She watched me whenever I touched something. I guess she felt like a [B]lack person would not have money to shop in that store.

-Black adult recalling a discriminatory encounter (Zou & Cheryan, 2017: 700)

People thought that I was uneducated and low class because of my race….

-Latino adult recalling a discriminatory encounter (Zou & Cheryan, 2017: 701).

Only two memories from two distinct individuals, but not unlike the memories recalled by other people of color (PoC) (Pérez, 2021).Footnote 1 In racially stratified polities like the United States, systemic discrimination by dominant group members (e.g., Whites) is a major tactic that keeps marginalized groups in “their place,” as highlighted by both epigraphs (Blumer, 1958; Bobo & Hutchings, 1996; Sidanius & Pratto, 1999; Sidanius et al., 1997; Kim 2003; Hutchings et al., 2011). Yet emerging work suggests that the precise locations of people of color within America’s racial hierarchy varies dramatically, a reflection of the very many real differences that distinguish PoC in terms of their arrival to the US, their treatment by US institutions, and their unique political aspirations and goals (Bobo & Hutchings, 1996; Kim, 2003; Masuoka & Junn, 2013; Zou & Cheryan, 2017). Consider as but one example the contrasting experiences of African Americans and Latinos. While the former population is largely a product of slavery and its many legacies (Carter, 2019), the latter population is a strong function of recent immigration flows and their demographic consequences (García & Sanchez, 2021). How, then, can these communities of color—and others— unify politically if their experiences in the US are so different?

Part of the answer, we contend, involves a deeper appreciation for the precise locations of PoC within America’s racial hierarchy and the types of discrimination they endure based on these positions. Accumulating work in psychology (e.g., Zou & Cheryan, 2017; Craig et al., 2022) and political science (e.g., Masuoka & Junn, 2013; Hutchings & Wong, 2014; Pérez & Kuo, 2021) suggests the unique stations of PoC within America’s racial order depend on two considerations: how foreign and how inferior is a racially marginalized group perceived to be in comparison to Whites?

Fig. 1
figure 1

Two axes of subordination

Full size image

As Fig. 1 indicates, Whites are positioned as the most superior and American racial group within our nation’s racial hierarchy, which reflects their dominant status vis-à-vis people of color. However, while Asian and Latino people are each stereotyped as foreign and un-American, Asian individuals are considered a more superior minority group than Latino and Black individuals. Equally important, although Black people are stereotyped as a relatively more American minority than Latinos and Asians (Carter, 2019), both Black and Latino individuals are stereotyped as more inferior minorities than Asians, as underscored by the model minority myth—the characterization of Asian individuals as more “well-to-do” and less impertinent than Black and Latino people (Tuan, 1998; Kim, 2003; Xu & Lee, 2013). The power of these two axes of subordination resides in their flattening of otherwise highly internally diverse populations. Black individuals are stereotyped as inferior—for example, as overwhelmingly uneducated and poor—despite significant socio-economic heterogeneity (Dawson, 1994). In turn, Latinos are stereotyped as foreign despite the fact that the majority of Latino individuals are US-born (García & Sanchez, 2021). Moreover, Asian individuals are stereotyped as un-American even though many of these individuals and their families have been in the US for generations (Tuan, 1998).

This relative positioning of racially minoritized groups does not preclude the formation of political alliances between groups (e.g., Jones-Correa 2011; Wilkinson, 2015; Benjamin, 2017; Merseth, 2017; Corral, 2020), but the nuances between these stations make it more difficult to foster interminority partnerships given the psychology of minoritized status (Pérez & Vicuña in press), the sense of relative deprivation and alienation among specific communities of color (e.g., Bobo & Hutchings 1996; Morín et al. 2011; Hutchings et al., 2011; Carey et al., 2016; Wilkinson & Bingham, 2016), and the enormous heterogeneity that exists between and within racially minoritized groups (e.g., White & Laird, 2020; Sanchez & Vargas, 2016; Wong et al., 2011). For example, the positioning of Black individuals as a more American group does not mean they are considered White (Sidanius & Pratto, 1999; Masuoka & Junn, 2013). Rather, it means that in light of Black people’s contributions to US culture and history, they are seen as a relatively better exemplar of an American when compared to other racially minoritized groups, such as Asian Americans and Latinos (Zou & Cheryan, 2017). This positioning makes Black individuals distinctive vis-à-vis other racially minoritized groups (Brewer, 1991). Distinctiveness is a strong human motive that propels individuals to preserve those attributes and statuses that make their ingroup special and unique (Jetten et al., 1999; Pickett & Brewer, 2005), which complicates the formation of interminority alliances, even when racially minoritized groups share a comparable station in society (Craig et al., 2012). The same logic applies to other racially minoritized groups, where a natural impulse exists to display ingroup favoritism, rather than a broader pro-minority outlook (e.g., Kim 2003; Kaufmann, 2004; Morin et al., 2011; McClain et al., 2011).

Some evidence for the politically catalyzing effects of one’s station in this hierarchy has accumulated across a series of experiments. These studies manipulate whether a specific community of color senses it is discriminated similarly to another minoritized outgroup, with downstream consequences for interminority unity in politics. For example, two published experiments show that exposure to information that Latinos are discriminated against as foreigners, similar to Asian individuals, causes Latinos to express greater solidarity with PoC, which then increases Latino support for public policies that implicate Asian individuals, such as more flexible policies toward highly educated and skilled immigrants (Pérez et al., 2021). In turn, Asian individuals who sense they are discriminated as foreigners, similar to Latinos, leads Asian adults to express more solidarity with PoC, which then increases their support for policies that target Latinos, including reducing the presence of the US Border Patrol agents along the US-Mexico border. In a similar vein, other research suggests that when Middle Eastern individuals sense they are discriminated as foreigners, similar to Latinos, they also express more solidarity with PoC, with downstream increases in support for more flexible policies toward Latinos (e.g., pathway to citizenship for undocumented Latino immigrants) (Eidgahy & Pérez, 2022). Critically, multiple survey analyses and other conceptually similar experiments affirm aspects of this mediated chain reaction that produces interminority solidarity (Chan & Jasso, 2021; Sirin et al., 2016;  Carey et al. 2016; Merseth, 2017; Jones-Correa 2011; Zamora, 2011), which increases confidence in this proposed framework.

Still, closer inspection of these findings reveals some major theoretic and empirical blind spots. The first one is weak engagement with the experience of Black individuals, their position in America’s hierarchy, and how it affects their politics toward other minoritized groups (Carter, 2019; Wilkerson, 2020; Benjamin, 2017). This omission matters because Black people have been part of the US since our country’s inception and America’s hierarchy originally emerged to marginalize and oppress them (alongside indigenous peoples) (Omi & Winant, 1986; Marx, 1998; Zou & Cheryan, 2017). Because of these legacies, Black individuals occupy a unique location in our nation’s racial order as an inferior group (similar to Latinos), but a more American minority (unlike Asian and Latino people) (Kim, 2003; Fouka & Tabellini, 2021). While some research suggests that activating Black individuals’ sense of being American can lead them to sometimes express less solidarity with people of color who are stereotyped as un-American and foreign (e.g., Asians, Latinos) (Pérez & Kuo, 2021), it remains unclear whether Black individuals will express greater solidarity with PoC if they sense they are similarly discriminated against like other inferior groups (e.g., Latinos, Arab Americans) (Zou & Cheryan, 2017).

Second, and related to the previous point, there is a glaring absence of evidence on the motivating impact of this inferiority dimension, which, in principle, is highly relevant to Black individuals, Latino people, and other PoC. This omission is important because without evidence of this pathway, the proposed notion of a more nuanced hierarchy (see Fig. 1) is reduced to a simpler one where only a group’s sense of foreigness matters. If true, a more parsimonious racial order would imply that Black solidarity with PoC is incredibly difficult to achieve, since shared discrimination would be restricted to these groups’ sense of being foreign and un-American. Yet as Tuan (1998: 8) explains, “Blacks may be many things” in the minds of people, “but foreign is not one of them.” Consistent with this view, experimental evidence contrasting the effects of the foreigner axis (operationalized via contributions to U.S. culture by one’s ingroup) with the inferiority axis (operationalized via reported hate crimes against one’s ingroup), finds that while a sense of shared inferiority catalyzes Black-Latino solidarity, a shared sense of being American fails to generate interminority solidarity among Black adults (Pérez et al. 2023a).

Our paper breaks new ground on these aspects of prior work by examining Black relations with Latinos and other PoC, which are often characterized by political conflict (e.g., McClain & Karnig 1990; Bobo & Hutchings, 1996; McClain et al., 2007; Hutchings et al., 2011; Morín et al., 2011; Carey et al., 2016; Wilkinson, 2015; Wilkinson & Bingham, 2016; Benjamin, 2017; Corral, 2020). To this end, we theorize how Black people’s position in America’s racial order motivates their sense of solidarity with PoC in politics. As we explain in more detail below, Black individuals are deemed a more inferior group, similar to Latinos and other minoritized groups. This leads to the possibility that highlighting the inferiority dimension in America’s hierarchy leads Black individuals to express more solidarity with people of color—with downstream consequences for their political attitudes.

To test these ideas, we ran two large-scale experiments with Black (N = 1064) and Latino (N = 996) adults. Both studies were pre-registered and randomly allocated participants to read (1) control information, or (2) information highlighting how, similar to another outgroup, one’s ingroup is treated as “second class citizens” (i.e., inferiority dimension). Participants in both studies then reported their sense of solidarity with PoC, followed by their support for policies that implicate other outgroups as well as their own ingroup.

As hypothesized, we find that Black and Latino adults who sense similar discrimination as inferior (“second-class citizenship”) expressed greater solidarity with people of color, which was then associated with increased support for initiatives that implicate both Black people (endorsing Black Lives Matter movement) and Latino individuals (pathway to citizenship). A meta-analysis further suggests this inferiority pathway is robust across both minoritized groups, thus providing evidence of its generality. In addition, a series of sensitivity analyses indicate this pathway is statistically viable, which fortifies ongoing work on the more nuanced nature of America’s racial hierarchy. We discuss these findings and their implications for theory-building and further experimentation in the realm of interminority politics in the US.

Theory and Hypotheses

The theoretic framework we develop to interpret our experimental results draws on insights from two literatures on the psychology of interminority relations. The first body of work we consult is social identity theory (SIT) and its offshoots (Turner et al., 1987; Huddy, 2001, 2013). According to this broad literature, the categorization of individuals into ingroups and outgroups generally devolves into the expression of ingroup favoritism: an attitudinal and behavioral bias toward one’s category or ingroup (Tajfel et al., 1971). A major motivation behind this effect is distinctiveness (Brewer, 1991). This motive drives people to uphold those attributes, values, and experiences that makes their ingroup special or unique, thus endowing an ingroup with cachet or positive value (Danbold & Huo, 2015; Pickett & Brewer, 2005). In light of ingroup favoritism, ingroup members coalesce behind their category, thus increasing the affective, cognitive, physical, and material space between them and an outgroup(s) (Tajfel & Turner, 1986).

For interminority politics, this research implies that unity between similarly situated groups—such as African Americans, Latinos, and other PoC—is generally difficult to achieve due to the enormous heterogeneity in historical, material, and political circumstances that characterize these groups, a broad pattern extensively documented by scholars across the last 30 years (e.g., McClain & Karnig 1990; Bobo & Hutchings, 1996; McClain et al., 2007; Hutchings et al., 2011; Morín et al., 2011; Carey et al., 2016; Wilkinson, 2015; Wilkinson & Bingham, 2016; Benjamin 2017; Craig et al., 2012; Corral, 2020). Indeed, per the distinctiveness motive highlighted above, similarly stationed outgroups are more inclined to display ingroup favoritism than intergroup unity (Pickett & Brewer, 2005). For example, Pérez (2021) reports a study showing that underlining racial discrimination against one’s ingroup (e.g., Latinos) leads individuals to express more favorable sentiments toward one’s own (i.e., pro-Latino sentiment), not broader favorability toward all groups who are similarly stationed to one’s group (i.e., pro-minority sentiment).

Given these findings, what does cause increases in solidarity between PoC? Ironically, other SIT research suggests that the very same processes just described can be redirected to achieve greater unity between different marginalized groups. Known as the Common Ingroup Identity Model (CIIM), this work suggests that focusing on categories that are broadly shared by two or more ingroups minimizes the affective and cognitive differences that ingroup favoritism produces (Gaertner et al., 1989, 1999). Here, the re-categorization of varied groups under a new, broadly shared ingroup leads individuals to extend the benefits of ingroup favoritism to members of the broader category held in common. For example, in political science, Transue (2007) shows that relative to priming racial identity, priming American identity leads White adults to express less racially conservative opinions. Similarly, Levendusky (2018) reveals that heightening people’s sense of being American reduces affective polarization between self-identified Democrats and Republicans. In both instances, the higher-order group (American) cognitively encapsulates the lower-order groups (race, party) that people belong to, thus extending the benefits of ingroup favoritism to the latter.

This dynamic is generally shaped by a similarity principle, where a shared ingroup minimizes the perceptual and affective differences between outgroups. For example, Cortland et al. (2017) report a series of experiments where highlighting a source of shared discrimination between marginalized groups with different identities (e.g., sexual and racial minorities) boosts support for policies that do not directly implicate one’s narrower ingroup (see also Craig et al., 2012, 2022; Sirin et al., 2016). When tied back to the configuration of America’s racial order in Fig. 1, it becomes clear that broad similarities can emerge in terms of how various communities of color are discriminated against. For example, research on Asian Americans, Latinos, and Middle Eastern and North Africans (MENAs) shows that when they are exposed to information highlighting similar discriminatory treatment as foreigners, they express more solidarity with PoC, which is then associated with downstream increases in support for policies that do not implicate one’s narrower ingroup (e.g., Asian American support for flexible policies toward undocumented Latino immigrants) (Eidgahy & Pérez, 2022; Pérez et al., 2021; see also Sirin et al., 2016; Craig et al., 2022).

Braiding these insights together, we derive and test two hypotheses about the position of people of color within America’s racial order and how this location can promote solidarity with PoC. Our primary focus here is African Americans, with Latinos as a comparison case. Our first hypothesis (H1) stipulates that a shared sense of discrimination as inferior should increase Black and Latino solidarity with people of color. That is, insofar as members of these two communities sense that their ingroup is similarly discriminated as another inferior outgroup, they should express greater solidarity with PoC. In turn, we hypothesize that a heightened sense of solidarity with PoC will boost support for policies that do not directly implicate one’s narrower ingroup (H2). This means that in light of solidarity with PoC, Black individuals should become more supportive of pro-Latino initiatives, including more flexible policies toward undocumented Latino immigrants, such as reductions in border patrol surveillance along the US-Mexico border and a pathway to citizenship for undocumented Latino immigrants. Similarly, Latinos should become more supportive of pro-Black efforts, such as endorsement of #BlackLivesMatter and policies to combat hate crimes against Black individuals. Table 1 lists our hypotheses for reference.

Table 1. Hypotheses about solidarity with people of color (PoC)
Full size table

Research Design: Parallel Experiments with Black and Latino Adults

We test our hypotheses using two pre-registered experiments with Black and Latino participants (https://aspredicted.org/see_one.php; see SI.1). In partnership with Dynata, an online survey platform, we recruited large samples of Black (N = 1064) and Latino (N = 996) adults to complete a brief 8-min survey. Although not yielded probabilistically, these opt-in samples are nonetheless highly heterogeneous (see SI.2) and an improvement relative to convenience samples of primarily student populations that are sometimes used in experimental work with communities of color (Pérez, 2021).

Both experiments employ a classic mediation design, where a treatment’s indirect effect on an outcome(s) is transmitted through a mediating variable (i.e., solidarity between PoC) that is observed rather than manipulated (Hayes, 2021). This design cannot definitively establish causality in the downstream path between our mediator and outcome(s). But our main goal is to appraise the viability of our theory, not the establishment of solidarity’s causal effects, which require additional studies with alternate designs (a point we reprise in the conclusion).

After consenting, participants answered a few demographic questions (i.e., age, gender, ideology) and an attention check (see SI.2 for balance checks). Participants were then randomly allocated to one of two conditions. The control group exposed participants to an article depicting the gradual extinction of giant tortoises (cf. Hopkins et al., 2020). In turn, participants in our inferiority condition were treated with an article describing continued discrimination of Black (or Latino) people as “second class citizens.”Footnote 2

Each experiment presented our treatments as news briefs focused on another racialized outgroup. This means that Black participants read articles largely focused on Latino people, while Latino participants read an article primarily focused on Black people. All manipulations and question wordings can be inspected in (SI.3). For Black participants, the inferiority manipulation consisted of a news brief titled “Despite Their Presence in the United States for Decades, Many Latinos are Still Treated as Second Class Citizens, As Evidenced by Hate Crimes Data.” The article highlights the continued prejudice and discrimination that many Latino individuals experience, with hate crimes data corroborating this trend. Critically, a shared sense of inferiority is introduced in passing by briefly noting that Latinos “are still viewed as second-class individuals, similar to many Black people.” In this way, we manipulate a sense of similarity in discrimination experiences between groups, which reflects our proposed mechanism and aligns with prior work showing how intergroup similarity is sufficient to trigger a sense of commonality between diverse individuals (e.g., Cortland et al., 2017). In this specific treatment, which was assigned to Black participants, 5 out of 246 words refer to “Black people”, which is 2% of the total. An effect from this treatment will suggest that a link between one’s ingroup and another minoritized outgroup must be made in order for solidarity to reliably increase.

To ensure that individuals were, in fact, treated, all participants completed a manipulation check asking them to indicate whether the information they read reflected the underlying thrust of each treatment.Footnote 3 Across both samples (N = 2060), a total of 250 participants failed this manipulation check, which is about 12% of our combined samples. Per our pre-registration, we exclude these participants from our analyses.Footnote 4 Table SI.6 shows our results are substantively identical if we include in our analyses those respondents who did not pass this manipulation check.

Following our manipulation checks, participants completed two (2) statements designed to capture our mediator, solidarity with people of color, with one item adapted from Leach et al. (2008) and another item being reverse-worded in order to mitigate possible acquiescence bias. The positively worded item (Leach et al., 2008) read “I feel solidarity with people of color, which include African Americans, Asian Americans, and Latinos.” This item reflects the degree to which an individual is committed to the ingroup in a specific context. Given that people of color is a superordinate identity (Pérez, 2021), this item references the major groups nested under this larger-order category. Although it is plausible that a reference to one’s specific ingroup captures solidarity with one’s racial community, rather than solidarity with other minoritized outgroups, prior work shows that a similar item gauging solidarity between PoC without reference to specific ingroups is still significantly associated with support for pro-outgroup policies (e.g., Black solidarity on this simpler item increases support for pro-Latino policies) (Pérez, 2021).

In turn, our reverse-worded statement on solidarity read “The problems of Blacks, Latinos, Asians, and other minorities are too different for them to be allies or partners.” This item highlights the heterogeneity of group interests among people of color as a way to provide a more stringent assessment of solidarity between people of color, since narrower ingroup interests can, in principle, compromise a sense of solidarity (Wilkinson, 2015; Benjamin, 2017). Participants answered both of these items on a scale from 1-strongly disagree to 5-strongly agree. We code and scale them so that higher values reflect stronger solidarity with PoC. Our goal was to scale these items to reduce the lower reliability that comes with using single items to measure a mediating variable (Hayes, 2021). This item pair correlates positively, but modestly, due to the reverse-worded item in this set.Footnote 5 However, SI.4 indicates that both items provide useful information on the underlying variable of interest (i.e., solidarity). Moreover, if we analyze each solidarity item separately, we find that our treatment impacts each solidarity item in a positive direction, but with mixed significance levels, which is consistent with the presence of random measurement error. Our analyses therefore follow our pre-registration by scaling these items as our measure of solidarity between people of color.

After our mediator, we tapped support for policies that implicate Latinos and Blacks, respectively (counter-balanced). Using the same 5-point scale, all participants indicated agreement with “Increasing the number of border patrol agents at the US-Mexico border,” “Renewing temporary relief from deportation for undocumented Latino immigrants brought to the US as children,” and “Granting a pathway to citizenship for undocumented Latino immigrants.” We code replies so that higher values reflect a pro-Latino response.Footnote 6

Table 2 Shared discrimination as Inferior produces more political unity between black and latino adults
Full size table

Using the same response scale as before, participants also indicated their agreement with “Limiting the protest activities of #BlackLivesMatter and other movements like it” and “Introducing harsh penalties for hate crimes committed against Black individuals.” Again, all items are coded so that higher values indicate greater support for the pro-Black position.Footnote 7

As an additional outcome in both studies, we combine our dependent measures into additive indexes where higher values signify stronger pro-Black and pro-Latino opinions. This has the added virtue of attenuating measurement error in our estimation (Brown, 2007), while giving us a sense of how broadly systematic any observed effects are (Broockman et al., 2017). In our analyses, all variables run along a 0–1 interval, allowing us to interpret our OLS coefficients as percentage-point shifts. In addition, we include liberal ideology as a covariate in the downstream path between our mediator and outcomes, to ensure that any associations between solidarity and our dependent variables are independent of this key source of public opinion among people of color (Pérez, 2021). All exact p-values reported in the text are two-tailed.

Results: Black and Latino Adults as Inferior Groups

We begin by assessing whether a sense of shared discrimination as inferior causes increases in solidarity with PoC among African American and Latino adults. Table 2 reports the relevant results for each group. Relative to the control, reading about Latinos’ inferiority causes Black adults to express marginally more solidarity with PoC (0.031, s.e.=0.016, p < .060), a shift of 3% points in our proposed mediator. This evidence is consistent with (H1). In further support of this hypothesis, we uncover a similar effect among Latinos, who in light of the inferiority treatment, reported about 3% points more solidarity with PoC as well (0.029, s.e.=0.015, p < .062).

With these effects bordering on statistical significance, we deviate from our pre-registration and re-assess them via a mini meta-analysis (Goh et al., 2016; Hopkins et al., 2020). We call attention to this deviation so that readers can judge how much weight to give this post-hoc analysis. Our meta-analysis allows us to decrease (random) measurement error, which can attenuate treatment effects, while also boosting statistical power. It further enables us to appraise whether the direction of our treatment generalizes, reliably, across both samples. We find that across both minoritized groups, our inferiority treatment reliably generates more solidarity with PoC among Black and Latino participants (d = 0.130, s.e. = 0.050, p < .010). Since d values are standardized mean differences, our estimate here suggests a significant treatment effect of a little more than one-tenth of a standard deviation, comparable to prominent published work using experiments (e.g., Kalla & Broockman, 2020, 2022).

Given this measureable increase in solidarity with PoC, what occurs next? Table 2’s remaining entries capture the downstream associations between solidarity and our suite of outcomes among African Americans. For parsimony’s sake, we mainly focus on this treatment’s effect on support for policies that implicate another minoritized outgroup (e.g., Black support for pro-Latino policies). However, we note that the downstream associations between solidarity with PoC and support for policies that implicate one’s specific ingroup are, as one might expect, relatively strong: a reflection of ingroup favoritism (Tajfel et al., 1971; for a review, see Pérez and Vicuña in press). The fact that the same directional pattern emerges when the outcomes implicate another outgroup is consistent with our claim that—in light of our treatment—a sense of solidarity broadly encapsulates one’s ingroup and other similarly stationed outgroups (i.e., Black and Latino adults). In this vein, we find that a unit shift in solidarity with PoC is associated with greater Black support for policies that implicate Latinos, such as opposition to border patrol increases along the US-Mexico Border (0.056, s.e.=0.047, p < ns), support for Deferred Action on Childhood Arrivals (DACA) (0.233, s.e.=0.046, p < .001), and support for a pathway to citizenship for undocumented immigrants (0.268, s.e.=0.046, p < .001). Indeed, combining these variables into a summated index reveals that, on average, the association between heightened solidarity and support for pro-Latino policies (scaled) is substantively and statistically meaningful (0.186, s.e.=0.031, p < .001). Figure 2 (panel A) depicts this mediated relationship among Black adults.

Among Latinos, a similar pattern occurs where heightened solidarity with PoC is significantly associated with support for policies that strongly implicate Black people. More specifically, a unit increase in solidarity with PoC is associated with greater Latino endorsement of efforts to combat hate crimes against Black individuals (0.436, s.e.=0.051, p < .001) and stronger Latino support for the Black Lives Matter Movement (0.385, s.e.=0.049, p < .001). In fact, when we combine this item pair into a simple additive index, we find that greater solidarity with people of color is associated with greater Latino backing of pro-Black policies (0.411, s.e.=0.034, p < .001). Figure 2 (panel B) depicts this mediated relationship among Latinos.

Fig. 2
figure 2

Solidarity with PoC mediates the impact of sensed Inferiority on black and latino adults

Full size image

Building on these findings for Black and Latino adults, respectively, we undertake two additional analyses. The first of these examines whether the mediated chain reaction depicted in Fig. 2 is reliably different from zero, a key consideration in analyses like this one (Hayes, 2021). Formally known as Average Causal Mediation Effects (ACMEs), this quantity allows us to have more confidence that the full pathway from shared inferiority to solidarity to support for policies implicating another outgroup is distinguishable from zero. The second of these analyses evaluates how robust these mediated relationships are to confounding since our mediator (solidarity with PoC) is measured, not manipulated.

We start with the ACME for Black people’s support for pro-Latino policies (Fig. 2a). If you recall, the effect of our inferiority treatment on Black solidarity with PoC just misses the 5% threshold for statistical significance, while the connection between solidarity and Black support for pro-Latino policies is highly significant. In essence, an ACME reveals whether the multiplication between the coefficients corresponding to both paths is reliably different than zero. Our investigation finds that the ACME in our sample of Black adults is positive (0.010), as expected, and reliable at the 8% level or better. We find a comparable set of results for our Latino sample. Specifically, we find that the ACME for our hypothesized pathway is positive (0.014) and reliable at the 7% level or better. Given the directional consistency in these results, we again meta-analyze these estimated ACMEs and find that across both African American and Latino adults, the pathway from treatment-to mediator-to outcome is significantly different from zero (Aroian test: 2.551, s.e. = 0.018, p < .001).

Building on this result, we next evaluate the sensitivity of this ACME to possible confounding in the downstream path between Black solidarity with PoC and their support for pro-Black policy. This is revealed by estimating rho (ρ), which tells us how correlated an omitted variable must be with our mediator for the proposed pathway to break down. By this metric, we find that our ACME is moderately robust, with ρ = 0.203. We also find that in this sample of Latinos, our proposed mediation mechanism is quite robust, with ρ = 0.382. Once we meta-analyze this result, we find that across our Black and Latino participants, our observed ACME is reasonably robust to confounding in the downstream path (ρ = 0.296), since an omitted variable (besides ideology) would have to be heartily associated with solidarity between people of color for this pathway to vanish to zero.Footnote 8

Discussion

We proposed that when individuals sense they are discriminated similarly to another minoritized outgroup, it causes them to feel greater solidarity with people of color, which then influences their support for policies that target and benefit that outgroup. We evaluated this proposition by manipulating Black adults’ sense that they are similarly discriminated against like Latinos as an inferior group. We found measureable evidence that this mediated chain reaction occurs just as we stipulated, with heightened solidarity with PoC leading Black adults to reliably support pro-Latino policies—a pattern that is statistically robust. Moreover, we found a similar pattern among Latinos, who, when they sense they are similarly discriminated as inferior, also become more supportive of pro-Black policies. Combined, these findings support H1 and H2 (see Table 1).

Implications

We began with an argument about America’s racial hierarchy, its configuration, and its political implications for people of color and the politics between themselves. The core of our idea is that insofar as one believes, as a member of a minoritized ingroup, that one’s community is discriminated similarly to another minoritized outgroup, a stronger sense of solidarity with people of color emerges, leading to increased public support for political initiatives that benefit other minoritized outgroups that are not one’s own. Our primary focus for this was Black people in the United States—a key group in the theatre of US interminority relations, whose specific dynamics and politics continue to be overlooked or underplayed. We therefore reasoned that an evaluation of our claims centered on the Black experience and their relations with Latinos would help us illuminate one of the more misunderstood cases of interminority conflict.

Two of our hypotheses suggested that insofar as Black Americans sense that they are discriminated against, similar to another outgroup, on the basis of their alleged inferiority, they will express more solidarity with PoC, which will then increase their support for policies that implicate the other minoritized outgroup (in this case, Latinos). Our expectations on this front were supported with evidence, not only from our sample of Black adults, but also from our sample of Latinos. Our results suggest that Black individuals who feel similarly discriminated to Latinos express more solidarity with people of color in general, which then heightens their endorsement of pro-Latino policies. Critically, this mental chain reaction is replicated among Latinos, who also become more supportive of pro-Black policies when they feel a sharpened sense of solidarity with PoC. To be sure, these findings are based on a manipulation that compactly highlighted similarity in discrimination experiences between groups (Cortland et al., 2017). Whether other manipulations with greater or lesser detail can spark the same reaction is both a theoretic and empirical matter, and one that can enhance the external validity of our results (i.e., similar effects across different participants, settings, and operationalizations of our treatment) (Campbell & Stanley, 1963). Indeed, as Druckman (2022) reminds us, all experimental designs sample from a wider universe of possible treatments and measures of mechanisms and outcomes. Viewed this way, additional research should operationalize inferiority in other ways that allow a clearer peek into the precise triggers of solidarity (e.g., perceptions of shared discrimination vs. direct experiences with discriminatory acts). We welcome future research that builds in this direction.

In addition, although our focus has been on African Americans and Latinos, we believe that our framework also applies to Asian Americans. In principle, our experimental framework should produce comparable findings among Asian Americans, provided that the axis of discrimination is the inferiority dimension, operationalized via community experiences with hate crimes, which African Americans, Asian Americans, and Latinos all experience to various degrees.

Furthermore, we think future work can further refine our understanding of how independent the two axes of subordination of minoritized groups really are (see Fig. 1). Some work finds the two dimensions are orthogonal to each other (Zou & Cheryan, 2017), while other work suggests they might be interrelated to some degree (Pérez & Kuo, 2021). Our stance is that this model of racial hierarchy is better seen, not as a full and granular representation of lived reality for minoritized groups, but rather, as a heuristic to guide the stipulation and evaluation of hypotheses related to interminority coalitions. Those efforts are likely to identify some circumstances when the experiences of some racial groups who are stereotyped as un-American but superior in the aggregate (e.g., Asian Americans) can also be construed as inferior when focused on subsets of the overarching racial category (e.g., Hmong individuals). Similarly, some racial groups who are stereotyped as inferior but American (e.g., Black individuals) might be construed as foreign when also focused on subsets of the larger racial group (e.g., Black immigrants). Our point here is simply that there is much unexplored heterogeneity within each of these four quadrants that is worthy of further theory-building and testing for the sake of effective cross-racial coalitions in politics.

We also believe that subsequent research should broaden the battery of measures available to appraise solidarity between people of color, our focal mediator. Although we have provided reasonable evidence of a mediated relationship between a racial group’s social rank, solidarity, and support for pro-outgroup policies, we also find that the measurement of solidarity can be systematically improved. Specifically, we discovered (but did not expect) our reverse-worded solidarity item to underperform as a measure of solidarity between PoC. Subsequent research should expand our set of questions in order to re-appraise solidarity between PoC in a more comprehensive way. By designing and adding more solidarity items, researchers can reassess the measurement of solidarity through more sophisticated analyses (e.g., item response theory) that require more than two items. One possibly fruitful direction here is the addition of items that draw on standard “linked fate” or “common fate” items that are influential in both political science (Dawson, 1994) and social psychology (Hornsey et al., 2003), which reflect a sense that specific goals can be achieved via cooperative efforts between individuals.

We see our reported evidence as supporting the viability of our proposed pathway, where a shared sense of inferiority stimulates greater solidarity with PoC: a key mechanism translating a sense of shared discrimination into cooperative interminority relations. Indeed, by a variety of tests (i.e., ACMEs) and diagnostics (sensitivity analyses), our results on this dimension of the hierarchy suggest that, within very specific bounds, we can consider our findings here causal. For example, the mediating role of solidarity is independent of liberal ideology (a strong correlate of political opinions among PoC) and reasonably robust to another potential confounder. Of course, with the viability of this chain reaction established, more direct evidence about the causal nature of these patterns seems worth obtaining by designing ways that manipulate both shared discrimination and a sense of solidarity (Hayes, 2021; Spencer et al., 2005).Footnote 9

Beyond the implications of our work for research on interminority politics, this pair of large-scale experiments also has implications for the “real-world” of politics. Although there are many on this front, we conclude our paper by focusing on what we think is the most important one. As our country continues to change demographically, people of color are becoming an increasingly larger segment of the mass public. Our findings suggest that building support for interminority endeavors is within our grasp. We do not see our findings as implying that one simple message will produce the political unity among disparate minoritized groups that we observed. Rather, the viability of our proposed pathways suggests that relevant stakeholders (e.g., elected officials, grassroots activists, policy-makers) can take this knowledge to create clear and compact messages that they, then, consistently disseminate to produce and sustain interminority coalitions. This last point is crucial given what we know about political communication. People of color, like other people in the U.S., regularly encounter an overwhelming and dizzying stream of information about how they should construe issues. Should African Americans, whose own communities continue to encounter a brutal police presence just worry about their own? Should Latinos, whose own communities continue experiencing immigration raids and other police-led incursions, similarly focus on their own? That would be the natural reaction, extensively documented by prior work (e.g., Tajfel et al., 1971; Wilkinson 2015; Benjamin, 2017). But what our research here suggests is that a focus on the aggressiveness of law enforcement, generally, is one that can unite and mobilize, so long as the message consistently focuses on the shared commonalities between minoritized groups; in this case, as allegedly inferior communities of color.