Objective: Despite documented racial disparities in all facets of the criminal justice system, recent laboratory attempts to investigate racial bias in legal settings have produced null effects or racial-bias reversals. These counterintuitive findings may be an artifact of laboratory participants' attempts to appear unprejudiced in response to social norms that proscribe expressions of racial bias against Black individuals. Furthermore, given pervasive stereotypes linking Black people with crime and heightened attention to issues of racial injustice in the legal system, laboratory participants may be especially likely to attempt to appear unprejudiced in studies examining judgments of Black individuals in legal as opposed to nonlegal contexts. Hypotheses: We predicted that counterintuitive race effects (null and pro-Black effects) are more likely to occur in laboratory research examining race in legal than in nonlegal contexts. Method: We conducted a quantitative review of race effects in three leading social psychology and legal psychology journals over the last four decades (Personality and Social Psychology Bulletin [PSPB]; Law and Human Behavior [LHB]; Psychology, Public Policy, and Law [PPPL]). We then conducted two experiments in which students (N = 314; Experiment 1) and Mechanical Turk workers (N = 695; Experiment 2) read descriptions of White and Black targets in either legal or nonlegal contexts and rated each target along various characteristics (e.g., dangerous, trustworthy). Results: Our analysis of the literature indicated that counterintuitive race effects were more frequent in studies examining race in legal compared with nonlegal contexts. Our experiments likewise revealed that pro-Black race effects were stronger in legal than in nonlegal contexts. Conclusions: Laboratory research on racial bias against Black people-especially in legal settings-may produce misleading conclusions about the effects of race on important real-world outcomes. Methodological innovations for studying racial bias are needed. Public Significance StatementAmple evidence suggests that Black individuals are systematically disadvantaged in the criminal justice system and that racial disparities in legal outcomes persist to the present day. We documented inconsistencies in research findings on racial bias in naturalistic and laboratory research and across social psychology and legal psychology journals. Our own experiments reveal that laboratory participants are more likely to exhibit pro-Black attitudes when evaluating targets in legal contexts than when evaluating targets in nonlegal contexts. We suggest that laboratory research on racial bias in the legal system may be in need of reform.

Objective: Recently, experimental work on racial bias in legal settings has diverged from real-world field data demonstrating racial disparities, instead often producing null or potential overcorrection effects favoring Black individuals over White individuals. We explored the role of social desirability in these counterintuitive effects and tested whether allowing participants to establish nonracist moral credentials increased their willingness to convict a Black defendant. Hypotheses: We predicted that establishing nonracist moral credentials would increase convictions of Black defendants-especially for participants likely to harbor racial bias and external motivation to control it. Method: In two experiments, we randomly assigned White mock jurors (Study 1: N = 1,018; Study 2: N = 1,253) to establish nonracist moral credentials by acquitting a Black defendant in an initial case, acquit a White defendant in the same case, or see no prior case. Next, they judged an ambiguous case against a Black (Studies 1 and 2) or White (Study 2) defendant. After choosing verdicts, they provided open-ended guesses of what the study was about. Participants completed measures of explicit prejudice, motivations to control prejudice, and political orientation. Results: Most participants who were asked to judge at least one Black defendant guessed that the study was about racial bias and convicted Black defendants less often than did those who guessed the study was about something else. White participants who established nonracist credentials were significantly more likely to convict Black defendants compared with White participants who did not establish nonracist credentials. Subsequent analyses revealed that conservatives showed this predicted credentialing pattern, whereas liberals did not. Credentialed liberals' convictions of Black defendants remained low; instead, they convicted White defendants more than did noncredentialed liberals. Conclusions: Social desirability plays a clear role in whether White people acquit Black defendants in experiments, which does not align with persistent racial bias in the legal system. Research participants' concern about looking prejudiced might undermine the validity of experiments investigating racial bias in legal settings by artificially inflating pro-Black judgments. The opportunity to credential oneself as nonracist, however, might make conservatives more comfortable making anti-Black legal judgments-whereas credentialed liberals continue to judge Black individuals more favorably than White individuals in legal settings. Public Significance StatementCriminal contexts are very racially charged today, and many people perceive the decision to convict a Black defendant to be about race and bias. The current research suggests that concerns about appearing racially biased in experimental settings lead to lower conviction rates for Black defendants. However, conservatives (but not liberals) are more willing to convict a Black defendant when given the opportunity to establish that they are not r...