Bringing People Together: Improving Intergroup Relations via Group Identity Cues

Abstract

Many U.S. cities pursue a “human relations” strategy in response to racial and ethnic group conflict. Reflective of Common Ingroup Identity theory, human relations practitioners emphasize a superordinate community identity among residents from different groups for the purpose of “bringing people together” in an effort to improve intergroup relations. Practitioners also encourage intergroup contact to promote positive change in attitudes. Herein, we test the influence of group identity cues and intergroup contact as predictors of perceived intergroup commonality. The findings suggest emphasizing a superordinate community identity increases feelings of commonality in the attitudes of Anglos and Latinos toward one another and toward African-Americans and Asians, while intergroup contact has no significant influence on intergroup attitudes. These findings contribute to the extant literature by simultaneously testing the relative effect of salient group identities on intergroup attitudes and expanding the focus beyond the binary comparison found in most studies of racial–ethnic relations.

Keywords

race human relations intergroup contact superordinate identity Latinos

With the recent intergroup conflicts in cities such as Philadelphia, Charlottesville, Ferguson, Miami, and New York over immigration, police brutality, and race-related issues, examining the effectiveness of local government strategies to resolve conflict and build positive relationships between racial and ethnic groups is imperative. In reaction to unrest, cities often increase policing and other enforcement activities to bolster public safety and penalize discriminatory practices. Another, arguably complementary, policy approach emphasizes “human relations” activities that focus on “bringing people together” through education, dialogue, and a superordinate identity emphasis (Ethington and West 1998; Saunders and Bang 2007; Torres 2017; Villasenor 2015). In fact, local governments often pursue both general policies. In seeking to reduce group tensions outside of law enforcement’s purview, more than 500 U.S. cities and counties have established human relations commissions (HRCs), many of which were constituted in the aftermath of race riots or other intergroup conflicts (U.S. Department of Justice 2003).¹ Yet, scholarship is limited in assessing the effectiveness of a human relations approach for improving perceptions between diverse local groups.

We argue that HRC attempts at bringing people together overlaps with the mechanism inherent in the formation of common ingroup identity (CII; see Gaertner et al. 1993). As such, and following (Turner et al.’s 1987, p. 60) insight that “. . . the attractiveness of an individual is not constant, but varies with ingroup membership,” our study represents a critical opportunity to extend the academic literature regarding social identity effects between racial groups by testing the efficacy of practitioner-created approaches to improving group relations.

To be clear, measuring the overall performance or impact of HRCs is not our purpose here. Instead, it is to test the likelihood of improving intergroup relations via a messaging intervention utilized by HRCs that emphasizes a superordinate community identity among local residents of different racial/ethnic groups. Specifically, we assess the influence of different group identity cues on perceptions of intergroup commonality using a novel questions-as-treatment survey experiment in Los Angeles (LA) County featuring Latino and Anglo subjects. As part of the study, we also examine the relationship of intergroup contact with perceptions of intergroup commonality. We consider these messages and contact elements to be antecedents of effective HRC intergroup efforts (Riek et al. 2010; Wilkinson 2014), thereby isolating their impact before scholars undertake a more comprehensive assessment of HRCs in the literature.

The usefulness of LA residents as subjects in this study is threefold. First, LA is the second largest and one of the most racially and ethnically diverse metropolitan areas in the United States. Second, LA has a storied history of racial and ethnic strife, perhaps the best-known example being the 1965-Watts uprising. Third, and in response to the Watts conflict, the city created its HRC (LAHRC) as a government-based effort to promote healthy intergroup relations. A common LAHRC strategy has been the use of programmatic messaging in its group mediation sessions. These messages feature superordinate community identity statements, which we use as randomly assigned treatments in our experimental design (Villasenor 2015). Though we can isolate only a segment of the LAHRC’s “on-the-ground” interpersonal messaging per se and test it in a survey context, the randomized treatments allow for direct tests of messaging effects on adult subjects residing within LAHRC’s operational area. As such, our assessment provides both an examination of existing identity theories and policy insights for the LAHRC and similarly situated local government entities across the United States.

Regarding intergroup contact, our findings suggest the influence of intergroup contact on intergroup attitudes is inconclusive at best. However, emphasizing a superordinate community identity is found to significantly increase feelings of political commonality in the attitudes of both Anglos and Latinos toward one another and toward African-Americans and Asians relative to other potentially salient identities, including a distinctive subgroup identity (i.e., racial, ethnic, or religious) and a combined or dual identity. Our research contributes to the extant literature in both social psychology (on common ingroup theory) and political science (on the politics of race and ethnicity) in two ways. First, it simultaneously tests the relative effect of salient group identities² on intergroup attitudes. In other words, this study investigates the effect of each of the three group identity cues—superordinate, subgroup, and combined identities—relative to the other two group identity cues and/or the control group (which received no cue). Second, it expands the focus beyond the binary comparison found in most studies of racial–ethnic relations to include the attitudes of Anglos and Latinos toward each other and also toward Asians and African-Americans. From the standpoint of local policy makers, our study provides a systematic assessment of a commonly used tactic to reduce or diffuse intergroup tension/conflict (superordinate identity messaging). Our findings corroborate the anecdotal reports of HRC personnel and other street-level bureaucrats.

Theoretical Underpinnings and Hypotheses

Social Identity Theory suggests people have a strong tendency to categorize and identify themselves and others in terms of group membership. Self-categorization provides people with a sense of where they stand in relation to others, helping to determine appropriate attitudes and behavior (Tajfel and Turner 1986). Dividing themselves and others into different groups reduces subjective uncertainty and may increase self-esteem, but it also contributes to prejudice, discrimination, and intergroup tension as individuals tend to favor members of their ingroup (the group they identify with) over outgroup members (Hogg and Mullin 1999; Hornsby 2008). That said, individuals have multiple group identities, and research shows the salience of a particular identity can vary when people are encouraged to reflect on different identity conceptions of themselves and others (Transue 2007). The multiple identity phenomenon provides the basis for our focus on ingroup and superordinate identities.

Superordinate Identity

The CII model is an important addition to earlier work on social identity and self-categorization. CII emphasizes recategorization, whereby members of different groups are induced to conceive of themselves as a single, more inclusive superordinate group rather than as completely separate groups (Gaertner et al. 1989). This theory predicts intergroup tension is reduced when people recategorize one another so the ingroup and outgroup(s) are subsumed into a superordinate category. A superordinate group emphasis “could focus members’ attention on similarities rather than differences between the groups” (Riek et al. 2010, p. 405). There is considerable support for CII theory in studies representing multiple research situations and groups (Gaertner et al. 1999; Riek et al. 2010; Transue 2007; Ufkes et al. 2012). For example, Transue’s (2007, p. 80) survey experiment finds that Anglos support tax increases “to improve educational opportunities for minorities” when an American identity is made more salient, while Ufkes et al.’s (2012) field study of a city district in the Netherlands suggests stronger district identities lead to more positive attitudes toward residents with different cultural backgrounds.

In their summary of CII research, Gaertner and Dovidio (2009) conclude that inducing members of different groups to think of themselves in terms of a superordinate group identity leads these different groups to think about, feel, and act more positively toward each other. This may be, in part, because invoking a superordinate identity does not require individuals to forsake their subgroup identities. Social psychology’s focus on whether (and to what extent) a common ingroup or superordinate identity might mitigate “aversive racism” (Dovidio and Gaertner 2004, p. 3) (and, arguably, other forms of “latent racism”) places our test of HRC messaging to reduce conflict between racial and ethnic groups squarely within that literature’s focus (see Dovidio and Gaertner 2004; Gaertner et al. 1989; Gaertner et al. 1993; Nier et al. 2001). Our study also extends the literature in applying the CII model to intergroup relations outside the “White/Black” focus found in the majority of social psychology CII studies on racism. For its part, the political science literature fosters mixed expectations for positive effects from superordinate identity cues across racial and ethnic groups.

Though Transue (2007) did not test for shifts in intergroup commonality perceptions in his experiment, his reported outcome suggests that the CII recategorization mechanism influences Anglos toward increased commonality perceptions with persons of color. But whether Latinos might be motivated to perceive commonalities through superordinate identity cues is an open question, particularly with regard to “Brown/Black” relations, which have been characterized in urban settings by competition and distrust (Bobo, D and Massagli 2001; Borjas 1999; Johnson and Oliver 1989; Meier and Stewart 1991).

Kauffman (2003, p. 208), focusing on the issue of “perceived intergroup commonality,” suggests that growing identity inclusivity among Latinos should increase perceived commonality with African-Americans (see also Sanchez 2008), while Wilkinson (2014) finds that Latino perceptions of commonality with African-Americans decrease as feelings of economic threat increase. She further finds an inverse relationship between increased Latino political influence and commonality with Anglos. Overall, then, the safest conclusion to draw from the political science literature is that perceived intergroup commonality is a variable phenomenon that scholars have previously linked to large-scale economic and political forces.

Importantly, these macro-level conditions are far beyond the ability of any urban government, let alone HRCs, to control. The charge of HRCs and local governments to improve intergroup relations does not take the structural limitations that HRCs work under into account (Ethington and West 1998; Torres 2017). But HRCs may also have an advantage in that they can tap elite-level persuasion to shape how members of different groups view the “other.” Both Kauffman (2003) and Gay (2006) comment on the importance of community elites in fostering improved intergroup relations, even if elite-level influence is more complicated and potentially less effective than conventional wisdom supposes (see Benjamin 2017; Wallsten and Nteta 2017). HRCs, though led and staffed by members of the local community, enjoy the government’s imprimatur, suggesting that the commissions can also function as elite influencers whose messaging cues can alter intergroup perceptions.

As such, based on social psychology and political science research, we hypothesize that:

Hypothesis 1 (H1): Superordinate Identity messaging will increase perceptions of commonality with other subgroups.

Intergroup Contact

Though not explicitly part of the CII model, the Contact Hypothesis typified by Allport’s (1954) work is considered a part of the contextual backdrop that helps determine CII effectiveness (though contact can be difficult to manipulate in the confines of a survey-embedded experiment; Gaertner and Dovidio 2005). Still, since intergroup contact is a critical part of the process in determining a common or superordinate identity, we leverage intergroup contact as a statistical covariate following determination of direct treatment effects in our randomized experiment.

Due to changing demographics, U.S. residents—especially those in urban areas—interact increasingly with people from diverse backgrounds, although outcomes from contact itself may vary. HRCs and similarly tasked organizations regularly work to facilitate intergroup contact, with the expectation that contact will improve intergroup perceptions (Saunders and Bang 2007; Torres 2017; Villasenor 2015). Allport’s (1954) Intergroup Contact Theory hypothesizes contact between racial and ethnic groups is more likely to be positive when the contact involves common goals, equal group status within the situation, institutional support, and no intergroup competition. Others argue positive intergroup contact is more likely to occur when the exchange occurs voluntarily. Still, there is no guarantee intergroup contact always results in positive outcomes (Ford 1986; Hewstone and Brown 1986).

The influence of contact on intergroup relations has been the subject of numerous studies (especially in social psychology) since the 1940s. A meta-analysis of 515 studies on intergroup contact with more than 250,000 subjects shows “intergroup contact typically reduces prejudice (mean r = –.21)” and may result in other positive changes in intergroup attitudes, including greater trust, empathy, and forgiveness for past transgressions (Pettigrew et al. 2011, p. 271). Importantly, the conditions identified by Allport (1954) to facilitate positive outcomes between racial and ethnic groups are not required (Pettigrew et al. 2011). This is critical because, at least in terms of intergroup relations in American cities, the sharing of common goals, equal status, institutional support, and the absence of intergroup competition are not likely to be in alignment at any one time.

Perception of intergroup commonality is not among the outcomes specifically considered in prior research on intergroup contact, but it has been regarded as a desirable outcome or condition in other studies of intergroup relations (Sanchez 2008). While perceiving commonality with other subgroups is not necessarily tantamount to reducing prejudice between groups, it is reasonable that increased perceptions of commonality with other racial/ethnic groups is a positive outcome and an important antecedent upon which to pursue the reduction of prejudice and discrimination (Plous 2003). Therefore, and based on the Pettigrew et al. (2011) meta-analysis, we hypothesize:

Hypothesis 2 (H2): Intergroup Contact will increase perceptions of commonality with other subgroups.

Data and Method

Our data come from a 20-minute survey administered online by Qualtrics, Inc. to 945 Latino (N = 533) and Anglo (N = 418) adults residing in LA County between August 3 and 15, 2015. The survey instrument was available in both English and Spanish. Different from other studies in the literature, our design features intergroup perceptions of Latinos and Anglos toward each other and toward Asians and African-Americans. The utility of this approach is found in the growing size of the Latino community in American cities, while maintaining comparative focus on Anglos, the socially and politically dominant group (Axt, Ebersole, and Nosek 2014).

The sample was drawn from an online panel of respondents maintained by Qualtrics where subject selection for survey participation matches as closely as possible the demographic characteristics and size of the three largest groups in LA County (which also resulted in the African-American subject pool being considerably smaller) (see Note 2). The use of online opt-in “convenience” samples has become popular in political science research leveraging survey-embedded experiments, and Mullinix et al. (2015, p. 109) report “considerably similarity” in treatment effects across 20 separate research design pairings where probability and online convenience samples were compared. The authors further point out that, with the growing rate of non-response to surveys and polls using probability samples, the assumed operational validity of probability samples falls into increasing doubt. Still, in terms of external validity, there are tradeoffs made when using an Internet panel such as Qualtrics. This mainly has to do with the degree of discrepancy between the panel and U.S. Census targets for an area. It is encouraging that a recent study by Heen, Lieberman, and Miethe (2014) shows Qualtrics to have the lowest average difference in panel characteristics versus the Census of major competing panel services, while Boas, Christenson, and Glick (2018) find that Qualtrics is the most representative versus competitors (including Facebook) in recruiting online samples.

Comparing our sample with Census Bureau’s population estimates for LA County (U.S. Census Bureau 2016) indicates the sample is very close to Census estimates for age, education, and income but moderately more female (data not shown). The continued drawback, however, is that the convenience sample may still differ from a population on dimensions that are unmeasured. But, from an applied perspective, and given LAHRC’s broad, intergroup focus across such a diverse geographic and population area, we argue that findings of a direct and statistically significant treatment effect in our experiment on messaging that aligns with existing LAHRC programming provides it and similar bodies with guidance that residents respond favorably to the theme of intergroup commonality (thereby encouraging replication and extension of this and similar studies with additional samples of varying type).

That said, even if the current sample is externally valid to the LA area in terms of LAHRC goals, there is the question of whether findings related to LA generalize to the rest of the country. We suggest that our findings should transfer to other settings with diverse populations by race and ethnicity. Our analysis has perhaps less relatability in more racially homogenous locales (i.e., those that look much less like LA). As such, we speak with greater clarity to HRCs and similar bodies in more racially diverse, and likely urban, settings.

Our questions-as-treatment experiment was embedded toward the beginning of the survey, with subjects randomly assigned to a treatment or the control condition via Qualtrics algorithm. The experiment included treatment cues in the form of statements that were randomly assigned for subjects to read just prior to the outcome question of interest on the survey. A control group of subjects did not receive the treatment cue prior to the outcome question being asked. Table 1 shows the wording of the treatment questions, including the distribution of treatment and control conditions across Latino and Anglo subjects.

Table 1.

Description of Variable Wording, Assignment, and Measurement.

Treatment GroupAssignments	Wording	No. of Subjects Assigned
Superordinate Identity Cue	“Los Angeles residents—no matter their race, ethnicity, or religion—may see community problems in ways that elected officials do not. This should be taken into account in any study of problems facing Los Angeles.”	116 Anglos 135 Latinos
Subgroup Distinctiveness Cue	“Los Angeles area residents of certain racial, ethnic, or religious backgrounds may see community problems differently than others. This should be taken into account in any study of problems facing Los Angeles.”	94 Anglos 127 Latino
Combined Cue	Superordinate and Subgroup Distinctiveness Cues	102 Anglos 135 Latinos
Control (no cues)		102 Anglos134 Latinos
	Political Commonality	Scale
	“Thinking about things like government services, political power, and representation, how much does your racial or ethnic group have in common with other groups in Los Angeles today? Would you say your group has a lot in common, some in common, little in common, or nothing at all in common with [African Americans, Asians, Whites, Latinos]?”	4 = A lot in common3 = Some in common2 = Little in common1 = Nothing at all in common
	Intergroup Contact Control	Scale
	“About how many of your friends/coworkers have a difference race or ethnicity than yours?	5 = More than 204 = 9–203 = 4–92 = 1–31 = None
	Demographic Controls	Coding
	Subject gender	Female = 1
	Subject party ID	Republican = 1, Democrat = 1
	Subject age	18–72, continuous
	Subject education	1 = College educated
	Subject annual income	1 = < $50,000

The first treatment emphasizes a superordinate identity by distinguishing between the superiority of residents’ perceptions of community problems versus those of elected officials and is related to the type of superordinate identity cues that Transue (2007) tested. The Superordinate Identity Cue reads: “Los Angeles area residents—no matter their race, ethnicity, or religion—may see community problems in ways that elected officials do not. This should be taken into account in any study of problems facing Los Angeles.”

The second treatment emphasizes subgroup distinctiveness in the perceptions of community problems. The Subgroup Distinctive Cue reads: “Los Angeles area residents of certain racial, ethnic, or religious backgrounds may see community problems differently than others. This should be taken into account in any study of problems facing Los Angeles.”

The third treatment combines the prior two cues and acknowledges the superiority of residents’ perceptions of community problems over those of elected officials while recognizing subgroup differences in problem perception (the cue statements were ordered at random for display to subjects). The Combined Identity Cue reads:

Los Angeles area residents—no matter their race, ethnicity, or religion—may see community problems in ways that elected officials do not. Still, residents of certain racial, ethnic, or religious backgrounds may see community problems differently than others. This should be taken into account in any study of problems facing Los Angeles.

Finally, a Control group of subjects received no identity treatment cues prior to the outcome variable being asked in the survey.

The outcome variable is a measure of intergroup commonality taken from a posttreatment survey question that read: “Thinking about things like government services, political power, and representation, how much does your racial or ethnic group have in common with other groups in Los Angeles today?” Subjects answered the question for each group that did not reflect their self-reported race–ethnicity. In addition to our direct treatment effects, we include the covariate of interest: intergroup contact. Taking our cue from prior explanations of Allport’s (1954) theory as a proxy for intergroup contact, we ask: “About how many of your friends/coworkers have a different race or ethnicity than yours?” Our covariate models include the following demographic control variables: age, sex (1 = female), partisan affiliation (Republican and Democrat; with Independent serving as the baseline category), education level, and income level. Table 1 includes a summary of variable coding, and the descriptive statistics are provided in the appendix.

Treatment Effects

Post hoc analysis of the differences between the treatment and the control responses for the dependent variable—Intergroup Commonality—shows the experiment’s power at .90 (p < .01). Additionally, a qualitative review of subjects’ answers to an open-ended question³ included on the survey questionnaire suggests that subjects were not aware of the purpose of the randomized treatment statements, and regressions on the treatment variables indicate treatment assignment was not predicted by subject sex, race, income, age, income, or partisanship (results not shown).

Table 2 reports the direct treatment effects and contains the difference of mean results on political commonality with other groups using Tukey’s honestly significant difference (HSD) critical values (p < .01). Since we report the mean treatment effects across Latino and Anglo subject subgroups, both of the randomly assigned groups have three separate sets of means reported in Table 2 (one for each of the racial/ethnic outgroups relative to the subject’s ingroup).

Table 2.

Comparison of Means on Political Commonality with Other Groups, According to Treatments (1–4 Scale).

Subjects’ Treatment Cues		Perceived Commonality with:
Anglos (N = 418)	African-Americans	Asians	Latinos
Superordinate Identity	3.39^a	3.30^a	2.97^b
Subgroup Distinctive	2.50	2.75	2.69
Combined Identity	2.41	2.70	2.63
Control	2.36	2.59	2.49
Latinos (N = 533)	African-Americans	Asians	Anglos
Superordinate Identity	3.10^a	3.05^a	3.23^a
Subgroup Distinctive	2.75	2.37	2.45
Combined Identity	2.65	2.33	2.32
Control	2.67	2.32	2.49

p < .01 for Tukey’s HSD pairwise comparisons between the Collective Identity treatment cue and the other three randomly assigned groups, including the control.

p < .01 for Tukey’s HSD pairwise comparisons between the Collective Identity treatment cue and Combined Identity treatment cue and control groups only.

The means indicate that the Superordinate Identity cue has a consistent statistical impact on perceptions of intergroup commonality among Anglos and Latinos compared with the control group and subjects exposed to the other two treatment cues. Specifically, and in robust confirmation of H1, the Superordinate Identity Cue’s effect among Anglo subjects is significantly different (at p < .01) in the pairwise comparison from the three other randomly assigned groups. Relative to the control group, the Superordinate Identity Cue increases Anglo’s perceived commonly with African-Africans by 1.03 points, Asians by .71 points, and Latinos by .48 points on the variable’s 1–4 scale. The effect pattern is similar among Latino subjects, where the Superordinate Identity Cue increases perceived commonality with Anglos by .74, Asians by .73, and African-Americans by .43 points on the 1–4 scale (relative to the control).

Comparing the ability of the Superordinate Identity Cue to produce the intended outcome, we find Anglos moving toward greater perceptions of commonality with African-Americans over the other two groups and Latinos moving toward greater perceptions of commonality with Anglos (although Asians are a close second). Interestingly, this observed difference between Anglos and Latinos is consistent with extant literature on race relations among Anglos, Latinos, and African-Americans. For example, extant literature finds Latinos express greater perceived commonality with Anglos than with African-Americans (Dyer, Vedlitz, and Worchel 1989; McClain et al. 2006; McClain and Stewart 2002). This line of research argues that similar socioeconomic challenges experienced by Latinos and African-Americans may serve to heighten political and economic competition between the two groups (Carey et al. 2015; Gay 2006; McClain et al. 2006; McClain and Stewart 2002). The heightened competition between Latinos and African-Americans serves to strengthen Latino perceived commonality with Anglos. Alternatively, others find evidence that Anglos feel they have more in common with African-Americans than Latinos due to common history and culture (Gusterson and Besteman 2009; McClain et al. 2006; McClain and Stewart 2002). Therefore, the ordering of perceived commonality among our Latino and Anglo respondents is in line with prior research.

Covariate Models

Next, we collapsed the 4-point Likert scale dependent variables into binary responses for use in logit covariate models that include the intergroup contact measure.⁴ The models contain the three treatment cues (the control group serves as the baseline category), our additional independent variable of interest (i.e., intergroup contact), and the subject demographic variables referenced above. Reported probabilities of the dependent variables’ value are calculated as the independent variable highlighted moves from its minimum to maximum value, with all other variables held at their means.⁵ The results for Anglo subjects are presented in Table 3.

Table 3.

Anglo Subjects’ Perceived Commonality with African-Americans, Asians, and Latinos.

	African-Americans		Asians		Latinos
	Coefficient(Robust SE)	Δ Minimum → Maximum Probability	Coefficient(Robust SE)	Δ Minimum → Maximum Probability	Coefficient(Robust SE)	Δ Minimum → Maximum Probability
SuperordinateIdentity Cue	2.23*** (0.36)	.41	1.43*** (0.35)	.24	0.620** (0.29)	.14
SubgroupIdentity Cue	0.084 (0.29)	.02	0.150 (0.32)	.03	0.032 (0.31)	.01
Combined Cue	−0.126 (0.28)	−.03	0.122 (0.30)	.02	0.093 (0.29)	.03
Intergroup Contact	0.175 (0.13)	.16	0.186 (0.13)	.14	0.100 (0.13)	.07
Age	−0.001 (0.01)	−.01	0.010 (0.01)	.10	−0.002 (0.01)	−.09
Female	−0.234 (0.23)	−.05	−0.683** (0.24)	−.13	−0.541** (0.22)	−.12
Democrat	0.350 (0.026)	.08	−0.278 (0.28)	−.06	0.165 (0.25)	.03
Republican	−0.270 (0.30)	−.06	−0.656** (0.30)	−.13	−0.220 (0.28)	−.06
College graduate	−0.020 (0.24)	.00	0.082 (0.24)	.02	0.134 (0.23)	.03
Income < 50,000	−0.314 (0.25)	−.07	−0.817*** (0.27)	−.18	0.101 (0.25)	.03
Constant	−0.342 (0.74)		0.458 (0.76)		0.417 (0.71)
Baseline	.64		.73		.64
Log-likelihood	−243.7		−227.2		−260.7
Wald chi-square	60.6		44.9		14.9
Probability > F	.000		.000		.137
Pseudo-R²	.13		.10		.03
N	418		411		410

Note. Unstandardized logistic regression coefficients.

p < .10. **p < .05. ***p < .01 in two-tailed tests.

The Superordinate Identity Cue effects remain robust. In the same pattern as the difference of means reported in Table 2, the probability changes in perceived commonality among Anglo subjects for each of the three other racial groups are also arranged in largest magnitude size from African-Americans to Asians to Latinos. Specifically, exposure to the Superordinate Identity Cue treatment increases the probability of Anglos perceiving commonality with African-Americans by .41, with Asians by .24, and with Latinos by .14. Meanwhile, and in contrast to H2, Intergroup Contact seemingly has no statistically significant effect on Anglo perceptions of commonality. The gender control is statistically significant for two groups: Anglo women have a .13 lower probability of perceiving commonality with Asians and a .12 lower probability of perceiving commonality with Latinos than compared with Anglo men. Anglo Republicans have a .13 decrease in the likelihood of perceiving commonality with Asians than Anglo independents. Finally, Anglos with lower annual income have a .18 lower probability of feeling commonality with Asians. Multiplicative interactions between the statistically significant covariates and the treatment variables were not significant (results not shown).

The covariate models for Latino subjects are presented in Table 4, where the probability differences are, again, ordered in magnitude the same as the mean differences in Table 2. The Superordinate Identity Cue treatment increases the probability of perceiving commonality with African-Americans by .19, with Asians by .27, and with Anglos by .28. Again, these differences in the group ordering in perceptions among Latinos and Anglos are consistent with the existing research, which demonstrates that Latinos feel more in common with Anglos than African-Americans, while Anglos feel more in common with African-Americans than Latinos. Like the Anglo findings, Intergroup Contact has no significant effect on Latinos’ perceptions.

Table 4.

Latino Subjects’ Perceived Commonality with African-Americans, Asians, and Anglos.

	African-Americans		Asians		Anglos
	Coefficient(Robust SE)	Δ Minimum → Maximum Probability	Coefficient(Robust SE)	Δ Minimum → Maximum Probability	Coefficient(Robust SE)	Δ Minimum → Maximum Probability
SuperordinateIdentity Cue	0.957*** (0.28)	.19	1.13*** (0.27)	.27	1.19*** (0.28)	.28
SubgroupIdentity Cue	0.057 (0.26)	.01	0.079 (0.26)	.02	−0.149 (0.26)	−.04
Combined Cue	−0.027 (0.25)	−.01	0.193 (0.25)	.05	−0.133 (0.26)	−.03
Intergroup Contact	0.035 (0.10)	.03	0.090 (0.10)	.09	0.060 (0.11)	.06
Age	−0.010 (0.01)	−.08	−0.010 (0.01)	−.09	0.010 (0.01)	.05
Female	−0.102 (0.21)	−.02	−0.638** (0.20)	−.16	−0.532** (0.21)	−.13
Democrat	−0.088 (0.21)	− 02	0.077 (0.20)	.02	0.188 (0.21)	.05
Republican	−0.088 (0.31)	−.02	0.545 (0.31)	.13	0.831** (0.33)	.20
College graduate	0.180 (0.20)	.04	0.112 (0.19)	.03	0.090 (0.20)	.02
Income < 50,000	−0.184 (0.20)	−.04	−0.166 (0.19)	−.04	−0.526** (0.20)	−.13
Constant	0.843 (.590)		−0.048 (0.57)		−0.117 (0.60)
Baseline		.67		.49		.54
Log-likelihood	−330.7		−347.0		−320.0
Wald chi-square	19.2		39.5		50.7
Probability > F	.040		.000		.000
Pseudo-R²	.03		.06		.09
N	533		534		505

Note. Unstandardized logistic regression coefficients.

p < .10. **p < .05. ***p < .01 in two-tailed tests.

And, as with the Anglo findings, there are no gender differences in perceptions of commonality with African-Americans, but Latinas show a .16 lower probability of perceived commonality with Asians and a .13 lower probability of perceived commonality with Anglos, holding all variables at their means. Republican Latinos have a .20 higher probability of perceiving commonality with Anglos than compared with independent Latinos. Finally, Latinos with lower incomes have a .13 lower probability of feeling commonality with Anglos than compared with Latinos with higher incomes. Again, multiplicative interactions between these significant covariates and the treatment variables were not statistically significant (results not shown).

Discussion

The covariate model findings are important in that they show the Superordinate Identity Cue’s consistent effect even when controlling for economic and political variables. These variables help to represent the larger structural context in which individuals find themselves and suggest that messaging efforts promoting CII may cut across macrolevel differences between racial and ethnic groups. The covariate model findings provide strong support for the CII theory (H1). Specifically, cueing a superordinate identity has significant positive impact on the attitudes of both Anglos and Latinos toward other racial–ethnic groups, although the magnitude of the superordinate priming effect varies considerably. The superordinate identity cue has the largest effect on the probability of Anglos perceiving commonality with African-Americans (.41) and the smallest effect on Anglos’ attitudes toward Latinos (.14). There is less variation in the effects of superordinate priming on Latino perceptions and the magnitudes of the predicted changes are all less than the largest effect predicted for Anglos. This finding is at odds with Ufkes et al.’s (2012) finding that culturally dominant individuals are more receptive to superordinate priming than culturally subordinate individuals. However, it could be that Anglos in LA County retain a dominant group mindset (despite being in the numeric minority), meaning they view Anglo culture and concerns as more prototypical than others. By contrast, Latinos, who are now the plurality of LA County residents, may still retain a subordinate group mentality (Ufkes et al. 2012). An alternate explanation is Latino identity is more contested and complicated than Anglo identity for reasons of nativity, legal status, language skills, and so on, making Latinos less receptive to superordinate identity cues (Fraga et al. 2010; Huddy 2003; Sanchez and Vargas 2016; Schildkraut 2007). Alternatively, some research also suggests that Latinos have a lower sense of linked fate than Anglos which could be moderating the size of the priming effect (Gay, Hochschild, and White 2016). Unfortunately, we lack data at this point to test these alternate explanations and, to our knowledge, there is no existing literature that speaks directly to our findings. Additional research is needed to understand the variation in receptivity to superordinate messaging, but this variation should not distract us from the overall finding of a significant positive effect of superordinate identity priming on Anglos’ and Latinos’ perceptions of commonality toward other racial–ethnic groups (H1).

There is seemingly no confirmation of Intergroup Contact theory (H₂) in our results. For both Anglos and Latinos, the coefficients for the Intergroup Contact variable are positive across the models, but none reach statistical significance. This could be due to limitations in our data. Because there were no questions in the survey that directly assess the amount of intergroup contact (i.e., the length of time spent with people of a different race or ethnicity) or the nature of that contact (i.e., whether that contact was frequent or infrequent, voluntary or involuntary, friendly or hostile, etc.), we chose to use a proxy measure. The question used to measure intergroup contact read: “About how many of your friends/coworkers have a different race or ethnicity than yours?” In hindsight, there is the potential for imprecision in this measure. For example, due to the wording of the question, we do not know the extent to which our subjects’ intergroup communication is with friends versus coworkers. Although coworkers may be friends, and vice versa, that is not always the case, and our data do not allow us to account for this distinction. Our findings serve to underscore the complicated nature of intergroup communication, an issue Pettigrew et al. (2011) identified.

The primary finding in our analysis, that is, superordinate priming increases perceptions of commonality with outgroups, should be encouraging for HRCs that include CII cues in their local work. To be clear, increased perceptions of intergroup commonality are not tantamount to improved intergroup relations between racial and ethnic groups in American cities. But improved perceptions between ingroup and outgroup members are arguably a key part of any reform mechanism that local HRCs can leverage, particularly since perceptions of intergroup commonality are thought to be precursors to coalition building (McClain and Stewart 2002; Wilkinson 2014).

Conclusion

This study extends the research in political science and social psychology that finds superordinate identity influences attitudes and behavior (e.g., Gaertner and Dovidio 2009; Transue 2007; Ufkes et al. 2012). Herein, we present evidence that effects from messaging stimuli premised on the CII model spark increased perceptions of intergroup commonality among Anglo and Latino subjects. Substantively, the results suggest emphasizing a superordinate community identity increases feelings of political commonality in the attitudes of both Anglos and Latinos toward one another and toward African-Americans and Asians.

CII appears efficacious among the LA subjects and represents evidence for HRC policy workers to build on in their daily efforts. Still, and though our study was conducted with subjects in their local milieu, we are also mindful of Huddy’s (2001) contention that identity priming does not always lead to positive outcomes, especially in the superordinate sense we assessed herein. Therefore, and because it is always possible for people to reject attempts to have their identities seem broadly similar with those around them (especially in racially and ethnically diverse settings), scholars and policymakers should continue to monitor intergroup responsiveness to the superordinate-oriented cues for any changes in effects that new circumstances may affect.

At the same time, future assessments will need to move beyond considering the impact of the superordinate identity messaging on intergroup attitudes in a single-shot research design. Subsequent research should investigate if, when, and how the effect of emphasizing the superordinate identity strengthens or diminishes over time. Furthermore, when and how does promotion of the superordinate identity influence specific social and political behaviors, institutional perceptions, stress, neighborhood cohesion, and so on.? Focus on these broader outcomes will provide a more nuanced understanding of whether and to what extent the superordinate identity and wider CII framework improves intergroup conditions relative to other approaches.

From a real-world perspective, our work also supplements the often-anecdotal assessments of successful intergroup relationship building from HRC personnel in their local mediations between conflicting racial and ethnic groups. Human relations practitioners can leverage the evidence in our survey experiment to initiate or continue messaging strategies, with the caveat that racial/ethnic group perceptions, like messaging effects, may be ephemeral. Policymakers will want to build on these findings, perhaps by encouraging an expansion of existing efforts to have members of different racial groups meet for mediated discussions over shared concerns across groups, as well as grievances relevant to particular minority groups. Policymakers might also want to explore additional public/private funding opportunities to allow HRCs to test and develop additional insight into effects from the CII-based cues on intergroup perception, including consideration of duration effects for the treatment and whether the treatments in studies conducted in the field can be replicated in a real-world setting.

Footnotes

Appendix

Summary Statistics.

Treatment GroupAssignments	Wording	No. of Subjects Assigned^a,b
Superordinate Identity Cue	“Los Angeles residents—no matter their race, ethnicity, or religion—may see community problems in ways that elected officials do not. This should be taken into account in any study of problems facing Los Angeles.”	116 Anglos 138 Latinos
Subgroup Identity Cue	“Los Angeles area residents of certain racial, ethnic, or religious backgrounds may see community problems differently than others. This should be taken into account in any study of problems facing Los Angeles.”	95 Anglos 128 Latino
Combined Cue	Collective and Subgroup Distinctiveness Cues	107 Anglos 135 Latinos
Control (no cue)		105 Anglos134 Latinos
Political CommonalityWording and Response	Latino Frequency (%^c)	Anglo Frequency (%)
“Thinking about things like government services, political power, and representation, how much does your racial or ethnic group have in common with other groups in Los Angeles today? Would you say your group has a lot in common, some in common, little in common, or nothing at all in common with [African Americans, Asians, Whites, Latinos]?”	w/African-Americans	w/African-Americans
	4 = 107 (20.1)	4 = 81 (19.4)
	3 = 245 (46.0)	3 = 170 (40.1)
	2 = 143 (26.9)	2 = 125 (29.9)
	1 = 38 (7.1)	1 = 42 (10.1)
	w/Asians	w/Asians
	4 = 84 (15.7)	4 = 93 (22.6)
	3 = 178 (33.3)	3 = 194 (47.2)
	2 = 203 (38.0)	2 = 95 (23.1)
	1 = 69 (12.9)	1 = 29 (7.1)
	w/Anglos	w/Latinos
	4 = 105 (20.7)	4 = 63 (15.4)
	3 = 163 (32.1)	3 = 199 (48.5)
	2 = 181 (35.6)	2 = 115 (28.1)
	1 = 59 (11.6)	1 = 33 (8.1)
4 = A lot in common3 = Some in common2 = Little in common1 = Nothing at all in common
Intergroup Contact Control	Latino Frequency (%)	Anglo Frequency (%)
“About how many of your friends/coworkers have a difference race or ethnicity than yours?”	5 = 13 (2.4)	5 = 08 (1.9)
	4 = 160 (29.9)	4 = 89 (21.0)
	3 = 193 (36.1)	3 = 136 (32.2)
	2 = 144 (26.9)	2 = 162 (38.3)
	1 = 25 (4.7)	1 = 28 (6.6)
5 = More than 204 = 9–203 = 4–92 = 1–31 = None
DemographicControls	Latino Frequency (%)	Anglo Frequency (%)
Female = 1	1 = 373 (69.7) 0 = 162 (30.3)	1 = 232 (54.9) 0 = 191 (45.2)
Republican = 1^d	1 = 70 (13.1) 0 = 465 (86.9)	1 = 111 (26.2) 0 = 312 (73.8)
Democrat = 1	1 = 285 (53.3) 0 = 250 (46.7)	1 = 175 (41.4) 0 = 248 (58.6)
Age (median/mean)	40/38.6	43/43.5
College graduate = 1	1 = 191 (35.7) 0 = 344 (64.3)	1 = 145 (34.3) 0 = 278 (65.7)
Income < 50,000 = 1	1 = 210 (39.3) 0 = 325 (60.1)	1 = 103 (24.4) 0 = 320 (75.6)

A total of 535 Latino subjects were randomly assigned one of the treatment or control conditions, but two subjects did not provide responses to the dependent variable questions. The difference of mean test N is 533. There is no evidence of nonrandom attrition from the assigned groups.

A total of 423 Anglo subjects were randomly assigned one of the treatment or control conditions, but five subjects did not provide responses to the dependent variable questions. The difference of mean test N is 418. There is no evidence of nonrandom attrition from the assigned groups.

Percentages may not sum to 100 due to rounding.

The baseline category for Partisanship was Independent. There were 137 Anglo and 180 Latino Independents.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

ORCID iD

Valerie Martinez-Ebers

Author Biographies

Valerie Martinez-Ebers is the University Distinguished Research Professor of Political Science and Director of Latino and Mexican American Studies at the University of North Texas. She holds a PhD in Political Science from Ohio State University.

Brian Robert Calfano is an associate professor of Political Science and Journalism at the University of Cincinnati. He holds a PhD in Political Science from the University of North Texas.

Regina Branton is a professor of Political Science and Associate Director of Latino and Mexican American Studies at the University of North Texas. She holds a PhD in Political Science from the University of Arizona.

References

Allport

Gordon W.

1954. The Nature of Prejudice. Cambridge: Addison-Wesley.

Axt

Jordan R.

Ebersole

Charles R.

Nosek

Brian A.

2014. “The Rules of Implicit Evaluation by Race, Religion, and Age.” Psychological Science 25:1804–15.

Benjamin

Andrea

. 2017. “An Experimental Test of the Co-ethnic Elite Cues Theory.” In Racial Coalition Building in Local Elections: Elite Cues and Cross-ethnic Voting, 54–98. Cambridge: Cambridge Univ. Press. doi:10.1017/9781108233644.003

Boas

Taylor C.

Christenson

Dino P.

Glick

David M.

2018. “Recruiting Large Online Samples in the United States and India: Facebook, Mechanical Turk and Qualtrics.” Political Science Research and Methods 8 (2): 232–50.

Bobo

Lawrence D.

Massagli

Michael P.

2001. “Stereotyping and Urban Inequality.” In Urban Inequality, edited by O’Connor

Alice

Tilly

Chris

Bobo

Lawrence

, 89–162. New York: Russell Sage Foundation.

Borjas

George

. 1999. Heaven’s Door—Immigration Policy and the American Economy. Princeton: Princeton University Press.

Carey

Tony E.

Jr. Martinez-Ebers

Valerie

Matsubayashi

Tetsuya

Paolino

Phillip

. 2016. “¿Eres Amigo o Enemigo? Contextual Determinants of Latinos’ Perceived Group Competition with African-Americans.” Urban Affairs Review 52:155–81.

Dovidio

John F.

Gaertner

Samuel L.

2004. “Aversive Racism.” Advances in Experimental Social Psychology 36:1–52.

Dyer

James

Vedlitz

Arnold

Worchel

Stephen

. 1989. “Social Distance Among Racial and Ethnic Groups.” Social Science Quarterly 70:607–16.

10.

Ethington

Philip J.

West

Christopher D.

1998. “The Challenge of Intergroup Relations in Los Angeles: An Historical and Comparative Evaluation of the Los Angeles City Human Relations Commission, 1966–1998.” Department of History and The Southern California Studies Center of the Univ. of Southern California, Los Angeles City Human Relations Commission, Los Angeles.

11.

Ford

W. Scott

. 1986. “Favorable Intergroup Contact May Not Reduce Prejudice: Inconclusive Journal Evidence, 1960–1984.” Sociology and Social Research 70:256–58.

12.

Fraga

Luis

Garcia

John

Hero

Rodney

Jones-Correa

Michael

Martinez-Ebers

Valerie

Gary

Segura G

. 2010. Latino Lives in America: Making It Home. Philadelphia, PA: Temple University Press.

13.

Gaertner

Samuel L.

Dovidio

John F.

2005. “Understanding and Addressing Contemporary Racism: From Aversive Racism to the Common Ingroup Identity Model.” Journal of Social Issues 61:615–39.

14.

Gaertner

Samuel L.

Dovidio

John F.

2009. “A Common Ingroup Identity: A Categorization-based Approach for Reducing Intergroup Bias.” In Handbook of Prejudice, Stereotyping, and Discrimination, edited by Nelson

Todd D.

, 489–505. New York: Psychology Press.

15.

Gaertner

Samuel L.

Dovidio

John F.

Anastasio

Phyllis A.

Bachman

Betty A.

Rust

Mary C.

1993. “The Common Ingroup Identity Model: Recategorization and the Reduction of Intergroup Bias.” European Review of Social Psychology 4:1–26.

16.

Gaertner

Samuel L.

Mann

Jeffrey

Murrell

Audrey

Dovidio

John F.

1989. “Reducing Intergroup Bias: The Benefits of Recategorization.” Journal of Personality and Social Psychology 57:239–49.

17.

Gaertner

Samuel L.

Dovidio

John F.

Nier

Jason A.

Ward

Christine W.

Banker

Brenda S.

1999. “Across Cultural Divides: The Value of Superordinate Identity.” In Cultural Divides: Understanding and Overcoming Group Conflict, edited by Prentice

Deborah A.

Miller

Dale. T.

, 173–2012. New York: Russell Sage Foundation.

18.

Gay

Claudine

. 2006. “Seeing Difference: The effect of economic Disparity on Black Attitudes toward Latinos.” American Journal of Political Science 50:982–97.

19.

Gay

Claudine

Hochschild

Jennifer

White

Ariel

. 2016. “Americans’ Belief in Linked Fate: Does the Measure Capture the Concept?” The Journal of Race, Ethnicity, and Politics 1 (1): 117–44. doi:10.1017/rep.2015.3

20.

Gusterson

Hugh

Besteman

Catherine

, eds. 2009. The Insecure American: How We Got Here and What We Should Do about It. Berkeley: Univ. of California Press.

21.

Heen

Miliaikeala

Lieberman

Joel

Miethe

Terance

. 2014. “A Comparison of Different Online Sampling Approaches for Generating National Samples.” Center for Crime and Justice Policy 1:1–8.

22.

Hewstone

Miles

Brown

Rupert

. 1986. “Contact Is Not Enough: An Intergroup Perspective on the ‘Contact Hypothesis’.” In Contact and conflict in intergroup encounters, edited by Hewstone

Miles

Brown

Rupert J.

, 1–44. Oxford: Basil Blackwell.

23.

Hogg

Michael A.

Mullin

Barbara-A

. 1999. “Joining Groups to Reduce Uncertainty: Subjective Uncertainty Reduction and Group Identification.” In Social Identity and Social Cognition, edited by Abrams

Dominic

Hogg

Michael A.

, 249–79. Malden: Blackwell.

24.

Hornsby

Matthew

. 2008. “Social Identity Theory and Self-Categorization Theory: A Historical Review.” Social and Personality Psychology Compass 2:204–22.

25.

Huddy

Leonie

. 2003. “Group Identity and Political Cohesion.” In Sears

David O.

Huddy

Leonie

Jervis

Robert

, eds. Oxford Handbook of Political Psychology. New York: Oxford University Press.

26.

Johnson

James

Oliver

Melvin

. 1989. “Interethnic Minority Conflict in Urban America: The Effects of Economic and Social Dislocations.” Urban Geography 10:449–63.

27.

Kaufmann

Karen M

. “Cracks in the Rainbow: Group Commonality as a Basis for Latino and African-American Political Coalitions.” Political Research Quarterly 56, no. 2 (2003): 199-210. doi:10.2307/3219898.

28.

McClain

Paula

Stewart

Joseph

Jr . 2002. Why Can’t We All Get Along? Racial and Ethnic Minorities in American Politics. Boulder, CO: Westview Press.

29.

McClain

Paula

Carter

Niambi

Soto

Victoria DeFrancesco

Lyle

Monica

Grynaviski

Jeffrey

Nunnally

Shayla

Scotto

Thomas

Kendrick

Alan

Lackey

Gerald

Cotton

Kendra

. 2006. “Racial Distancing in a Southern City: Latino Immigrants’ Views of Black Americans.” The Journal of Politics 68:571–84.

30.

Meier

Kenneth J.

Stewart

Joseph

Jr.

1991. “Cooperation and Conflict in Multiracial School Districts.” Journal of Politics 53:1123–33.

31.

Mullinix

Kevin J.

Leeper

Thomas J.

Druckman

James N.

Freese

Jeremy

. 2015. “The Generalizability of Survey Experiments.” Journal of Experimental Political Science 2:109–38.

32.

Nier

Jason A.

Gaertner

Scott L.

Dovidio

John F.

Banker

Brenda S.

Ward

Christine M.

Rust

Mary C.

2001. Changing Interracial Evaluations and Behavior: The Effects of a Common Group Identity. Group Processes & Intergroup Relations, 4(4), 299–316. https://doi.org/10.1177/1368430201004004001

33.

Pettigrew

Thomas F.

Tropp

Linda R.

Wagner

Ulrich

Christ

Oliver

. 2011. “Recent Advances in Intergroup Contact Theory.” International Journal of Intercultural Relations 35:271–80.

34.

Plous

Scott

. 2003. Understanding Prejudice and Discrimination. New York: McGraw-Hill.

35.

Riek

Blake M.

Mania

Eric W.

Gaertner

Samuel L.

McDonald

Stacy A.

Lamoreaux

Marika J.

2010. “Does a Common Ingroup Identity Reduce Intergroup Threat?” Group Processes and Intergroup Relations 13:403–23.

36.

Sanchez

Gabriel R.

2008. “Latino Group Consciousness and Perceptions of Commonality with African Americans.” Social Science Quarterly 89:428–44.

37.

Sanchez

Gabriel R.

Vargas

Eduardo

. 2016. “Taking a Closer Look at Group Identity: The Link Between Theory and Measurement of Group Consciousness and Linked Fate.” Political Research Quarterly 69:160–74.

38.

Saunders

Kenneth L.

Bang

Hyo Eun

. 2007. “A Historical Perspective on U.S. Human Rights Commissions.” In Executive Session Papers: Human Rights Commissions and Criminal Justice, Edited by Beeman

Marea L.

, 1-16. Cambridge: Kennedy School of Government, Harvard Univ.

39.

Schildkraut

Deborah

. 2007. “Defining American Identity in the Twenty-First Century: How Much ‘There’ Is There?” Journal of Politics 69:597–615.

40.

Tajfel

Henry

Turner

John C.

1986. “The Social Identity Theory of Intergroup behavior.” In Psychology of Intergroup Relations, edited by Worchel

Stephen

Austin

William G.

, 7–24. Chicago: Nelson-Hall.

41.

Torres

Carlos

. 2017. Interview with director, Pittsburgh Human Relations Commission on April 15, Pittsburgh, PA.

42.

Transue

John E.

2007. “Identity Salience, Identity Acceptance, and Racial Policy Attitudes: American National Identity as a Uniting Force.” American Journal of Political Science 51:78–91.

43.

Turner

John C.

Hogg

Michael A.

Oakes

Penelope J.

Reicher

Stephen D.

Wetherell

Margaret S.

1987. Rediscovering the Social Group: A Self-Categorization Theory. Cambridge: Basil Blackwell.

44.

Ufkes

Elze G.

Otten

Sabine

Van Der Zee

Karen I.

Giebels

Ellen

Dovidio

John F.

2012. “Urban District Identity as a Common Ingroup Identity: The Different Role of Ingroup Prototypicality for Minority and Majority Groups.” European Journal of Social Psychology 42:706–16.

45.

U.S. Census Bureau. 2016. “American Community Survey 1-Year Estimates. Los Angeles County Estimates for Age, Sex, Education, Income and Homeownership Were generated by Valerie Martinez-Ebers on June 1, 2018 Using American FactFinder.” Accessed May 20, 2019 https://factfinder.census.gov/faces/nav/jsf/pages/community_facts.xhtml

46.

U.S. Department of Justice. 2003. “Guidelines for Effective Human Relations Commissions.” Accessed May 20, 2019 https://www.justice.gov/archive/crs/pubs/gehrc.pdf.

47.

Villasenor

Patricia

. 2015. Interview with Director, Los Angeles Human Relations Commission on April 29, Los Angeles, CA.

48.

Wallsten

Kevin

Nteta

Tatishe M.

2017. “Race, Partisanship, and Perceptions of Inter-Minority Commonality.” Politics, Groups, and Identities 2:298–320.

49.

Wilkinson

Betina Cutia

. 2014. “Perceptions of Commonality and Latino–Black, Latino–White Relations in a Multiethnic United States.” Political Research Quarterly 67:905–16.