causes, but do civil rights claims actually resonate? Building on social movements, collective memory, and public opinion scholarship, we conceptualize civil rights claims in three non-mutually-exclusive ways: as a highly resonant “master frame” grounded in core American ideals of equal rights, as an appeal to the idealized memory of the Civil Rights Movement, and as racialized messaging that is likely to provoke backlash. Using these conceptualizations, we derive expectations about the effectiveness of civil rights claims across diverse issues, beneficiaries, and audiences, which we test using two large-scale survey experiments. Respondents viewed “civil rights” very positively in the abstract and broadly agreed about the meaning in both closed and open-ended survey responses: civil rights are about ensuring equal rights and treatment, rather than addressing material needs. Yet, surprisingly, framing contemporary problems—even unequal treatment—as civil rights violations reduced support for government intervention. Indeed, we find widespread frame backfire: civil rights framing was counterproductive across issues (material deprivation, unequal treatment), beneficiaries (African Americans, Mexican Americans, White Americans, undocumented Mexican immigrants), and audiences (liberals, conservatives, Whites, African Americans, Latinos). Given the consistently negative effects across respondents, these findings cannot be adequately explained as racialized backlash. Instead, we propose that civil rights claims evoke comparisons to the historic Civil Rights Movement, making contemporary hardships appear less significant and prompting unfavorable contrasts with idealized claims-making of the past. Our findings challenge assumptions that frames resonate when they align with audiences’ values or appeal to positive collective memories; indeed, invoking idealized memories risks undermining support for contemporary causes.
Silva, Fabiana. "What predicts employer discrimination? The role of implicit and explicit racial attitudes." Social Science Research 108 (2022).
Abstract: Hiring discrimination against black jobseekers remains prevalent in the United States. Yet, despite strong evidence about the existence of employer discrimination, we know relatively little about what motivates employers' discriminatory behavior. I draw on an original two-wave study with a sample of white hiring agents to examine whether respondents' explicit (deliberate) and implicit (automatic) racial attitudes predict their evaluations of white and black job applicants. Building on dual-process models of the attitude-behavior relationship, I theorize that the hiring process at many U.S. organizations—characterized by ambiguity, time pressure and distractions, and the legitimacy of emotions as a decision-making tool—encourages decision-making based on implicit rather than explicit cognition. Consistent with this theorization, I find implicit, but not explicit, racial attitudes predict respondents' evaluations of white applicants, and of black applicants relative to white applicants. This suggests hiring agents’ behavior reflects implicit biases, rather than deliberate avoidance. The results further highlight the role of pro-white bias, not solely anti-black sentiment, in explaining discrimination: implicit attitudes were associated with bias in favor of white jobseekers, not only with discrimination against black jobseekers relative to white jobseekers. Finally, in open-ended responses, hiring agents explain their racially-motivated evaluations without invoking race, suggesting the ambiguity of the hiring process enables them to justify their behavior as colorblind. Together, these findings illustrate how employers can portray an egalitarian image while engaging in racially-motivated behavior.
Voss, Kim, Fabiana Silva, and Irene Bloemraad. "The limits of rights: claims-making on behalf of immigrants." Journal of Ethnic and Migration Studies 46, no. 4 (2020): 791-819.
Abstract: Activists do not just ‘name’ problems faced by migrants; they ‘frame’ them, constructing a particular meaning of the social world. Activists in the United States are especially likely to use rights language. Some appeal to human rights; others call on the history and resonance of civil rights. Those who contest immigrant inclusion often instead evoke ‘American values’. Are these competing frames persuasive? Drawing on a survey experiment of California voters, we examine whether these frames affect support for undocumented immigrants and U.S. citizens in need. We find that although respondents agree that food insecurity, sexual harassment, and inadequate health care violate the human rights of citizens and noncitizens equally, a human rights frame does not equalise support for government action to address the situation. Indeed, overall, respondents are much less supportive of government action for undocumented immigrants than citizens; neither rights nor value frames mitigate this inequality. The civil rights frame, relative to the American values frame, actually decreases respondents’ support for government action, for citizens and noncitizens alike. The type of hardship also matters: in scenarios concerning sexual harassment, legal status is not a barrier to claims-making. These findings reveal some limits of rights language for mobilisation around immigration.
Silva, Fabiana. "The strength of Whites’ ties: How employers reward the referrals of Black and White jobseekers." Social Forces 97, no. 2 (2018): 741-768.
Abstract: Sociologists commonly point to jobseekers’ racially segregated networks and employers’ discriminatory behavior to explain racial inequality in employment. Network scholars argue that, given segregated networks and black and white employees’ unequal position in the labor market, employers’ reliance on employee referrals reproduces black disadvantage. Scholars of discrimination focus instead on employers’ unequal treatment of equally qualified black and white jobseekers. Drawing on an original experiment with a sample of white individuals with hiring responsibilities, I seek to bridge these literatures by examining whether respondents’ racial prejudice affects how they reward employee referrals of black and white applicants from black and white employees. I use a measure of implicit prejudice that is resistant to social desirability and that can capture biases among people who genuinely believe they are unbiased. Whether evaluated by low-prejudiced or high-prejudiced respondents, white applicants benefit greatly from same-race referrals. In contrast, black applicants do not benefit from same-race referrals, even when they are evaluated by low-prejudiced respondents. In fact, black applicants only benefit from having a referral when two conditions are met: the referring employee is white and they are evaluated by a relatively low-prejudiced respondent. These findings suggest that in addition to their disadvantage in access to employee referrals, black jobseekers suffer from a disadvantage in returns to these referrals.
Bloemraad, Irene, Fabiana Silva, and Kim Voss. "Rights, economics, or family? Frame resonance, political ideology, and the immigrant rights movement." Social Forces 94, no. 4 (2016): 1647-1674.
Abstract: Although social movement scholars in the United States have long ignored activism over immigration, this movement raises important theoretical and empirical questions, especially given many immigrants' lack of citizenship. Is the rights “master” frame, used extensively by other US social movements, persuasive in making claims for noncitizens? If not, which other movement frames resonate with the public? We leverage survey experiments—largely the domain of political scientists and public opinion researchers—to examine how much human/citizenship rights, economics, and family framing contests shape Californians' views about legalization and immigrants' access to public benefits. We pay particular attention to how potentially distinct “publics,” or subgroups, react, finding significant differences in frame resonance between groups distinguished by political ideology. However, alternative framings resonate with—at best—one political subgroup and, dauntingly, frames that resonate with one group sometimes alienate others. While activists and political theorists may hope that human rights appeals can expand American notions of membership, such a frame does not help the movement build support for legalization. Instead, the most expansive change in legalization attitudes occurs when framed as about family unity, but this holds only among self-reported conservatives. These findings underscore the challenges confronting the immigrant movement and the need to reevaluate the assumption that historically progressive rights language is effective for immigrant claims-making.
Sole-authored. R&R.
Abstract: How do Americans ethnoracially classify others? While external ethnoracial classifications profoundly shape inequality, we know relatively little about their determinants. Drawing on a nationally representative conjoint experiment (N=1,606 observers; N=23,953 classifications), I theorize and examine the effects of racial (physical appearance), ethnic (names and language), status, and partisanship markers on White, Black, and Hispanic classification. Crucially, I find that high rates of Hispanic classification depend on both physical appearance and ethnic markers. Thus, despite the longstanding classification of Hispanics as an ethnic group that can be of any race, strong ethnic markers alone do not lead to high Hispanic classification. Further, contrary to the view of the Hispanic category as fundamentally elusive, Americans widely agree about the category’s core, even as there is ambiguity regarding its boundaries. These findings suggest that “Hispanic” is a meaningful shared category that is best conceptualized as an ethnorace. For Black and White classification, physical appearance dominates: distinctive appearances are necessary and sufficient for high classification rates. Yet, Hispanic ethnic markers meaningfully decrease White and Black classification—even among targets with distinctive appearances—indicating some Americans define these categories as excluding people with Hispanic origins, regardless of appearance. The study findings have implications for the conceptualization, measurement, and signaling of race, and for debates about the future of race in the US.
Abstract: This study identifies a novel mechanism through which referral-based hiring may disadvantage Black, Latino, and Asian jobseekers: many White hiring agents perceive these groups’ same-race referrals as reflecting in-group favoritism rather than credibly signaling candidate quality. Thus, racial minorities’ same-race referrals may be dismissed by skeptical White hiring agents. Drawing on an original two-wave experimental survey of White hiring agents (N=1,572) with open-ended responses, I found that while fewer than 20% report that White employees prefer to refer in-group candidates rather than the “best qualified” candidates, the figures are much larger for Asian (30%), Hispanic (42%), and Black (44%) employees. These racial differences stem from an important asymmetry: factors expected to undermine the credibility of minority referrals strongly influenced referral perceptions, but factors expected to enhance their credibility did not. Referral perceptions were best explained by accounts of perceived group interests, including racial group consciousness and racial resentment, which emphasize minorities’ motivation to advance their in-group. In contrast, the distinct reputational vulnerabilities that motivate minority employees to recommend qualified candidates were not salient to White hiring agents. Notably, even many racially sympathetic White hiring agents attributed favoritism to minority referrals, framing it as a legitimate response to injustice rather than as improper bias. Yet, even when viewed sympathetically, the perception of favoritism implies that minority same-race referrals are weak quality signals. Ultimately, this study extends scholarship on referral-based inequality beyond access and mobilization to hiring agent perceptions, showing that once minorities successfully mobilize referrals, racialized perceptions can still undermine their value.
[PDF]
Abstract: To what extent do racialized social markers, such as education, partisanship, and language, influence how people are ethnoracially classified by others in the US? While some scholars argue social markers meaningfully affect external classification, others reject this view, emphasizing physical appearance’s dominance. At stake is whether individuals’ race is fixed based on appearance or fluid based on social characteristics. Empirical findings are inconsistent, and key uncertainties remain regarding whether, when, and which social markers causally influence classification. Using a preregistered nationally representative conjoint experiment (N=3,127 participants; N=46,552 classifications), I examine how social markers affect White, Black, and Hispanic classification. Results indicate that understanding social markers’ influence requires distinguishing between ethnic (name and language) and non-ethnic (status and partisanship) markers and between contexts where appearance is observable versus unobservable. Ethnic markers substantially influenced classification for all three categories regardless of appearance observability—even affecting classification for targets with racially unambiguous appearances. Non-ethnic markers, however, only mattered when appearance was unobservable, with no influence even on targets with ambiguous appearances. These findings identify a central but overlooked role for ethnic markers in external classification and only a narrow role for status and political markers. The study has implications for understanding racial categorization, fluidity, and stereotypes, and for the measurement of race and discrimination.