Why do group-based inequalities exist? Cesario argues that the standard research paradigm in experimental studies of bias overstates the role of stereotypes (categorical bias) in decision-makers' perceptions and behavior. The harm in this, he suggests, is not merely the transmission of a “skewed … understanding of the human mind” (sect. 1, para. 3) to the wider culture, but the promotion of ineffective interventions – eliminating decision-maker bias – for addressing group disparities. Cesario's proposed alternative approach is: Detailed analysis of relevant decisions to ensure experimental tasks are valid representations of real-world processes; studying relevant “behavioral, personality, or other individual differences” (sect. 8, para. 3) between groups; and contrasting the effect size of categorical bias with other contributors to group disparities, particularly behavioral and personality group differences. We certainly agree that interventions aimed at eliminating decision-maker bias (e.g., blinding resumes) will not result in equal outcomes between groups. However, using gender disparities in labor market outcomes as an example, we disagree that Cesario's proposed approach will bring us closer to the goal of understanding or addressing group disparities.

Decades of scholarship have built an understanding of a gender system that (together with other interlocking hierarchical systems such as race and class) sustains inequalities of resources and authority via multiple, cumulative processes at the individual, interpersonal, institutional, and macro levels (Ridgeway & Correll, Reference Ridgeway and Correll2004). Within a gender system framework, then, the psychological processes identified by social psychologists simulate a single snapshot in time of just one of myriad interacting and dynamic mechanisms maintaining group disparities in status and resources. For this reason, we assume that there is broad consensus regarding Cesario's claim that simply eliminating decision-maker bias in any one particular context will not end group disparities in outcomes. Theorists of group inequalities have long recognized that formal equality on its own is inadequate to remedy the disadvantages of competing in a market in which the dominant group has already set the norms, practices, and standards (Young, Reference Young and Young1990). Contra Cesario, we, therefore, doubt that many social psychologists believe that decision-maker bias alone can largely explain gender disparities.

Indeed, it's for this reason that research on the effects of stereotypes goes far beyond the “standard experimental approach” described by Cesario: from developmental psychology exploring the relations between toy exposure at home and gender stereotypical play (Boe & Woods, Reference Boe and Woods2018); to psychobiological investigations of children's responsiveness to contrived gender cues and labels (Hines et al., Reference Hines, Pasterski, Spencer, Neufeld, Patalay, Hindmarsh and Acerini2016); to social psychological investigations of the effects of gendered stereotypes on career interest in science, technology, engineering, and mathematics (STEM) (Cheryan, Drury, & Vichayapai, Reference Cheryan, Drury and Vichayapai2013; Cheryan, Plaut, Davies, & Steele, Reference Cheryan, Plaut, Davies and Steele2009); to macro-level cross-national analysis showing that stronger gender stereotypes about mathematics among adolescents can explain cross-cultural variation in the gender gap in interest in a STEM career (Breda, Jouini, Napp, & Thebault, Reference Breda, Jouini, Napp and Thebault2020).

Cesario acknowledges that there are many distal causes of group differences, potentially including social and structural ones, but regards these as, “irrelevant because these causes are separable from the question of whether group disparities are because of biased decision-making for specific outcomes. For example, the reasons why men and women differ in their interest in things versus people is a separate question from whether faculty search committees are biased against women in hiring for STEM positions” (sect. 1, para. 6). However, because of the complex ways in which these distal processes interact with and shape group differences (e.g., Stephens, Markus, & Fryberg, Reference Stephens, Markus and Fryberg2012), many social psychologists do not regard them as “irrelevant.” Instead, they understand them as integral and interrelated parts of the system they are helping to unpack in their study of the effects of stereotypes.

Taking this broader view makes clear why Cesario's implicit framework, in which distal (and subsequent) effects are considered irrelevant, and whatever can't be explained by categorical bias is attributed to group differences in behavior, inadvertently encourages a false binary between “bias” and “group differences” as explanations of disparities. We agree that Cesario's suggestions will make for more accurate assessments of the contribution of decision-maker bias in a single decision-making context – and, in doing so, reduce the risks of allocating disproportionate resources to interventions likely to have modest or minimal effects. However, the research questions motivated by this implicit framework will not contribute much to understanding the complex dynamics that give rise to group disparities, and risk inappropriately rationalizing them.

Indeed, the latter point is illustrated by Cesario's stance that statistical discrimination (i.e., basing judgment of a member of a group in part on your “priors” about that group) is “a core tenet of good prediction” (sect. 5, para. 7). Thus, he is critical of the fact that “in studies of STEM hiring, the single relevant piece of information is the qualification of the applicant as revealed by the resume; being influenced by anything other than this information is treated as biased, erroneous decision-making” (sect. 5, para. 6). We are not exactly sure what other information (priors) Cesario thinks should influence recruiters. Is it the cross-culturally and ethnically variable sex difference in mathematical ability at the right-hand tail (Hyde, Lindberg, Linn, Ellis, & Williams, Reference Hyde, Lindberg, Linn, Ellis and Williams2008; Penner, Reference Penner2008)? Is it the gender gap in enjoyment of science among school-children that elsewhere is reversed (Stoet & Geary, Reference Stoet and Geary2018)? Is it men's lower contribution to unpaid labor (Hess, Ahmed, & Hayes, Reference Hess, Ahmed and Hayes2020) that provides them with more time to devote to advancement in a career in which long hours and the “zero-drag worker” are the norm (Williams, Reference Williams2001; Williams & Smith, Reference Williams and Smith2015)? Is it simply the fact that white men are the best represented demographic among scientists and engineers (National Science Board, 2018)? The use of any such priors in an employment context doesn't just risk a discrimination lawsuit. It also serves to rationalize the status quo, and to maintain and reproduce the reality of those priors.

A more detailed understanding of decision-maker judgments will not help our understanding of inequalities if it renders invisible other contributing factors and dynamics. What we need from social psychology is research investigating how psychological processes are shaped by, and contribute to, interpersonal, institutional, and macro-level factors that sustain group-based inequalities. The good news is such research is already flourishing.