Introducing the Sociopolitical Motive × Intergroup Threat Model to Understand How Monoracial Perceivers’ Sociopolitical Motives Influence Their Categorization of Multiracial People

Theoretical overview

SDT integrates individual-, group-, and societal-level factors to explain intergroup prejudice, discrimination, and inequality (Sidanius & Pratto, 1999). It observes that members of higher status groups are on average more motivated to maintain the hierarchical status quo than members of lower status groups (Lee et al., 2011; Sidanius & Pratto, 1999), though individuals within groups vary meaningfully in their preference for societies to be organized as group-based hierarchies (i.e., their social dominance orientation [SDO]; Ho, Sidanius, et al., 2015; Pratto et al., 1994). SDT is relevant to understanding multiracial categorization because multiracial people often combine groups that differ in their social status, and thus challenge the status differentiation inherent in the hierarchical status quo (which those high in SDO seek to protect and those low in SDO seek to attenuate).

Additional theoretical background pertaining to high status group perceivers

According to the logic of SDT, when perceivers from higher status groups are categorizing multiracial people who combine their own group with a lower status group, those who are higher in SDO should categorize them as outgroup members (i.e., use hypodescent), particularly when the status hierarchy is threatened and thus requires action for group boundaries to be maintained (Knowles et al., 2009; Thomsen et al., 2008). Indeed, the reemergence of the one-drop rule during the Jim Crow Era of the early 1900s, an era during which White Americans were trying to reassert their dominance, offers a vivid example of how dominant group members who prefer the hierarchical status quo use hypodescent to reinforce the extant hierarchy.

Evidence among high status group perceivers

Ho et al. (2013) applied a social dominance framework to understand why White Americans sometimes categorize Black-White multiracial people according to a rule of hypodescent. In two studies, Ho et al. found that White Americans who were higher (vs. lower) in SDO were most likely to categorize (on a continuous scale) a Black-White multiracial target as more Black than White, particularly when they perceived that the hierarchical status quo was under threat (i.e., when they were experimentally primed with the idea that Black people are rising and White people are sinking in status vs. when the hierarchy was said to be stable; see also Chen, de Paula Couto, et al., 2018, who directly replicated this finding in their study 3). ⁴ This is consistent with the prediction that individuals who prefer the hierarchical status quo will act in ways to reify extant group boundaries, particularly when the hierarchy is under threat. Of note, this exemplifies how particular sociopolitical motives interact with unique, theoretically relevant threats, to predict multiracial categorization.

Kteily et al. (2014) examined how White individuals higher (vs. lower) on SDO perceived the Whiteness of the Tsarnaev brothers—who planted bombs killing three people and injuring hundreds during the 2013 Boston Marathon—in the immediate aftermath of the bombing. At that point in time, there was a widely publicized debate over their racial group membership: On the one hand, the brothers were originally from the Caucasus region, which gave name to the term “Caucasian”; on the other, they did not share the Western European heritage of prototypical White Americans. Consistent with their argument that high SDO individuals would favor excluding low status ambiguous targets from high status groups (to help maintain status differentiation between groups), these authors found that White perceivers higher versus lower on SDO were less likely to perceive the (low status) brothers as White. In study 3 of their paper, Kteily et al. directly manipulated the status of a racially ambiguous Black-White target (by describing him as a successful entrepreneur vs. an unemployed electrician). Consistent with SDT’s predictions, they found that individuals higher on SDO were more likely to categorize the target as Black (vs. White) when he was portrayed as lower (vs. higher) in status. Like Ho et al. (2013) and Chen, de Paula Couto, et al. (2018), this demonstrates higher SDO perceivers’ sensitivity to maintaining hierarchical differentiation between high and low status groups. Notably, the Ho et al. and Kteily et al. experiments demonstrate that status hierarchies can be threatened in different ways—whereas Ho et al. operationalized hierarchy threat as a higher status group’s loss of resources (and a lower status group’s gain; i.e., as a realistic threat; Stephan et al., 1999; see also Blumer, 1958), Kteily et al. operationalized hierarchy threat as the potential incorporation of a low status target in a high status group (i.e., loss of high status via “contamination”). Notably, this latter operationalization of hierarchy threat might combine elements of both realistic threat—to the extent the incorporation of low status targets can dilute the resources of a high status group’s—as well as symbolic threat, to the extent that a low status target is perceived to have less desirable cultural features. Regardless of how a hierarchy is threatened, SDT predicts that threats to the hierarchical status quo can lead to hypodescent, particularly in the eyes of higher SDO perceivers, a prediction that has now received substantial empirical support (see Ho et al., 2017, studies 1A and 1B, for additional evidence of a positive relation between SDO and hypodescent among White perceivers).

Of note, some work finds evidence consistent with the idea that groups use hypodescent to defend the hierarchy without necessarily finding that this is moderated by levels of SDO. In particular, in three studies (studies 1, 3A, and 3B), Cooley et al. (2018) either presented White participants with a Black-White racially ambiguous target in the context of a Black group or the same target alone. These authors proposed that White perceivers would find biracial targets in the context of a Black group (vs. alone) threatening because that would evoke concerns that White people are losing their numerical majority status. Furthermore, the authors predicted that this would be especially true when perceivers are already primed to think about a “majority-minority” future, but not when they were assured that the proportional representation of White Americans was stable (i.e., White dominance was assured). In line with these expectations (and the general tenets of SDT), White perceivers who were already primed to think about a majority–minority shift, and who were presented with a Black-White target in the context of a Black group, were most likely to use hypodescent. Still, these researchers found no evidence that this pattern depended on perceivers’ levels of SDO (i.e., their motivation to maintain the extant hierarchy), as SDT would have predicted. This result may suggest that the participants in Cooley et al.’s work were specifically motivated to maintain their ingroup’s (i.e., White Americans’) status in the face of a threat to their numerical majority status, rather than being motivated to maintain the extant hierarchy per se (captured by SDO). This highlights the need to differentiate between hierarchy-relevant and group-relevant motives, a point we return to below.

Additional theoretical background pertaining to lower status group perceivers

For high status group perceivers, engaging in hypodescent is compatible both with upholding the social hierarchy and with excluding targets from the ingroup. For low status group perceivers, however, employing hypodescent when categorizing a multiracial target involves inclusion of the target into the ingroup, even as it may also have the consequence of applying a rule that has historically helped uphold racial hierarchy.

This tension helps to explain how theorizing derived from SDT supports opposing possibilities for how SDO might relate to lower status group perceivers’ categorization of multiracial people who combine their own group with a higher status group. On the one hand, lower SDO members of lower status groups might engage in hypodescent (i.e., categorize multiracial people as ingroup members) more than their higher SDO counterparts because for lower status group members, hypodescent involves inclusion into the ingroup, and low SDO individuals might be more inclined to be inclusionary, particularly to the extent that they perceive that multiracial people face discrimination (Craig & Richeson, 2016; for evidence that lower SDO individuals are more sensitive to discrimination, see Ho, Sidanius, et al., 2015; Kteily et al., 2017; Thomsen et al., 2010).

On the other hand, higher SDO members of lower status groups might engage in hypodescent (ingroup categorization) more than their lower SDO counterparts to the extent that hypodescent is part of a suite of conservative policies that help to uphold the hierarchy that higher SDO individuals favor. Indeed, past research has proposed and empirically supported the idea that members of lower status groups who are higher in SDO are more politically conservative, opposing policies that would bring about greater racial equality (despite the fact that such opposition would harm their own group; Ho, Sidanius, et al., 2015).

Evidence among lower status group perceivers

Research supports the idea that those lower in SDO include multiracial people in the ingroup. Ho et al. (2017) found across seven studies with 3,447 African American participants that SDO was consistently negatively related to perceiving discrimination against multiracial people, which in turn was positively related to linked fate (the belief that Black and Black-White multiracial people share a common destiny) and, therefore, a higher likelihood of categorizing Black-White multiracial people as ingroup members (Black). Thus, these data did not support the possibility that lower status group members higher in SDO will use hypodescent as a function of their general tendency toward supporting more conservative policies. It is possible that although hypodescent can be seen as a hierarchy-maintaining categorization rule, that is not the default assumption among perceivers from lower status groups—the default may instead be to see hypodescent as inclusive. An experimental design that frames hypodescent as either an inclusionary phenomenon, or one that maintains the hierarchical status quo, may help identify the conditions under which lower versus higher SDO perceivers from lower status groups use hypodescent. In particular, it may be that when hypodescent is perceived as an inclusionary phenomenon, lower (vs. higher) SDO perceivers will use it more. But when it is seen as a hierarchy-supporting phenomenon, the reverse might be true.

There is also some evidence suggesting the possibility that among members of lower status groups with intermediate status in the racial hierarchy, neither lower nor higher SDO perceivers use hypodescent. Specifically, although Chen et al. (2019) were not focused on the relation between SDO and hypodescent, they reported in three studies that Asian American perceivers’ level of SDO was unrelated to the categorization of Asian-White multiracial people as either ingroup or outgroup members (and furthermore, that Asian people tended to categorize Asian-White multiracial people as more White than Asian). Thus, Asian Americans, who constitute a lower status group that is closer in status to White than to Black Americans (Kahn et al., 2009), do not show the positive relation between SDO and hypodescent found among White Americans, a high status group in the United States. Nor do they show the negative relation between SDO and hypodescent found among Black people, a low status group in the United States. It may be that similar to Black perceivers, Asian perceivers do not see hypodescent as a hierarchy-maintaining categorization rule (at least not by default). At the same time, one reason why lower SDO Asian people might not use hypodescent is because Asian-White multiracial people are relatively well accepted by White people (at least relative to Black-White multiracial people; Ho et al., 2011; Pew Research Center, 2015) and lower SDO Asian people might, therefore, not feel as much linked fate with Asian-White multiracial people as Black people perceiving Black-White multiracial people do. Again, an experimental manipulation that frames hypodescent as either an inclusionary phenomenon or as one that maintains the status quo may interact with SDO, such that SDO will be positively associated with hypodescent when it is seen as a hierarchy maintaining phenomenon but negatively associated when it is seen as an inclusionary phenomenon.

Additional theoretical background pertaining to third-party group perceivers

Individuals who ideologically support group-based hierarchies (those higher in SDO) endorse hierarchy-maintaining measures, regardless of the status of their own group and even regardless of whether their own group is part of the hierarchy under consideration (Ho et al., 2012; Ho, Sidanius, et al., 2015; Kteily et al., 2012). This is consistent with the idea that SDO is a generalized orientation toward intergroup hierarchy and inequality, rather than simply an index of support for ingroup dominance. Because SDO predicts hierarchy-maintaining beliefs, perceptions, and behavior in general, even in cases where one’s own group is not directly affected (and indeed, even when it may harm one’s own group), one would expect higher SDO among third-party perceivers to be related to hierarchy-maintaining multiracial categorization patterns (e.g., categorizing Black-White multiracial people as Black) even when the categorization outcome does not directly affect the perceiver’s group (i.e., when the perceiver is a third-party group member). For example, whereas a higher SDO Asian American perceiver may not use hypodescent when categorizing an Asian-White person—because that entails the inclusion of a target in the ingroup—she may use hypodescent when categorizing a Black-White person, because that does not entail being inclusive, and reifies existing racial group boundaries. Likewise, even a perceiver who resides in a different society, such as a higher SDO European perceiver, might categorize a Black-White American multiracial target as more Black than White to satisfy their preference for hierarchical social arrangements.

Theoretical overview

SJT proposes that individuals may be motivated to defend existing social arrangements because doing so “addresses—at least subjectively, if not objectively—underlying epistemic motives to reduce uncertainty and ambiguity; existential motives to assuage threat and insecurity; and relational motives to coordinate social relationships and achieve a sense of shared reality” (especially with other system justifiers; Jost, 2019, p. 275). For example, Hennes et al. (2012) found that individuals who were dispositionally higher in epistemic, existential, and relational motives held more politically conservative, system-justifying beliefs. Thus, from a system justification perspective, it follows that social perceivers may engage in hypodescent for two reasons: (a) because doing so reduces some uncertainty and ambiguity in the racial hierarchy that interracial relationships and the existence of multiracial people potentially introduce (i.e., hypodescent addresses the need for order and certainty in the racial order) and (b) because hypodescent may help to stabilize the existing system, and, therefore, lead perceivers high in existential needs for security to be more confident in the system’s ability to defend against threats. It is somewhat less clear how hypodescent might serve relational needs, though endorsing hypodescent might increase a sense of belonging among perceivers embedded in social circles opposed to interracial marriage and “racial mixing” (just as rejecting hypodescent might do the same for those embedded in progressive social circles particularly supportive of interracial relationships). To the extent each of epistemic, existential, and relational needs also leads to a general tendency to defend and bolster the status quo, they may have indirect effects on hypodescent through their effects on general system-justifying beliefs (e.g., negative stereotyping of racial minority groups; we discuss below how this can relate to hypodescent).

One difference between SDT and SJT regards the systems to which they relate. SJT predicts hypodescent only to the extent that engaging in hypodescent helps to maintain one’s own existing social system—because it is only one’s own system that can satisfy the basic needs that lead one to defend the status quo in the first place (Krosch et al., 2013). SDT, however, predicts that hypodescent should emerge as a function of a more general and abstract motivation to maintain hierarchical organization between groups, and thus predicts hypodescent in any system where hypodescent helps to maintain the extant hierarchy. For example, SDO predicts a preference for hierarchy in novel social systems that perceivers are not a part of (Ho et al., 2012; Kteily et al., 2012). ⁵ Thus, in contrast to predictions from SDT, a system justification perspective would predict that Americans would lack the inclination to categorize Black-White Swedes as more Black than White, because the structure of Sweden’s social system does not impinge directly on the nature of their own (i.e., using hypodescent in this case would fail to satisfy Americans’ epistemic, existential, or relational concerns).

The preceding theoretical overview does not need to be augmented for high status perceivers, so we next turn to evidence among these perceivers.

Evidence among high status group perceivers

In the only paper examining multiracial categorization from a system justification perspective (Krosch et al., 2013), the authors used political conservatism as a proxy for system justification, on account of the fact that conservatives have heightened epistemic, existential, and relational needs to justify the system (Jost et al., 2009). Krosch et al. found that White American participants who were politically conservative were more likely to categorize ambiguous Black-White targets as Black (see also Caruso et al., 2009, who showed that conservatives were more likely than liberals to believe that a darkened image of Barack Obama was more representative of him than a lightened image). Consistent with SJT, this was particularly true when they were making categorization judgments about targets belonging to a motivationally relevant system (the United States) rather than a system that did not affect them (Canada). In another study in the same paper, however, the effects of conservatism on racial categorization were not mediated by the need for structure, an epistemic motive that SJT predicts should have played a role. This is particularly unexpected from the standpoint that from among epistemic, existential, and relational motives, hypodescent arguably most directly serves epistemic needs for order and stability in the racial hierarchy.

More research is needed to test the effects of system justification motives on multiracial categorization. It would be useful for future studies to directly measure motives theorized to underpin system justification (i.e., epistemic, existential, and relational motives) to examine their direct effects on multiracial categorization; to date, only epistemic motives have been measured directly. It would also be useful to include a general measure of system justification (e.g., Kay & Jost, 2003), to examine its effects on hypodescent. Importantly, such research would need to assess system- rather than group-based existential and relational motives (e.g., assess whether an American participant believes that she depends on the U.S. government for her safety and security, or whether she seeks to promote greater harmony among all Americans). Measures that do not specify system-based needs could conflate system justification effects on multiracial categorization with “group justification” effects (e.g., a desire to maintain a strong ethnic ingroup to defend against external threats or a desire for shared reality or cohesion within the ethnic ingroup leading to the rejection of multiracial people).

Additional theoretical background pertaining to lower status group perceivers

Distinguishing it from most other theories, SJT stipulates that lower status group perceivers often justify the existing social system, even though doing so would seem at odds with their own group’s interest. Indeed, SJT was originally formulated to explain why members of lower status groups do not challenge existing social arrangements as much as one might expect if they were primarily interested in their own group’s interests (Jost & Banaji, 1994). Thus, SJT would predict that members of lower status groups should endorse hypodescent, categorizing multiracial people as ingroup members. Importantly, however, SJT would predict that this would obtain not because of any inclusionary tendencies rooted in perceptions of shared discrimination. Rather, from an SJT perspective, low status group members should employ hypodescent to the extent that they believe doing so helps reify extant group boundaries and maintain the system of social stratification, which is perceived to serve epistemic needs for order and existential needs for a stable system to defend against external threats.

Evidence among lower status group perceivers

Research is needed to test predictions derived from SJT concerning the use of hypodescent among lower status perceivers. Notably, based on Ho et al.’s (2017) finding that hypodescent is primarily used by Black perceivers who wish to challenge the system rather than justify it (i.e., those who are lower in SDO), it is possible that perceivers from lower status groups do not construe hypodescent as a means to bolster the status quo, and therefore would not use hypodescent.

Although research is needed to directly test how lower status perceivers who have higher (vs. lower) system justification needs categorize multiracial people in general, there is one study (Stern et al., 2016, study 3) testing the system justification perspective among Black perceivers in a political context. This study examined the perception of a racially ambiguous political candidate that Black American perceivers were led to believe they agreed (vs. disagreed) with, and under conditions where the national office the candidate was running for was said to be unstable (vs. stable). This study found that when the system was portrayed as stable, Black perceivers thought that a darker (vs. lighter) image of a candidate they agreed (vs. disagreed) with was most representative. However, when the system was portrayed as unstable—a realistic threat, which presumably activates the motive to bolster the system (Kay & Friesen, 2011)—Black perceivers chose a lighter (vs. darker) photo of the candidate as most representative. These authors interpreted this pattern to signify that when a threat to the system is salient, lower status group members (like higher status group members) will prefer strong candidates (those they agree with in this case) who are representative of the status quo, which includes preferring whiter or lighter skinned candidates; this would presumably be driven by those individuals who are higher in the epistemic and existential needs theorized to underpin system justification, though that was not assessed in this study.

Additional theoretical background pertaining to third-party group perceivers

From the perspective of SJT, individuals who are motivated to justify the status quo because of epistemic needs for order or existential needs for a stable system should use hypodescent even if they do not share a racial identity with the target, as long as they are embedded within the same social system. For example, an Asian American who wants to bolster racial group boundaries due to needs for order and certainty in such boundaries should, in theory, use hypodescent when categorizing a Latino-White American multiracial individual. If the perceiver resides in a different social system from the target, however, they should not be biased in categorizing the target according to SJT, because categorization biases in this case would not help them bolster a system that meets their personal needs.

Theoretical overview

Individuals who perceive that the world is a dangerous place are more likely to prioritize deference to ingroup authority figures and social norms, and to support the punishment of norm violators (i.e., endorse right-wing authoritarian values; Altemeyer, 1998; Duckitt et al., 2002). These values are thought to help groups coordinate effectively against perceived threats to the ingroup’s safety and security. Because individuals higher in RWA aim to uphold ingroup norms, they have been shown to express prejudice toward outgroups that are perceived to have norms that differ from the ingroup’s norms, such as immigrant and racial minority outgroups (e.g., Duckitt et al., 2002; Thomsen et al., 2008). Multiracial people may be perceived to be nonnormative for at least two reasons. First, they sometimes combine majority groups with minority groups that are themselves perceived to be nonnormative from the perspective of majority group perceivers. Second, even when their minority parent group is not particularly stigmatized, multiracial people may still be perceived as nonprototypical of the ingroup. Thus, right-wing authoritarians may exclude multiracial people out of concerns that they will insufficiently adhere to ingroup norms (i.e., represent a symbolic threat). Notably, unlike social dominance and system justification motives, which clearly operate at the system level (i.e., they relate to the maintenance of the extant social system), authoritarian motives can operate at the group or system level, depending on whether a perceiver has in mind the subordinate (i.e., racial) or superordinate (i.e., national) ingroup, a point we elaborate on below.

The preceding theoretical overview does not need to be augmented for high status perceivers, so we next turn to evidence among these perceivers.

Evidence among high status group perceivers

In addition to examining SDO’s effect on multiracial categorization, Kteily et al. (2014) also examined the unique effects of RWA among White American participants. These authors expected that RWA would predict the exclusion of racially ambiguous targets over and above any effects of SDO, because beyond being low status, racially ambiguous targets may be seen as norm violators. Consistent with this theorizing, these authors found that RWA predicted the categorization of the (norm-violating) Tsarnaevs as non-White, controlling for the independent effects of SDO. In a subsequent study, they manipulated whether an individual Black-White target was nonconformist (vs. conformist; i.e., a symbolic threat). Consistent with a role for authoritarianism, those high in RWA were especially likely to categorize the target as Black when led to believe he was nonconformist (vs. conformist).

Additional theoretical background differentiating lower (vs. higher) status group perceivers

In theory, regardless of perceiver group status, a perceiver who is motivated to defend ingroup norms may regard multiracial targets with suspicion, because which racial group norms the target subscribes to is uncertain. However, this may lead lower (vs. higher) status group perceivers to arrive at different categorization outcomes, depending on whether the perceiver has in mind national ingroup norms, or racial ingroup norms. This is because for high status group perceivers, national and racial ingroup norms should overlap (e.g., “American = White”; Devos & Banaji, 2005). Thus, regardless of whether a high status group perceiver is thinking about national norms or racial ingroup norms, they should categorize multiracial people as racial outgroup members.

In contrast, for lower status group perceivers, national and racial ingroup norms are less compatible (Devos & Banaji, 2005; Sidanius et al., 1997). Therefore, if a lower status group perceiver wants to protect their racial ingroup norms, they will categorize multiracial people as racial outgroup members; but if they want to protect national ingroup norms, they may categorize multiracial people as members of the lower status racial ingroup (again, because national norms are less compatible with the lower status racial ingroup norms).

Thus, depending on which social context and attendant social norms a higher RWA perceiver from a lower status group has in mind, they may categorize multiracial targets as racial outgroup or ingroup members. For example, an Asian American social perceiver categorizing an Asian-White multiracial target and more focused on the American context would likely perceive the target’s Asian ancestry as counternormative (relative to the national norm, which equates American with White; Devos & Banaji, 2005; Zou & Cheryan, 2017) and thus categorize the target as more Asian than White. In contrast, an Asian American perceiver more focused on Asian norms might instead categorize the same Asian-White multiracial target as an outgroup member (White) because of uncertainty over whether the target adheres to racial ingroup norms. Which category norms are most salient for a perceiver could vary as a function of chronic factors such as level of identification with the respective racial versus national categories and/or on external contextual factors (e.g., whether a perceiver is categorizing a target around the time of Chinese New Year’s or during the Fourth of July).

Evidence among lower status group perceivers

The theory of RWA as applied to the categorization of multiracial targets has only been tested among White Americans. Thus, research is needed to test the hypothesis that higher RWA can lead to categorizing a target as belonging to the high or low status category, depending on what norms are salient in the perceiver’s mind. Future research should examine whether RWA, identification with a lower status ethnic minority group, and national identification interact to predict hypodescent among lower status group perceivers.

Additional theoretical background pertaining to third-party group perceivers

The preceding discussion about RWA effects can theoretically be applied to the case of third-party group members categorizing multiracial people with whom they do not share racial heritage. For example, it is theoretically plausible that an Asian American perceiver higher in RWA categorizing a White–Latino American multiracial target might believe that the target represents a threat to American ingroup norms, insofar as he associates the United States with Whiteness (Devos & Banaji, 2005) and, therefore, sees Latino people as being more “foreign” (Zou & Cheryan, 2017). As in the case of perceivers from lower status groups, whether RWA effects hold for perceivers from third-party racial groups likely depends on the extent to which the perceiver is invested in national norms (i.e., the perceiver is a third-party from the racial group standpoint, but not a third-party from the national standpoint). Because RWA may reflect a desire for individuals to adhere to ingroup norms per se, whether it be national or racial ingroup norms (Duckitt, 1989), third-party perceivers from outside (vs. inside) the multiracial target’s social system may not be biased in their categorization as a function of RWA.

Theoretical overview

At its core, social identity theory proposes that in addition to defining the self in personal terms (e.g., individual traits and attributes), we often define ourselves in terms of our group memberships. Furthermore, to the extent someone sees a group identity as an important part of how they define themselves, they will act to make favorable comparisons between the ingroup and outgroups, to achieve a positive social identity (Ellemers & Haslam, 2012). A natural extension of social identity theory, then, is the ingroup overexclusion hypothesis. This hypothesis holds that because individuals are centrally concerned about the integrity and strength of social groups to which they belong, they are wary about admitting “undesirable outsiders” who may negatively affect the ingroup (Castano et al., 2002; see Yzerbyt et al., 2000, for a review of ingroup overexclusion hypothesis outside the specific context of multiracial categorization). Because multiracial people are not prototypical exemplars of any one racial group, and because they might sometimes combine one’s ingroup with a stigmatized outgroup, it follows that individuals who are highly identified with a group and who are, therefore, most concerned about its strength may be wary of including multiracial people in the ingroup.

In addition, some theorizing in the social identity tradition holds that ingroups are “bounded communities of mutual cooperation and trust” and as such, “in-group behavior is governed by norms and sanctions that reinforce expectations of mutual cooperation and trustworthiness” (Brewer, 2007, p. 732). Thus, another concern that monoracial people may bring to bear in their categorization is whether multiracial people—who may have allegiances with multiple racial groups—will be trustworthy or loyal to the ingroup. Concerns that multiracial people do not identify strongly with the ingroup, and may therefore be disloyal, could in turn lead to the outgroup categorization of multiracial people. Notably, unlike SDT and SJT, which posit that some individuals seek to maintain the social hierarchy irrespective of whether it benefits the ingroup, social identity theory (SIT) focuses on individuals’ concerns about their ingroup per se.

The preceding theoretical overview does not need to be augmented for high status perceivers, so we next turn to evidence among these perceivers.

Evidence among high status group perceivers

As with the other theories reviewed, most of the evidence in support of the ingroup overexclusion account has been conducted among high status group members. In one study consistent with the ingroup overexclusion hypothesis, Castano et al. (2002) found that Northern Italians (a higher status ethnic group) who identified strongly with the ingroup (i.e., those who are theoretically most invested in protecting the ingroup) were more likely than their more weakly identified counterparts to categorize racially ambiguous faces that combined Northern and Southern Italian features as Southern Italian (a lower status ethnic group).

Knowles and Peng (2005) also applied a social identity framework to White Americans’ categorization of Black-White racially ambiguous faces. They measured White participants’ racial identity with an implicit measure called the White Identity Centrality Implicit Association Test. Consistent with Castano et al. (2002), they demonstrated that White perceivers who strongly (vs. weakly) identified as White were more likely to categorize ambiguous faces as Black. Similarly, Gaither et al. (2016) found that White perceivers’ racial identity (collective self-esteem) predicted a tendency to categorize racially ambiguous Black-White faces as Black.

Notably, studies done following the social identity (ingroup overexclusion) approach have not examined moderators of social identification effects. For example, according to this perspective, one might expect that when a target combines a more (vs. less) stigmatized outgroup with the higher status ingroup, strongly identified high status group members might be especially likely to reject the target from the ingroup to protect the ingroup from the entry of a lower status person (because including someone with a stigmatized identity may represent both a realistic and symbolic threat to the ingroup). Thus, studies manipulating the degree to which a multiracial target’s outgroup parent group is stigmatized would provide further evidence that ingroup overexclusion effects stem from a desire to protect the ingroup from undesirable outsiders in particular (rather than all outsiders equally).

Additional theoretical background pertaining to lower status group perceivers

The idea that ingroups are bounded communities of cooperation and trust may predict the exclusion of multiracial people among members of lower status groups under some circumstances. In particular, if monoracial perceivers from a lower status group perceive that a multiracial target may not themselves want to identify with their lower status parent group, and might be disloyal to the ingroup, they might exclude multiracial people out of concerns that including potentially disloyal group members would undermine norms of trust and cooperation (Brewer, 2007).

That said, lower status group members may sometimes want to include rather than exclude multiracial people. This may happen if the inclusion of multiracial people is perceived to strengthen rather than weaken the ingroup, and/or if there is a specific reason to believe that multiracial people will be loyal to the monoracial ingroup. For example, in the 1990s, racial minority civil rights leaders (who presumably identified strongly with their respective racial minority groups) protested the potential addition of a multiracial category to the U.S. Census, for fear that multiracial people would stop identifying with their minority parent group and, therefore, cause those groups to lose political power (Prewitt, 2013).

Furthermore, according to optimal distinctiveness theory (Pickett et al., 2002), an extension of the social identity framework, individuals who feel a need to belong will be motivated to join groups that are inclusive or to perceive their current groups as more inclusive (than they would if the need to belong were not salient). One way to perceive a group as more inclusive is to increase its size (larger groups are perceived to be more inclusive than smaller groups; Pickett et al., 2002). This might be particularly relevant for members of groups that represent a numerical minority, compared with members of numerical majority groups, which may already be seen as inclusive (Gaither et al., 2016). Thus, the need to belong, or feel included, may also lead minority group members to include multiracial people as a means of making their group feel more inclusive.

Evidence among lower status group perceivers

Although some of these theoretical predictions about perceivers from lower status groups, grounded in social identity theory, have not been tested directly, there are some data consistent with each of these ideas. First, Chen et al. (2019) found across five studies that Asian Americans (total N = 1,589) are concerned about the identity preferences of Asian-White multiracial people (i.e., they express uncertainty about whether Asian-White multiracial people actually want to be seen as Asian), and therefore are more likely to distrust them and perceive them to be more White than Asian. This is consistent with the notion that group members will exclude ambiguous others who they perceive to undermine the integrity of the ingroup (Yzerbyt et al., 2000), and that group members prioritize trustworthiness and loyalty (Brewer, 2007). This was especially the case when Asian participants perceived that the ingroup was discriminated against, and consequently questioned the identity preferences and loyalty of Asian-White multiracial people. Theoretically, this analysis would apply to a perceiver from any group who believes that a multiracial person may negatively affect the integrity of the ingroup, but Chen et al. reasoned that the concern may be especially salient among group members perceiving targets who can plausibly “pass” as members of another, higher status, group. ⁶

Second, supporting the possibility that lower status group members may, under some circumstances, believe that including multiracial people will strengthen the ingroup, Chen et al. (2019; study 2) also found that African Americans who were more (vs. less) strongly identified with being Black were more likely to include rather than exclude Black-White multiracial people from the ingroup. The fact that individuals who are invested in the well-being of the Black ingroup (i.e., those who were most strongly identified) include multiracial people suggests that these perceivers felt that doing so would strengthen the ingroup, or at minimum, that they were not concerned it would hurt it. Third, consistent with optimal distinctiveness theory (Leonardelli et al., 2010), Gaither et al. (2016) found that Black participants who were asked to recall a time they were socially excluded (compared with those who recalled a time they were included, or with those who were not reminded about social inclusion or exclusion), became more likely to categorize a Black-White target as Black. Research demonstrating that including multiracial people leads to perceptions that the ingroup is more inclusive, and more supportive of inclusion needs, would help support this theory. It could also be informative to examine the types of exclusion that were recalled—if in many cases, participants recalled experiences with racial discrimination, these results may also be consistent with work demonstrating that perceptions of discrimination can lead Black people to feel linked fate with multiracial people and thus categorize them as Black.

Additional theoretical background pertaining to third-party group perceivers

Unlike the theories already discussed, social identity theory cannot readily be applied to the categorization of targets that do not share a racial heritage with the perceiver (or to targets who are outside of a perceiver’s social system). This is because this theory was formulated to account for why individuals privilege ingroups specifically, and the categorization of targets that do not share racial heritage does not have any direct implications for the ingroup’s well-being (even if it matters for the stability of the social system).

Theoretical overview

In many cases, a perceiver’s racial prejudice may be caused by the sociopolitical motives we have discussed in the preceding sections (Duckitt et al., 2002; Jost & Banaji, 1994; Sidanius & Pratto, 1999; Tajfel & Turner, 1986). However, much of the literature on multiracial categorization has not specified the source of a perceiver’s prejudice when examining the prejudice–categorization relation. Thus, we review work on the prejudice and categorization relation here, and, in the section “Toward a Theoretical Integration” below, discuss how prejudice might in some cases mediate the effects of SDO, system justification, RWA, and social identity, on categorization.

A long tradition in intergroup relations research has examined how racial prejudice, or the negative evaluation of a social group, can influence downstream beliefs about and behavior toward members of the group, ranging from discriminatory resource allocation to race-based opposition to political candidates (Kurdi et al., 2019; Sears & Henry, 2003). It follows, then, that perceivers who dislike one of a multiracial person’s racial groups may also reject the multiracial person.

Prejudice might lead to outgroup categorization because when perceivers evaluate one of a multiracial target’s parent groups more favorably than the other, they may more heavily weigh the target’s negatively evaluated ancestry (compared with the target’s positively evaluated ancestry) when categorizing the target. This hypothesis derives from a large body of research on “negativity bias,” showing that humans generally give greater weight to negatively evaluated entities compared with positively evaluated ones (e.g., negative events affect one’s mood more than positive events; Skowronski & Carlston, 1989; Taylor, 1991, cited in Rozin & Royzman, 2001).

In addition to exerting a main effect on categorization, a perceiver’s prejudice might interact with other perceiver beliefs and target characteristics to influence categorization. First, perceivers who hold an essentialist view of race—that is, who believe that traits that are stereotypically associated with racial groups can be biologically transmitted to children—may be particularly likely to differentially weigh a disfavored ancestry. This is because they have a conceptual basis for the differential weighing process (i.e., they think in terms of “essences” that are amenable to differential weighing). For example, if a perceiver holds an anti-Black/pro-White bias, and believes that a Black person and a White person who have a child together pass on their “Black essence” and “White essence” (i.e., essentialist reasoning about how racial traits are biologically transmitted), they may then more heavily weigh the child’s Black ancestry and categorize the child as more Black than White (Ho, Roberts, & Gelman, 2015).

Second, a perceiver’s prejudice may interact with a target’s characteristics to predict hypodescent, particularly when the target has stereotypical characteristics. A target’s possession of stereotypical traits or features can, on its own, influence highly consequential outcomes. For example, in a powerful demonstration, Eberhardt et al. (2006) showed that the stereotypicality of Black defendants—whether they had phenotypic features that were stereotypic of Black people—predicted the likelihood they were sentenced to death. Still, stereotypic traits or features of the target may not influence all perceivers equally. In particular, perceivers who are lower in prejudice may be able to control the influence of stereotypical features, avoiding their effects on the perceivers’ perceptions of and behavior toward targets (Devine, 1989; Plant & Devine, 1998). Thus, to take the example of death sentencing, a judge who is relatively low in prejudice may notice a defendant’s stereotypical features but nevertheless go out of her way to avoid being unduly influenced by these features (e.g., strictly focusing instead on the facts of the case and legal precedent). Similarly, whereas a highly prejudiced perceiver may allow stereotypic target features to guide their categorization of the target, perceivers lower in prejudice may attempt to ignore such features, and use other considerations to guide their categorization instead (e.g., consciously reflecting on the fact that target phenotype and behaviors are poor predictors of group membership).

Unlike some of the previous theories, perceiver prejudice should operate similarly across high status, lower status, and third-party perceivers. That is, whereas the group that a perceiver is most prejudiced against might differ as a function of the perceiver’s group status, the manner in which that prejudice may come to influence categorization (e.g., through differential weighing of a target’s ancestry), should be the same (e.g., more heavily weighing a target’s disfavored ancestry). That said, we note one way in which predictions may diverge as a function of perceiver group status below.

The preceding theoretical overview does not need to be augmented for high status perceivers, so we next turn to evidence among these perceivers.

Evidence among high status group perceivers

Consistent with the possibility that prejudice has a main effect on categorization, Kemmelmeier and Chavez (2014) found that explicit prejudice among White participants (here, anti-Black symbolic racism) was related to choosing a darker image of Barack Obama (who is widely known to have one Black and one White parent) as representative of him.

Consistent with the theory that prejudiced perceivers will more heavily weigh entities they perceive as negative (vs. positive)—particularly if they also engage in essentialist reasoning about race—Ho, Roberts, and Gelman (2015) demonstrated that intergroup bias may work in conjunction with biological essentialism to bias categorization. These authors found that White perceivers who believed that racial group membership is biologically determined rather than socially constructed, and who favored White over Black people, weighed a Black-White target’s “Black genes” more heavily than the target’s “White genes” in categorizing the target (we place the term “Black genes” and “White genes” in quotes to indicate that this reflects essentialist reasoning about race rather than scientific reality; Smedley & Smedley, 2005). Interestingly, and further demonstrating the interactive effects of intergroup prejudice and essentialism, White individuals who exhibited a pro-Black/anti-White bias, and who held essentialist beliefs, showed the opposite tendency, categorizing ambiguous targets as White.

Consistent with the theory that perceivers’ prejudice will interact with a target’s attributes to predict hypodescent, several studies have found that implicit prejudice predicts White perceivers’ categorization of racially ambiguous Black-White targets as more Black than White, especially when target faces appear angry rather than happy (Black people are stereotyped as aggressive; Dunham, 2011; Hugenberg & Bodenhausen, 2004; Hutchings & Haddock, 2008; Miller et al., 2010).

Notably, Dunham (2011) argues and empirically demonstrates that this finding may be due to a tendency to negatively stereotype outgroup members in general (rather than being dependent on stereotypes of Black people in particular). Dunham first replicates the finding described in the last paragraph, an effect that was originally documented by Hugenberg and Bodenhausen (2004). He then uses a new memory task, in which participants are first assigned to one of two novel groups that are differentiated by color (i.e., a “red group” or a “blue group”). Following group assignment, participants learn the faces of 10 people in their novel ingroup and 10 that are in the novel outgroup. After the learning phase, participants have to recall which faces were associated with which group, and furthermore, faces are shown with a happy or angry expression. Dunham found that during the recall phase, participants were more likely to incorrectly recall that angry (vs. happy) faces belonged to the novel outgroup rather than the novel ingroup, suggesting that humans associate outgroup members in general with anger, rather than Black people specifically. Furthermore, Dunham demonstrated that implicit novel outgroup prejudice was related to the tendency to associate angry novel group individuals with a novel outgroup. This suggests that humans may generally be predisposed to fear outgroup members (compared with ingroup members), a tendency that can lead to an outgroup categorization bias, particularly when the perceiver is prejudiced against the outgroup and an ambiguous target appears threatening. In a similar vein, Miller et al. (2010) demonstrated that the relationship between implicit prejudice and the tendency to categorize angry faces as Black was especially pronounced among White participants who “felt chronically vulnerable to interpersonal threats” and were experimentally induced to be frightened (by watching Silence of the Lambs).

Stepanova and Strube (2012) also investigated the effects of implicit anti-Black prejudice on the categorization of racially ambiguous targets among a group of majority White participants, demonstrating prejudice interacted with skin color such that when the target was darker, participants who were relatively high in prejudice were more likely than those lower in prejudice to categorize the targets as “African American.” Somewhat unexpectedly, however, when the target was lighter but still ambiguous, higher prejudice participants were actually more likely to categorize the target as Caucasian compared with lower prejudice participants.

Additional theoretical background pertaining to lower status group perceivers

In general, prejudice should influence the outgroup (vs. ingroup) categorization of multiracial people in a similar way among lower status group perceivers. That is, lower status group perceivers who harbor prejudice toward one of the target’s parent groups should similarly give greater weight to a target’s disfavored ancestry in the categorization process, especially if they also tend to use essentialist reasoning about race (Ho, Roberts, & Gelman, 2015). For example, a Black perceiver who is anti-White and uses essentialist reasoning should in theory categorize a Black-White target as White. Furthermore, a lower status group perceiver who is prejudiced against a higher status group and sees an angry multiracial target might similarly categorize the target as a member of the higher status outgroup (Dunham, 2011). However, if a lower status group perceiver holds the same negative stereotypes of the lower status ingroup as higher status group perceivers do, and is prejudiced against the ingroup, they may also associate an ambiguous target that is stereotypic of the ingroup with the ingroup (Hugenberg & Bodenhausen, 2004)

Of note, among Black perceivers in particular, the “angry = outgroup” thesis advanced by Dunham (2011; i.e., the idea that outgroup members are associated with anger because they are more threatening than ingroup members) and the “angry = Black” thesis advanced by Hugenberg and Bodenhausen (2004; i.e., the idea that particular social groups, such as Black people, are associated with anger/threat) lead to different predictions. Specifically, if a Black perceiver categorizing an angry Black-White multiracial target is pro-Black/anti-White, and the perceiver associates outgroup members with anger (Dunham, 2011), one can straightforwardly predict that the perceiver will categorize the target as White. If instead, the perceiver is pro-Black/anti-White but nonetheless associates Black people with anger (due to universal knowledge of group stereotypes; Devine, 1989; Hugenberg and Bodenhausen, 2004), the perceiver’s pro-Black/anti-White prejudice may counteract the effects of the stereotype—in this case, the target’s angry expression and its stereotypical association with Black people would predict Black categorization but the perceiver’s pro-Black/anti-White prejudice would predict White categorization, making the categorization outcome less certain. The fact that a target characteristic such as anger and a perceiver sociopolitical belief such as pro-Black prejudice can pull categorization in opposite directions further reinforces the importance of taking into account both perceiver and target characteristics for a complete understanding of multiracial categorization.

Evidence among lower status group perceivers

To date, there is no research examining whether prejudice does in fact operate in the same fashion to bias multiracial categorization among lower status group perceivers, an important question for future research to address.

Additional theoretical background pertaining to third-party group perceivers

Prejudice against one of a multiracial person’s parent groups should also influence multiracial categorization among third-party group perceivers. For example, an Asian American perceiver categorizing a Black-White multiracial person might categorize the target as Black, if she holds anti-Black prejudice. Furthermore, essentialist reasoning about race (Ho, Sidanius, et al., 2015), and/or an angry (vs. happy) target expression (Hugenberg & Bodenhausen, 2004) might moderate the prejudice–categorization relation among third-party perceivers, as it does among first-party perceivers. Of note—and unlike categorization driven by system justification motives—this reasoning should apply even for perceivers outside of the target’s social system (e.g., a European perceiver categorizing a Black-White American).

Exogenous intergroup motives and the mediating role of prejudice

Following the extant literature, this model specifies as exogenous variables four sociopolitical motives (person variables) that explain why higher status group perceivers might categorize a target with shared racial ancestry as an outgroup (vs. ingroup) member. In particular, research to date suggests that high status group members exclude multiracial people (categorize them as members of their lower status parent group) because they are

Notably, some work on multiracial categorization has focused on prejudice as a predictor of categorization, with other work emphasizing the direct effect of motives such as SDO, system justification, RWA, and social identification on categorization. A long tradition of research in intergroup relations outside of multiracial categorization has considered how each of these motives can shape prejudice (e.g., Duckitt et al., 2002; Jost & van der Toorn, 2012; Sidanius & Pratto, 1999; Tajfel & Turner, 1986). Putting these together, we argue that SDO, system justification, RWA, and racial identification may shape multiracial categorization both directly as well as indirectly, by shaping prejudice toward a target’s parent racial groups. Hence, we have proposed that perceivers’ prejudice against a target’s outgroup parent group (or lower status parent group among third-party group perceivers) should be considered as endogenous to the four motives outlined above and as a proximate mediator of these motives on categorization. Although we specify prejudice as a possible mediator, in many cases, the four motives will have direct effects as well. For example, a person dispositionally high in system justification motives might, on the one hand, directly use hypodescent to bolster the status quo (i.e., by reifying extant group boundaries). On the other hand, system justification motives may lead to prejudice against a multiracial person’s minority parent group (because the perception that the social system is fair can lead perceivers to attribute a minority group’s disadvantage to internal attributes of group members vs. environmental factors), with this prejudice in turn leading to hypodescent (i.e., an indirect effect from system justification to hypodescent, via prejudice). As another example, a highly identified high status group perceiver might directly categorize a multiracial person as an outgroup member, because categorizing ambiguous targets as outgroup members allows for greater differentiation between the ingroup and outgroups (Knowles & Peng, 2005). In addition, we know that highly identified members of high status groups are often prejudiced against outgroup members (Jardina, 2019), including a multiracial person’s minority parent group (because intergroup prejudice allows for positive differentiation), and this prejudice may mediate the effects of social identification on categorization.

Distinct intergroup threats as moderators

Our model takes into account how situationally salient intergroup threats may (differently) moderate the effects of each motive, leading to outgroup categorization (directly and perhaps also indirectly through prejudice). Thus, our framework of categorization encourages theorizing about which of several motives may be particularly active in driving categorization at any given moment, depending on which particular threats are present and perceived. Specifically, we separately consider threat to hierarchy, threat to ingroup norms, and threats to ingroup status. We note here that threats to the hierarchy and to ingroup status can take the form of either material/economic/physical threats (i.e., realistic threats) or cultural/value threats (i.e., symbolic threats; Stephan et al., 2016). For example, a group can be threatened both realistically, to the extent that it faces economic competition from another group, and symbolically, to the extent that potential new group members hold different cultural values.

The role of perceiver group status

Our model takes into account the ways in which perceiver group status moderates the relation between sociopolitical motives and categorization. First, it highlights that whereas both system- and group-level motives predict outgroup categorization among higher status group perceivers, the effects of sociopolitical motives on categorization are more varied among lower status group perceivers. Second, we note that only system-level motives are relevant for third-party perceivers. We expand on these two points next.

Tests to dissociate the role of distinct sociopolitical motives and threats are needed

In this section, we summarize and underscore an important contribution of our theoretical synthesis, and follow with a detailed empirical example. In particular, our integrative model suggests that to understand the unique contributions of each intergroup motive, and how each motive works in conjunction with distinct intergroup threats, future research should (a) simultaneously assess multiple motives to examine their incremental contribution to multiracial categorization, controlling for one another; and (b) systematically manipulate distinct intergroup threats to reveal dissociations in which threats activate which motives. In doing so, future research can show how distinct motives and distinct threats operate in conjunction to influence monoracial perceivers’ categorization of multiracial people.

More research is needed on perceivers from lower status and third-party groups

The possibility that some motives may have different effects for perceivers from lower (vs. higher) status groups, and among third-party perceivers, highlights the need to include ethnic minority research participants when testing theories such as social identity theory’s ingroup overexclusion hypothesis (i.e., the relation between racial identification and outgroup categorization, to take but one example where effects theoretically should diverge). At minimum, investigators need to be clear about whether the theoretical framework they are using should apply to all perceiver groups irrespective of status (and if not, why, in principle, perceiver group status may be an important moderator).

The inclusion of perceivers from different groups could also help to arbitrate between different theoretical accounts. For example, SJT suggests that perceivers from higher status, lower status, and third-party groups should all use hypodescent to maintain the hierarchical status quo. In contrast, an authoritarianism perspective would expect that, when perceivers are motivated to protect racial ingroup norms, (a) members of higher status groups will use hypodescent; (b) members of lower status groups will not use hypodescent, but rather will categorize multiracial people as outgroup members belonging to the higher status outgroup; and (c) third-party perceivers will not be invested in how the target is categorized one way or the other.