Cesario's analysis of a Missing Forces Flaw in experimental research implies that social psychology equates intergroup bias with group stereotypes and conceptualizes stereotypes as the sole factor underlying group disparities (e.g., regarding policing outcomes; science, technology, engineering, and mathematics (STEM) participation; and school discipline). This analysis lacks consideration of many critical elements from the intergroup literature. Ample social psychological theorizing and research suggests that discrimination and the resulting group disparities are not only related to stereotypes (i.e., representations of characteristics of social groups), but also to various forms of prejudice (e.g., Dixon, Levine, Reicher, & Durrheim, Reference Dixon, Levine, Reicher and Durrheim2012), conceptualized as evaluative, affective, or emotional responses to social groups (Fazio, Jackson, Dunton, & Williams, Reference Fazio, Jackson, Dunton and Williams1995; Gawronski & Bodenhausen, Reference Gawronski and Bodenhausen2006), ingroup favoritism (Brewer, Reference Brewer1999; Greenwald & Pettigrew, Reference Greenwald and Pettigrew2014), or dehumanization of outgroups (e.g., Haslam & Loughnan, Reference Haslam and Loughnan2014).

Cesario's oversimplified depiction of social psychology ignores the tripartite attitude framework (e.g., Eagly & Chaiken, Reference Eagly and Chaiken1993), which has been dominant in intergroup research and theorizing since the 1990s (e.g., Haddock, Zanna, & Esses, Reference Haddock, Zanna and Esses1993; Jackson et al., Reference Jackson, Hodge, Gerard, Ingram, Ervin and Sheppard1996). According to this framework, intergroup attitudes contain cognitive, affective, and behavioral components, typically conceptualized as stereotypes, prejudice, and discrimination. Studies have repeatedly documented that these components account for unique variance in group attitudes (e.g., Haddock et al., Reference Haddock, Zanna and Esses1993; Stangor, Sullivan, & Ford, Reference Stangor, Sullivan and Ford1991). Most importantly, prejudice does not merely follow from stereotypes. For example, recent experimental studies support bidirectional causal relations between prejudice and stereotypes (e.g., Kurdi, Mann, Charlesworth, & Banaji, Reference Kurdi, Mann, Charlesworth and Banaji2019a; Phills, Hahn, & Gawronski, Reference Phills, Hahn and Gawronski2020). Some theoretical accounts even presume that prejudice and stereotypes are unrelated because they arise from fundamentally distinct semantic versus affective processes (e.g., Amodio & Devine, Reference Amodio and Devine2006; Brigham, Reference Brigham1971). Consequently, prejudice and stereotypes have been conceptualized as both antecedents and consequences of discriminative behaviors and group disparities.

The target article's depiction of experimental social psychology does not capture these theoretical complexities nor does it consider prejudice as an important missing force explaining group disparities. Experimental research has provided ample evidence that prejudice relates to discriminatory intergroup behaviors. One recent meta-analysis found that racial prejudice was related to discriminatory workplace outcomes (e.g., regarding selection and performance evaluation; Jones et al., Reference Jones, Sabat, King, Ahmad, McCausland and Chen2017). Another meta-analysis even concluded that racial prejudice tends to be “twice as closely” related to discrimination than stereotypes or beliefs (Talaska, Fiske, & Chaiken, Reference Talaska, Fiske and Chaiken2008, p. 263). Furthermore, meta-analyses on experimental studies on implicit cognition and micro-level interracial interactions suggest that prejudice is related to subtle behavioral effects, although average effects vary substantially (e.g., Cameron, Brown-Iannuzzi, & Payne, Reference Cameron, Brown-Iannuzzi and Payne2012; Greenwald, Poehlman, Uhlmann, & Banaji, Reference Greenwald, Poehlman, Uhlmann and Banaji2009; Kurdi et al., Reference Kurdi, Seitchik, Axt, Carroll, Karapetyan, Kaushik and Banaji2019b; Oswald, Mitchell, Blanton, Jaccard, & Tetlock, Reference Oswald, Mitchell, Blanton, Jaccard and Tetlock2013).

Of course, lab research cannot provide all answers regarding real-world group disparities. Consequently, recent empirical approaches have begun focusing on macro-level relationships between intergroup bias and real-world group disparities, consistent with the idea that “racism is more than the sum of the prejudice held by individuals in a system” (Wessells & Dawes, Reference Wessells, Dawes, Stevens and Gielen2006, p. 271). Many of these studies are inspired by the bias of crowds model (Payne, Vuletich, & Lundberg, Reference Payne, Vuletich and Lundberg2017) and rely on a novel approach, in which individual measures of stereotypes and prejudice are aggregated at geographic levels (e.g., U.S. counties) to investigate their associations with societal outcomes. For example, Riddle and Sinclair (Reference Riddle and Sinclair2019) demonstrated that racial disparities in school disciplinary outcomes were related to regional-level prejudice: Black students were more likely disciplined in U.S. counties with higher levels of racial prejudice by White residents, and this effect was robust across a number of metrics of school discipline. Using a similar approach with massive datasets of over 100 million police traffic stops, Stelter, Essien, Sander, and Degner (Reference Stelter, Essien, Sander and Degner2022) observed that Black drivers were disproportionately stopped in U.S. counties with higher levels of racial prejudice and threat stereotypes. These relationships were stronger and more robust for measures of prejudice than for measures of threat stereotypes. Furthermore, Hehman et al. (Reference Hehman, Flake and Calanchini2018) observed that Black people were disproportionately killed by police in regions with higher levels of stereotyping and (to a lesser extent) prejudice by Whites. Importantly, these relationships were even observed when controlling for local violent crime rates. Lastly, macro-level studies have observed relationships between self-reported prejudice and racial disparities in health outcomes (e.g., regarding circulatory diseases; preterm births; Leitner, Hehman, Ayduk, & Mendoza-Denton, Reference Leitner, Hehman, Ayduk and Mendoza-Denton2016; Orchard & Price, Reference Orchard and Price2017). Such findings contradict assumptions about real-world group differences and stereotype accuracy as a major missing force explaining racial disparities in school disciplinary policy, policing, and other societal outcomes.

Together, findings from both micro- and macro-level studies suggest that prejudice is an important force explaining discriminatory behavior, potentially affecting group disparities. These findings have important implications, because they demonstrate that discrimination is not only related to how people think about stigmatized groups (i.e., stereotypes), but also to how people feel about stigmatized groups (i.e., prejudice). For this reason, we disagree with the target article's assessment that “the information that [people] … have come to learn as being probabilistically accurate in their daily lives” (sect. 5, para. 4) should be regarded as the major missing force explaining group disparities in the lab or field.

In conclusion, Cesario is correct to point out limitations to the interpretability and external validity of experimental social psychological research, and we agree with the target article's assessment that real-world phenomena necessitate multi-causal explanations. But we do not see the call to abandon experimental research about group disparities as justified. Instead, a systematic combination of experimental research and field studies should enhance the ecological validity of social psychology research (Dasgupta & Stout, Reference Dasgupta and Stout2012; Mortensen & Cialdini, Reference Mortensen and Cialdini2010) and investigate relationships between stereotype- or prejudice-related behavior and group disparities. Ideally, field observations of real-world phenomena are supplemented with additional information (e.g., by decision makers), whereas experimental research on basic mechanisms of intergroup processes might benefit from linking it more closely to behavioral contingencies observed in the real world. Such a full-cycle integration of experimental and field research would be best positioned to further our understanding of the causes of real-world group disparities and help develop effective interventions to reduce them.