#NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity

Abstract

Donald Trump won the 2016 presidential election largely due to support from White Americans. This win created a new sociopolitical reality in which White Americans as a group became associated with Trump and his anti-egalitarianism. Four studies (N = 3,245) explored how liberal-leaning White Americans negotiate their racial identity to contend with group-image threat arising from the association between their racial ingroup and Trump. Trump-related group-image threat (i.e., White Americans’ support for Trump’s anti-egalitarianism or his continuation in office) led liberal-leaning White Americans to disidentify from their racial ingroup. In turn, racial disidentification predicted greater signaling of egalitarian beliefs (i.e., expressing intentions to advocate for racial equity and supporting policies designed to benefit racially minoritized groups) and behaviors (i.e., donating money to racial equity-focused organizations). These results suggest that the process of negotiating Trump-related group-image threat has implications for both White Americans’ racial identities and ongoing efforts to achieve racial equity.

Keywords

White identity racial disidentification group-image threat Trump egalitarian

“It’s draining to live everyday knowing that most of your own demographics are soulless at best and evil at worst. #NotAllWhites.”

—Jake Turk (@71djt, Twitter, July 24, 2018)

While politics are notoriously divisive, the 2016 presidential election exposed deep political and cultural divisions in the United States (Jacobson, 2016). Then-candidate Donald Trump denigrated and proposed policies adversely targeting minoritized groups (e.g., Muslims, Latin@s, immigrants, and women; Conway et al., 2017). Despite tremendous public backlash, Trump won the election, in large part, due to White Americans’ support. Exit polls revealed that the majority of White Americans (54%) voted for Trump, with White Americans comprising 88% of Trump’s voter base (Pew Research Center, 2018a). Throughout his presidency, Trump continued to promote anti-egalitarian ideas and policies (e.g., separating migrant parents and children and defending White supremacists; Parker & Nakamura, 2017; Southern Poverty Law Center, 2020). Nonetheless, White Americans’ support persisted, with polls consistently showing higher Trump approval ratings among White Americans relative to other racial groups (Gallup, 2018, 2019). As a result, a new sociopolitical reality emerged in which White Americans became associated with Trump’s anti-egalitarian views. While this new reality may not pose a problem for White Trump supporters, liberal-leaning White Americans may experience psychological tension arising from the association of their racial group with anti-egalitarianism (Kiley, 2017).

Indeed, since the election, many liberal-leaning White Americans responded to this sociopolitical reality by racially disidentifying (i.e., condemning, differentiating, and distancing from their racial ingroup), as reflected in the #NotAllWhites Tweet above. Evidence of White racial disidentification also appeared in actions outside of social media. Following the murder of George Floyd, many liberal-leaning White Americans marched alongside Black and Brown demonstrators, holding signs denouncing White privilege and complicity in racism (e.g., “White silence is violence”). In fact, nearly half of protesters nationwide identified as White, and the vast majority as liberal-leaning (Harmon & Tavernise, 2020; Pew Research Center, 2020).

This article empirically examines how this sociopolitical reality shapes how liberal-leaning White Americans negotiate the tension between their personal egalitarian beliefs and the anti-egalitarian assumptions associated with their racial ingroup. Drawing on group-image threat and racial identity research, we explore the psychological and behavioral consequences of the White American–Trump association. We theorize that liberal-leaning White Americans contend with these consequences by disidentifying from their racial group and signaling egalitarianism (e.g., supporting efforts to end racism).

Group-Image Threat and White Identity in the Sociopolitical Climate During Trump’s Presidency

Because people use social identities to define and evaluate themselves (Hogg, 2003; Tajfel & Turner, 1986), seeing one’s ingroup in a positive light is beneficial for positive self-views (Crocker et al., 1994; Tajfel & Turner, 1979). When individuals encounter negative portrayals of their ingroup (Turner et al., 1987), including stereotypes (Barrie et al., 2016; Ward, 2004), negative group evaluations (Rabinovich et al., 2014; Rabinovich & Morton, 2010), and salient ingroup transgressions (Doosje et al., 1998; Shuman et al., 2018), self-views come under threat. Negative ingroup portrayals conflict with the individuals’ need to maintain positive self-views, creating psychological tension known as group-image threat (Iyer et al., 2007; Knowles et al., 2014; Rabinovich & Morton, 2010). To protect their self-views, individuals must resolve the psychological tension arising from these portrayals (Sedikides & Gregg, 2003).

One way individuals respond to group-image threat is by disengaging from their threatened identity. Specifically, individuals can suppress their group identity (i.e., reduce the extent to which they view the identity as central to their sense of self; Branscombe et al., 2012; Hodson & Esses, 2002; Marshburn & Knowles, 2018) or disidentify from the group (i.e., actively rejecting one’s group membership, which manifests as feeling detached from, dissatisfied with, and dissimilar to other group members; Becker & Tausch, 2014; Matschke & Fehr, 2017). For example, encountering negative stereotypical group representations leads racial/ethnic minorities to suppress (Arndt et al., 2002; Jasinskaja-Lahti et al., 2009) and disidentify (Mussweiler et al., 2000; Yip, 2016) from their racial/ethnic group.

Although White identity is understudied relative to racially minoritized identities, researchers and theorists increasingly highlight the dynamism and complexity of White identity (Knowles & Peng, 2005; McDermott & Samson, 2005; Wong & Cho, 2005). Research shows that the extent to which White Americans are conscious of and embrace their racial identity depends, in part, upon how Whiteness and White Americans’ social positioning are constructed (Jardina, 2019; Knowles & Peng, 2005; McDermott & Samson, 2005). Moreover, experimental data reveal that, like members of other racial groups, White Americans may disengage from their racial identity in response to group-image threat. Specifically, White Americans suppress their racial identity when confronted with negative ingroup portrayals highlighting White privilege (Chow et al., 2008; Knowles et al., 2014; Powell et al., 2005). For instance, upon learning that the Black–White achievement gap arises from systemic biases favoring White people, White Americans reported lower racial identity centrality (Chow et al., 2008).

We contend that the political climate during Trump’s presidency, in which White Americans are associated with Trump and his anti-egalitarian views, shapes how White Americans engage their racial identities. For conservative-leaning White Americans, who support Trump, this association may not create tension or foster racial disengagement. In contrast, liberal-leaning White Americans are likely to experience group-image threat arising from the conflict between anti-egalitarian portrayals of their racial group and their pro-egalitarian self-views. Research on White identity has yet to explore whether White Americans racially disidentify in response to group-image threat. However, given the real-world evidence that many liberal-leaning White Americans expressed dissatisfaction with their racial group (e.g., #NotAllWhites on Twitter and participating in George Floyd protests), we propose that liberal-leaning White Americans will actively disidentify from their racial group in response to Trump-related group-image threat.

Does Disidentification Facilitate Pro-Egalitarian Social Changes?

Racial identities are social identities that extend beyond the individual. Research reveals that how White Americans engage their racial identity has seemingly inconsistently effects on how they think about and behave toward ingroup (e.g., #NotAllWhites) and outgroup members (e.g., participation in George Floyd protests). On one hand, weak White identification predicts greater sympathy for Black Americans, more guilt about White privilege, stronger support for policies promoting diversity, and less prejudice (Andreychik & Gill, 2009; Branscombe et al., 2007; Jardina, 2020; Levin et al., 1998; Sidanius et al., 2004; Swim & Miller, 1999). On the other hand, weak White identification predicts more negative attitudes toward diversity, denial of White privilege, opposition to affirmative action, and less respect for racial minorities (Goren & Plaut, 2012; Knowles et al., 2014; Linnehan et al., 2006).

One possible explanation for White identity’s seemingly contradictory intergroup implications is that the consequences of White disengagement depend upon the underlying cause or motive for disengagement. In this article, we propose that liberal-leaning White Americans racially disidentify to protect their egalitarian self-image, which is threatened by the association between their racial group and Trump’s anti-egalitarian views. We predict that liberal-leaning White Americans’ racial disidentification, driven by the defense of their pro-egalitarian self-images, will increase their likelihood of signaling egalitarian beliefs and engaging in egalitarian behaviors.

The Current Studies

Across four studies, we explore a contemporary form of group-image threat among liberal-leaning White Americans, namely, the White American–Trump association. We examine whether Trump-related group-image threats lead liberal-leaning White Americans to racially disidentify (Studies 1a–3). Moreover, we examine whether disidentification has positive implications for intergroup relations (Studies 2 and 3) such that racial disidentification leads liberal-leaning White Americans to signal and act upon their egalitarian beliefs.

Studies 1a and 1b

Studies 1a and 1b examined the effect of Trump-related group-image threat on liberal-leaning White Americans’ racial disidentification. We manipulated group-image threat via information about the extent to which White Americans agree with Trump’s anti-egalitarianism. We hypothesized that information suggesting that White Americans agree (vs. disagree) with Trump would increase racial disidentification among liberal-leaning White participants.

Method

Participants

Study 1a

We recruited 509 non-Hispanic, U.S. born, mono-racial White adults from Amazon’s Mechanical Turk (MTurk) to participate in a 15-min online survey in exchange for US$1.50.¹ In this and all subsequent studies, we excluded participants who failed to correctly respond to both manipulation check questions, as specified during preregistration. Across all four studies, results and interpretations remain unchanged when analyses include excluded participants (see Table S1 in the Supplemental file for analyses). In Study 1a, we excluded 31 participants, leaving a final sample of 478 (57% women²; M_age = 37.08 years; 46% without a 4-year college degree). A sensitivity analysis indicated that this sample size can detect small effects (f² ≥ .027) with 80% power (Faul et al., 2009).

Study 1b

We recruited a separate sample of 1,069 non-Hispanic, U.S. born, mono-racial White adults from MTurk to participate in a 15-min online survey in exchange for US$1.50. Using the same preregistered exclusion criteria as Study 1a, we excluded 64 participants, giving a final sample of 1,005 participants (51% women; M_age = 37.73 years; 43.5% without a 4-year college degree). A sensitivity analysis indicated that this sample size can detect small effects (f² ≥ .011) with 80% power (Faul et al., 2009).

Procedure

Procedures for Studies 1a and 1b were identical, except Study 1a included a no-information control condition, whereas Study 1b did not.³ Thus, we describe the procedures for both studies together.

Participants completed a measure of political ideology and then were randomly assigned to condition: Whites Support Trump, Whites Oppose Trump, or a no information control condition (Study 1a only). Participants in the control condition proceeded directly to the measures, while participants in the Whites Support Trump and Whites Oppose Trump conditions first viewed results from an ostensible survey describing White Americans’ agreement with five of Trump’s anti-egalitarian statements about people of color and immigrants (see Figure 1; see Supplemental File “Materials” for full stimuli).

The Whites Support Trump condition was intended to amplify group-image threat for liberal-leaning Whites by depicting White Americans as agreeing with Trump’s anti-egalitarianism:

Last month, a nonpartisan think tank conducted a survey with a nationally representative sample of 12,000 White adults (age 18-82) to assess the extent to which they agreed with President Trump’s beliefs about different social groups and issues. Overall, the survey revealed that the majority of White adults agreed with President Trump’s opinions and beliefs. This was especially true with respect to issues related to immigration and race.

This condition then included pie charts of survey results indicating that 60% to 70% of White respondents agreed with five anti-egalitarian Trump views (see Figure 1).

Figure 1.

Example pie chart presented to participants in the Whites Support Trump condition (A) and the Whites Oppose Trump condition (B) used in Studies 1a-2.

The Whites Oppose Trump condition suggested that most White Americans disagreed with Trump’s views. Participants viewed the same survey description as in the Whites support Trump condition. However, the description stated that “Overall, the survey revealed that the majority of White adults disagreed with President Trump’s opinions and beliefs.” In the corresponding pie charts, we reversed the percentages depicted in the Whites Support Trump condition to indicate that only 30% to 40% of White respondents agreed with Trump’s anti-egalitarianism (see Figure 1). This range of agreement reflected Trump’s approval rating at the time of Study 1a, which fluctuated between 30% and 40% (FiveThirtyEight, n.d.).

After reviewing survey results, all participants completed a measure of racial disidentification, followed by manipulation checks (not included in the control condition) and demographic questions.

Measures

Political ideology

One item assessed the extent to which participants’ beliefs leaned liberal versus conservative: “When you think about your political beliefs, how would you describe yourself in general?” (1 = Extremely conservative, 2 = Conservative, 3 = Somewhat conservative, 4 = Slightly conservative, 5 = Slightly liberal, 6 = Somewhat liberal, 7 = Liberal, 8 = Extremely liberal). Overall, the sample leaned liberal (Study 1a: M = 4.96, SD = 2.13; Study 1b: M = 4.90, SD = 2.24). See Table 1 for the distribution of political ideology in all studies.

Table 1.

Distribution of Political Ideology in Study 1a, 1b, 2, and 3 Samples.

Political ideology	Study 1a (%)	Study 1b (%)	Study 2 (%)	Study 3 (%)
Extremely conservative	5.6	5.6	4.7	5.6
Conservative	12.1	16.9	13.9	15.4
Somewhat conservative	7.5	8.5	9.1	6.5
Slightly conservative	16.3	12.4	14.6	12.8
Slightly liberal	18.2	12.4	16.1	15.0
Somewhat liberal	7.7	9.6	10.1	7.3
Liberal	18.2	21.5	18.5	21.0
Extremely liberal	14.2	13.1	13.0	13.4

Racial disidentification

We assessed racial disidentification, or the extent to which participants felt detached from (i.e., cognitive separation between self-concept and group identity), dissatisfied with (i.e., negative evaluations of the group), and dissimilar to (i.e., seeing the self as different from other group members) White Americans. We used nine items (randomized) adapted from Becker and Tausch’s (2014) Measure of Disidentification to refer to White Americans (e.g., detachment: “I feel a distance between myself and White Americans as a group”; dissatisfaction: “Being a White American gives me a bad feeling”; dissimilarity: “I have nothing in common with most White Americans.”). We averaged across items such that higher scores indicated greater racial disidentification (Study 1a: α = .94, M = 3.38, SD = 1.47; Study 1b: α = .94, M = 3.32, SD = 1.45).⁴

Manipulation checks

Exclusion Criteria. Two questions assessed whether participants correctly interpreted the survey results. Both questions were multiple-choice and included options stating, “I am not sure” and “I do not remember.” Participants who chose one of these options were excluded. One question asked, “Which racial group did the survey respondents belong to?” The correct response in both conditions was “White American.” The second question asked participants to select one of two statements that best described the survey results. The correct response was “The majority of respondents agreed with Trump’s perspectives on race and immigration issues” in the Whites Support Trump condition, and “The majority of respondents disagreed with Trump’s perspectives on race and immigration issues” in the Whites Oppose Trump condition.

Perceived Credibility of Manipulation. Two items (randomized) assessed the extent to which participants found the manipulation credible: “The survey results are similar to what I have heard from other media sources” and “I believe that the survey results are accurate”; 1 = Strongly disagree to 7 = Strongly agree. We averaged across items such that higher scores indicated higher perceived credibility (Study 1a: α = .86, M = 4.34, SD = 1.66; Study 1b: α = .82, M = 4.00, SD = 1.55). We did not exclude participants based on their responses to these two items.

Trump’s Perceived Representativeness (Study 1b only). To ensure that our manipulation materials shaped participants’ perception of the association between White Americans and Trump, we included an additional three-item manipulation check (randomized) in Study 1b assessing the extent to which participants believed Trump’s views represented White Americans (e.g., “I think that President Trump’s values reflect the values of most White Americans”; 1 = Strongly disagree to 7 = Strongly agree). We averaged across items such that higher scores indicated greater perceived representativeness (α = .88, M = 3.31, SD = 1.73).

See Table S4 in the Supplemental file for bivariate relationships between all variables in this and all subsequent studies.

Analytic Strategy

Across all studies, our central hypothesis was that liberal-leaning White Americans would racially disidentify as a result of Trump-related group-image threat. To test this hypothesis, we conducted linear regression analyses predicting racial disidentification from political ideology, condition, and their interaction. In all models, we mean-centered political ideology and dummy coded condition, with the Whites Support Trump condition (i.e., group-image threat for liberal-leaning Whites) as the reference category. To probe interactions, we followed the pairwise inference method using the Johnson–Neyman (J-N) technique (Johnson & Neyman, 1936) outlined by Hayes and Montoya (2017).⁵ Specifically, we identified the point on the political ideology scale at which condition had a significant effect on racial disidentification. We conducted all regression analyses and J-N procedures in SPSS version 25 using the PROCESS macro (Hayes, 2018).

Results

Because Study 1a and 1b results were highly consistent, we describe them together, noting divergence.

Manipulation checks

Credibility ratings were significantly above the scale midpoint in Study 1a, M = 4.34, SD = 1.66, t(320) = 3.69, d = .21, p < .001, and did not differ from the midpoint (i.e., “Neither disagree nor agree”) in Study 1b, M = 4.00, SD = 1.55, t(1,004) = .07, d = .02, p = .945. In other words, Study 1a participants found the manipulation relatively credible, and Study 1b participants found the manipulation neither high nor low in credibility.

Notably, across both studies, participants in the Whites Oppose Trump condition (Study 1a: M = 4.52, SD = 1.58; Study 1b: M = 4.10, SD = 1.52) viewed the manipulation information as more credible than participants in the Whites Support Trump condition, Study 1a: M = 4.16, SD = 1.72, t(319) = 1.96, d = .22, p = .051; Study 1b: M = 3.90, SD = 1.57, t(1,003) = 2.02, d = .12, p = .044. In this and all subsequent studies, results and interpretations remained unchanged when analyses include perceived credibility as a covariate (see Table S5 in the Supplemental file).

Trump’s perceived representativeness (Study 1b)

Importantly, the additional manipulation check in Study 1b indicated that the manipulation had a small but statistically significant effect on the extent to which participants perceived Trump as representative of White Americans. Participants in the Whites Support Trump condition (M = 3.53, SD = 1.74) perceived Trump as significantly more representative of White Americans than those in the Whites Oppose Trump condition (M = 3.10, SD = 1.69), t(1,003) = 4.05, d = .26, p < .001.

Does group-image threat increase racial disidentification among liberal-leaning White Americans?

Following the analytic strategy, we tested the hypothesis that exposure to information that the majority of White Americans agree with Trump’s anti-egalitarianism (Whites Support Trump condition) would amplify disidentification among liberal-leaning Whites, compared with both information that the majority disagree (Whites Opposed Trump condition) and the no information control condition (Study 1a only).

In both Studies 1a and 1b, the interaction between political ideology and Whites Oppose Trump versus White Support Trump conditions predicted racial disidentification (see Table 2 for model results; Figure 2). That is, the effect of political ideology on racial disidentification differed for participants in Whites Support Trump and Whites Oppose Trump conditions. Using the J-N technique to probe this interaction revealed that, as predicted, liberal leaning White participants disidentified more in the Whites Support Trump condition than in the Whites Oppose Trump (Study 1a region of significance: above 5.23 on the political ideology scale; Study 1b region of significance: above 2.10 on the political ideology scale). Interestingly, for Study 1b, greater disidentification also occurred among moderately conservative-leaning White participants.

Table 2.

Studies 1 to 3 Results of Models Predicting Racial Disidentification With Whites Support Trump Condition as the Reference Category.

Variable	B (SE)	Study 1a			95% CI_β	B (SE)	Study 1b			95% CI_β
Variable	B (SE)	β	t	p	95% CI_β	B (SE)	β	t	p	95% CI_β
Intercept	3.58 (.11)	.14	1.91	.057	[−.004, .28]	3.85 (.05)	.37	11.18	<.001	[.30, .43]
Political ideology	0.40 (.05)	.57	8.15	<.001	[.44, .71]	.51 (.02)	.79	23.94	<.001	[.73, .86]
X1 (Whites Support Trump vs. Whites Oppose Trump)	−0.24 (.15)	−.16	1.60	.110	[−.36, .04]	−1.02 (.07)	−.71	15.52	<.001	[−.79, −.62]
X2 (Whites Support Trump vs. Control)	−0.40 (.15)	−.27	2.64	.008	[−.47, −.07]	—	—	—	—	—
Political Ideology × X1	−0.15 (.07)	−.22	2.16	.002	[−.50, −.11]	−.29 (.03)	−.45	9.89	<.001	[−.54, −.36]
Political Ideology × X2	−0.21 (.07)	−.30	3.06	.031	[−.43, −.02]	—	—	—	—	—
Identical model using the Whites Oppose Trump condition as the reference category
X3 (Whites Oppose Trump vs. Control)	−0.16 (.15)	−.11	1.06	.288	[−.31, .09]
Political Ideology × X3	0.05 (.07)	.08	.76	.449	[−.12, .28]
Variable	B (SE)	Study 2			95% CI_β	B (SE)	Study 3			95% CI_β
Variable	B (SE)	β	t	p	95% CI_β	B (SE)	β	t	p	95% CI_β
Intercept	3.37 (.06)	.16	3.14	.002	[.06, .25]	3.39 (.07)	.15	2.93	.004	[.05, .25]
Political ideology	0.39 (.03)	.64	13.17	<.001	[.54, .73]	.32 (.03)	.56	10.86	<.001	[.46, .66]
X1 (Whites Support Trump vs. Whites Oppose Trump)	−0.30 (.09)	−.23	3.27	.001	[−.37, −.09]	−.25 (.09)	−.20	2.66	.008	[−.34, −.05]
X2 (Whites Support Trump vs. Control)	−0.32 (.09)	−.24	3.51	<.001	[−.38, −.11]	−.31 (.09)	−.25	3.44	<.001	[−.39, −.11]
Political Ideology × X1	−0.15 (.04)	−.24	3.41	<.001	[−.38, −.10]	−.20 (.04)	−.35	4.78	<.001	[−.49, −.21]
Political Ideology × X2	−0.07 (.04)	−.11	1.56	.120	[−.24, .03]	−.08 (.04)	−.15	1.99	.046	[−.29, −.002]
Identical model using the Whites Oppose Trump condition as the reference category
X3 (Whites Oppose Trump vs. Control)	−0.01 (.09)	−.01	.13	.895	[−.15, .13]	−.07 (.09)	−.05	.73	.468	[−.20, .09]
Political Ideology × X3	0.08 (.04)	.13	1.85	.064	[−.01, .27]	.12 (.04)	.20	2.76	.006	[.06, .35]

Note. Study 1a df = 471; Study 1b df = 1001; Study 2 df = 842; Study 3 df = 907. SE = standard error.

Figure 2.

The relationship between political ideology and racial disidentification across all studies (A: Study 1a; B: Study 1b, C: Study 2, and D: Study 3).

Study 1a also revealed a significant interaction between political ideology and the Whites Support Trump versus Control conditions. In line with hypotheses, the J-N analysis revealed that liberal-leaning White participants disidentified more in the Whites Support Trump condition than in the Control condition (Study 1a region of significance: above 3.41 on the political ideology scale).

The interaction between political ideology and the Whites Oppose Trump versus Control conditions of Study 1a was not significant, suggesting relatively equivalent levels of threat in the two conditions for all participants. Study 1b did not include a control condition for analogous comparisons.

Together, these findings support our hypothesis that learning that the majority of White Americans agree with Trump’s anti-egalitarianism increased racial disidentification among liberal-leaning participants, relative to learning that the majority of White Americans disagree, or relative to the control condition.

Discussion

As hypothesized, Studies 1a and 1b demonstrated that group-image threat, operationalized as White Americans’ support for Trump’s anti-egalitarianism, increased racial disidentification among liberal-leaning White Americans. However, Study 1b showed that moderately conservative-leaning participants also racially disidentified when they learned that White Americans support (vs. oppose) Trump’s anti-egalitarianism. One possibility is that conservative-leaning participants in Studies 1a and 1b differed in ways that our study design cannot identify (e.g., because we used a single-item measure of political ideology). Another possibility is that real world events that occurred after Study 1a changed the nature of the effect for conservative-leaning participants. For example, we conducted Study 1b one day before the release of the Mueller report regarding Russian interference in the 2016 election (Figure 5). Given the immense media attention to the investigation, the Whites Support Trump condition may have led participants to associate their racial ingroup with Trump’s purported crimes in addition to anti-egalitarianism. This additional association may have created group-image threat among a wider range of White Americans, including conservatives who were less supportive of Trump.

In addition, although participants in Study 1b found the manipulation neither high nor low in credibility, by assessing Trump’s perceived representativeness, Study 1b provided additional support for the manipulation’s effectiveness. Participants perceived Trump as more representative of White Americans, suggesting the manipulation amplified the association between White Americans and Trump, as intended. These results also support our assertion that the White American–Trump association is a key component of the group-image threat liberal-leaning White Americans experienced in the sociopolitical moment during Trump’s presidency.

Study 2

Study 2 replicated and extended Studies 1a and 1b by exploring the intergroup implications of White Americans’ racial disidentification. We hypothesized that Trump-related group-image threat would increase racial disidentification among liberal-leaning White Americans, leading them to affirm their personal values by signaling egalitarian beliefs.

Method

Participants

We recruited 923 non-Hispanic, U.S. born, mono-racial White adults from MTurk, who had not participated in previous studies, to participate in exchange for US$1.50. Using the same exclusion criteria as Studies 1a and 1b, we excluded 71 participants, giving a final sample of 849 participants (57% women; M_age = 40.22 years; 48.5% without a 4-year college degree). A sensitivity analysis indicated that this sample size can detect small effects (f² ≥ .015) with 80% power (Faul et al., 2009).

Procedure

Study 2 used the same procedure, manipulation (Whites Support Trump, Whites Oppose Trump, and no information Control), and measures of political ideology (M = 4.93, SD = 2.13) and racial disidentification (α = .95, M = 3.16, SD = 1.30) as Study 1a. The exception being that (a) after completing the racial disidentification measure, participants completed the perceived credibility (α = .78, M = 4.07, SD = 1.56) and perceived representativeness (α = .89, M = 3.11, SD = 1.66) manipulation checks; and (b) participants responded to three measures assessing egalitarian intentions.

Measures

Intentions to advocate for racial equity (close-ended)

We assessed intentions to advocate for racial equity using a five-item scale (randomized) adapted from the individual advocacy subscale of Pieterse et al.’s (2016) Anti-Racism Behavioral Inventory (e.g., “When I hear people telling racist jokes or using racial stereotypes, I will confront them”; 1 = Extremely unlikely to 7 = Extremely likely). We averaged across items such that higher scores indicated stronger intentions to advocate for racial equity (α = .81, M = 4.73, SD = 1.34).

Intentions to advocate for racial equity (open-ended)

We also assessed intentions to advocate for racial equity using an open-ended question: “What will you personally commit to doing in the future to reduce inequality (e.g., health, wealth, education, employment, housing, incarceration, etc.) between White people and racially minoritized groups (e.g., Black, Latino, and Native Americans)?” Two trained coders blind to hypotheses and condition coded the number of actions participants listed (κ = 0.92, p < .001; see Supplemental File “Study 2 Codebook”). Listing more actions indicated stronger intentions to advocate for racial equity (M = 1.17, SD = 1.23).

Policy support

We assessed support for policies designed to benefit racially minoritized groups using an eight-item scale (randomized) adapted from Iyer et al.’s (2003) Support for Compensatory Policy and Support of Equal Opportunity Policy measures (e.g., “Due to a history of discrimination, American society has a responsibility to create special programs to provide Black, Latino, and Native American people with jobs”; 1 = Strongly disagree to 7 = Strongly agree). We averaged across items such that higher scores indicated stronger policy support (α = .96, M = 4.11, SD = 1.66).

Results

Perceived credibility of manipulation

Consistent with Study 1b, credibility ratings (M = 4.07, SD = 1.56) did not significantly differ from the scale midpoint (4), t(547) = .98, d = .04, p = .329. In other words, participants found the information neither high nor low in credibility. Consistent with both Studies 1a and 1b, participants in the Whites Oppose Trump condition (M = 4.21, SD = 1.55) found the manipulation to be more credible than those in the Whites Support Trump condition (M = 3.93, SD = 1.72), t(546) = 2.11, d = .18, p = .035.

Trump’s perceived representativeness

As in Study 1b, the manipulation had a small but statistically significant effect on the extent to which participants perceived Trump as representative of White Americans. Participants in the Whites Support Trump (M = 3.32, SD = 1.66) condition perceived Trump to be significantly more representative of White Americans compared with participants in the Whites Oppose Trump condition (M = 2.89, SD = 1.65), t(546) = 3.03, d = .26, p = .003.

Does group-image threat increase racial disidentification among liberal-leaning White Americans?

Following the analytic strategy outlined for Studies 1a and 1b, we tested the hypothesis that liberal-leaning White Americans would racially disidentify when exposed to information suggesting that the majority of White Americans agree (Whites Support Trump condition) versus disagree (Whites Oppose Trump condition) with Trump’s anti-egalitarianism. As hypothesized, and consistent with prior studies, the interaction between political ideology and the Whites Oppose Trump versus Whites Support Trump conditions predicted racial disidentification (see Table 2 for model results). In line with hypotheses, the J-N analysis revealed that liberal-leaning White participants disidentified more in the Whites Support Trump compared with Whites Oppose Trump condition (region of significance: above 4.17 on the political ideology scale).

Diverging from Study 1a, the interaction between political ideology and the Whites Support Trump versus Control conditions predicting racial disidentification was not statistically significant. Because an omnibus interaction test is often insensitive, especially when involving a non-crossover (vs. crossover) interaction, it can fail to reject a null hypothesis even when the particular predicted pattern is visually evident (Rosnow & Rosenthal, 1989). Moreover, statistical significance is not a prerequisite for probing an interaction term (Hayes & Montoya, 2017). Therefore, as in prior studies, we probed this interaction using J-N analysis, which corroborated our hypothesized pattern (Figure 2): liberal-leaning White participants disidentified more in the Whites Support Trump condition than in the Control condition (region of significance: above 3.71 on the political ideology scale).

Finally, replicating Study 1a, the interaction between political ideology and the Whites Oppose Trump versus Control conditions was not significant, suggesting relatively equivalent levels of threat in the two conditions for all participants.

Together, these results suggest that learning that the majority of White Americans agree with Trump’s anti-egalitarianism increased racial disidentification among liberal-leaning White participants.

Does racial disidentification subsequently lead to signaling of egalitarian intentions?

To examine whether racial disidentification stemming from Trump-related group-image threat would lead liberal-leaning White Americans to express more egalitarian intentions, we conducted a series of moderated mediation analyses using PROCESS Model 7, seed 27, with 5,000 bootstrap resamples (see Figure 3 for the diagram of the tested model). By specifying the seed, each model tested the same effects of political ideology, condition, and their interaction on racial disidentification as reported above (see PROCESS documentation for discussion of multiple dependent variables; Hayes, 2018). Only the estimates of the effect of racial disidentification on egalitarian intentions differed across models, as each model assessed the effect of racial disidentification on a different operationalization of egalitarian intentions. We used bootstrapping estimation to probe conditional indirect effects (1SD below political ideology mean vs. mean vs. 1SD above mean). We did not use the J-N procedure because the sampling distribution of conditional indirect effects violated the J-N procedure assumption (Hayes, 2018).

Figure 3.

Conceptual representation of the moderated mediation model tested in Studies 2 and 3.

Across all three measures, greater racial disidentification predicted greater signaling of egalitarian intentions (all ps < .001; see Table 3 for model results; also see Table S6 in the Supplemental file for similar results from a more saturated model). Moreover, the indirect effect of political ideology on egalitarian intentions through racial disidentification was moderated by condition. Consistent with our hypothesis, racial disidentification in response to information that the majority of White Americans agree with Trump’s anti-egalitarianism increased liberal-leaning (vs. conservative-leaning) White participants’ intentions to advocate for racial equity and support policies designed to benefit racial minorities.

Table 3.

Study 2 Regression Results for Moderated Mediation Models Predicting Egalitarian Intentions: Close-Ended Intentions to Advocate for Racial Equity (Model 1), Open-Ended Intentions to Advocate for Racial Equity (Model 2), and Policy Support (Model 3), With the Whites Support Trump Condition as the Reference Category.

Predictor	Racial disidentification (mediator variable model)				95% CI
Predictor	b	SE	t	P	95% CI
Intercept	.20	.06	3.15	.002	[.08, .33]
X1 (Whites Support Trump vs. Whites Oppose Trump)	−.30	.09	3.27	.001	[−.49, −.12]
X2 (Whites Support Trump vs. Control)	−.32	.09	3.51	<.001	[−.49, −.14]
Political ideology	.39	.03	13.17	<.001	[.33, .45]
X1 × Political Ideology	−.15	.04	3.40	<.001	[−.23, −.06]
X2 × Political Ideology	−.07	.04	1.56	.120	[−.15, .02]
Predictor	Close-ended intentions to advocate for racial equity (dependent variable Model 1)				95% CI
Predictor	b	SE	t	P	95% CI
Intercept	4.65	.08	62.56	<.001	[4.50, 4.79]
X1	0.09	.11	0.87	.384	[−0.12, 0.30]
X2	0.16	.10	1.53	.127	[−0.05, 0.36]
Racial disidentification	0.40	.03	12.14	<.001	[0.34, 0.47]
Political ideology	Conditional indirect effects of X1 (Whites Support Trump vs. Whites Oppose Trump condition)			Bootstrap 95% CI
Political ideology	Bootstrap indirect effect		Bootstrap SE	Bootstrap 95% CI
−1SD	.004		.05	[−.10, .10]
M	−.12		.04	[−.20, −.05]
+1SD	−.25		.06	[−.37, −.13]
Index of moderated mediation	−.06		.02	[−.10, −.02]
Political ideology	Conditional indirect effects of X2 (Whites Support Trump vs. Control condition)			Bootstrap 95% CI
Political ideology	Bootstrap indirect effect		Bootstrap SE	Bootstrap 95% CI
−1SD	−.07		.04	[−.16, .02]
M	−.13		.04	[−.20, −.06]
+1SD	−.18		.06	[−.30, −.07]
Index of moderated mediation	−.03		.02	[−.06, .01]
Predictor	Open-ended intentions to advocate for racial equity (dependent variable Model 2)				95% CI
Predictor	b	SE	t	P	95% CI
Intercept	1.20	.06	18.76	<.001	[1.07, 1.32]
X1	−0.10	.09	1.04	.297	[−0.28, 0.08]
X2	0.01	.09	0.06	.954	[−0.17, 0.18]
Racial disidentification	0.47	.03	16.62	<.001	[0.42, 0.53]
Political ideology	Conditional indirect effects of X1 (Whites Support Trump vs. Whites Oppose Trump condition)			Bootstrap 95% CI
Political ideology	Bootstrap indirect effect		Bootstrap SE	Bootstrap 95% CI
−1SD	.004		.06	[−.11, .12]
M	−.14		.05	[−.24, −.06]
+1SD	−.29		.07	[−.44, −.15]
Index of moderated mediation	−.07		.02	[−.12, −.03]
Political ideology	Conditional indirect effects of X2 (Whites Support Trump vs. Control condition)			Bootstrap 95% CI
Political ideology	Bootstrap indirect effect		Bootstrap SE	Bootstrap 95% CI
−1SD	−.08		.05	[−.19, .02]
M	−.15		.04	[−.24, −.07]
+1SD	−.22		.07	[−.36, −.08]
Index of moderated mediation	−.03		.02	[−.07, .01]
Predictor	Policy support (dependent variable Model 3)				95% CI
Predictor	b	SE	t	P	95% CI
Intercept	4.00	.08	47.36	<.001	[3.83, 4.16]
X1	0.25	.12	2.07	.039	[0.01, 0.49]
X2	0.11	.12	0.94	.248	[−0.12, 0.34]
Racial disidentification	0.68	.04	18.13	<.001	[0.61, 0.76]
Political ideology	Conditional indirect effects of X1 (Whites Support Trump vs. Whites Oppose Trump condition)			Bootstrap 95% CI
Political ideology	Bootstrap indirect effect		Bootstrap SE	Bootstrap 95% CI
−1SD	.01		.09	[−.16, .17]
M	−.21		.07	[−.34, −.08]
+1SD	−.42		.10	[−.62, −.22]
Index of moderated mediation	−.10		.03	[−.16, −.04]
Political ideology	Conditional indirect effects of X2 (Whites Support Trump vs. Control condition)			Bootstrap 95% CI
Political ideology	Bootstrap indirect effect		Bootstrap SE	Bootstrap 95% CI
−1SD	−.12		.07	[−.27, .03]
M	−.22		.06	[−.34, −.10]
+1SD	−.31		.10	[−.51, −.12]
Index of moderated mediation	−.05		.03	[−.10, .01]

Note. Whites Support Trump n = 283, Whites Oppose Trump n = 265, Control n = 301. Political ideology and racial disidentification were mean-centered; condition was dummy coded such that Whites support Trump condition was the reference category. Unstandardized coefficients are reported because standardized coefficients are not available for moderated mediation models in the PROCESS macro (Hayes, 2018). Bootstrap sample size = 5,000. SE = standard error; CI = confidence interval; SD = standard deviation.

Discussion

Replicating Studies 1a and 1b, Study 2 demonstrated that group-image threat (i.e., the majority of White Americans support Trump’s anti-egalitarianism) led liberal-leaning White Americans to racially disidentify. Study 2 also provided initial evidence that racial disidentification increased liberal-leaning White Americans’ intentions to advocate for racial equity and support policies benefiting racial minorities. However, it is unclear whether these increased egalitarian intentions translate into actions (i.e., the intention-behavior gap; Sheeran, 2002). In Study 3, we explore this question.

Revisiting the inconsistency in disidentification among conservative-leaning participants in Studies 1a and 1b, Study 2 results better align with Study 1a such that the association of White Americans with Trump had little to no effect on conservative-leaning White Americans’ racial identification. While these results point to a lack of group-image threat among conservative-leaning White Americans, we anticipate that this group does experience Trump-related group-image threat in other Trump-related domains. Conservatism is associated with greater acceptance of inequality (Jost et al., 2007), but does not center on anti-egalitarianism in the same way that liberalism does (Sidorsky, 2016). Thus, while conservative-leaning White Americans may not enjoy being associated with Trump’s anti-egalitarianism, the association may not affect their racial identification. In Study 3, we further explore the effect of the White American–Trump association among conservative-leaning White Americans by activating a different Trump-related negative ingroup portrayal than in Studies 1a to 2.

Study 3

Study 3 expanded on prior studies by (a) testing another trigger of Trump-related group-image threat—information that the majority of White Americans support versus oppose Trump’s continuation in office, following accusations of his abuse of power and obstruction of justice that resulted in Trump’s first impeachment trial (Fandos & Shear, 2019), (b) examining whether racial disidentification stemming from group-image threat generalizes to conservative-leaning White Americans when the threat arises from a tension between their racial identity and political ingroup’s power, and (c) examining whether racial disidentification increases liberal-leaning White Americans’ egalitarian behaviors (i.e., monetary donations to racial equity-focused organizations). Conceptually replicating previous studies, we hypothesized that information suggesting that White Americans support Trump’s continuation in office would increase racial disidentification and subsequently egalitarian behaviors among liberal-leaning White participants. We also hypothesized that, because Trump’s loss of power threatens their political ingroup’s power, conservative-leaning White participants would racially disidentify in response to information suggesting that White Americans oppose Trump’s continuation in office.

Method

Participants

We recruited 956 non-Hispanic, U.S. born, mono-racial White adults from MTurk, who had not participated in previous studies, to participate in exchange for US$1.50. Using the same preregistered exclusion criteria as Studies 1a to 2, we excluded 42 participants, giving a final sample of 914 (46% women; M_age = 40.65 years; 45.5% without a 4-year college degree). A sensitivity analysis indicated that this sample size can detect small effects (f² ≥ .014) with 80% power (Faul et al., 2009).

Procedure

We ran Study 3 during a 24-hr window immediately before the House of Representatives voted on Trump’s first impeachment. We used the same procedure and measures of political ideology (M = 4.95, SD = 2.22), racial disidentification (nine-item; α = .91, M = 3.20, SD = 1.26), and manipulation credibility (α = .89, M = 4.16, SD = 1.59) as Study 3 with the following exceptions: (a) we measured egalitarian behaviors (i.e., donations to organizations that promote racial equity) instead of intentions; and (b) the group-image threat manipulation was framed in terms of Trump’s impeachment (vs. anti-egalitarian beliefs).

As in Studies 1a and 2, participants in the Control condition proceeded straight to the measures after rating their political ideology, while participants in the Whites Support Trump and Whites Oppose Trump conditions viewed results from an ostensible survey. The Whites Support Trump condition portrayed White Americans as largely supportive of Trump’s continuation in office by disapproving of his impeachment. The survey description stated: “The survey revealed that the majority of White adults disapprove of the impeachment of President Trump.” The corresponding pie charts indicated that 30% to 40% of White respondents “supported the current impeachment inquiry of President Trump,” “[. . .] believed that the House of Representatives should impeach President Trump,” and “[. . .] believed that the Senate should convict and remove President Trump from office” (see Figure 4; see Supplemental File “Materials” for full stimuli).

In contrast, the Whites Oppose Trump condition portrayed White Americans as largely opposed to Trump’s continuation in office by approving of his impeachment. The survey description stated, “The survey revealed that the majority of White adults approve of the impeachment of President Trump.” The corresponding pie charts showed reversed percentages from the Whites Support Trump condition, indicating that 60% to 70% of White respondents supported Trump’s impeachment (see Figure 4).

Figure 4.

Example pie chart presented to participants in the Whites Support Trump condition (A) and the Whites Oppose Trump condition (B) in Study 3.

We decided upon this range of agreement based on public opinion regarding impeachment. At the time of Study 3, 40% of American adults in general, 15% of Republicans, and 85% of Democrats approved of Trump’s impeachment (FiveThirtyEight, 2020). However, there were no available polls regarding White Americans’ opinions specifically. Because White Americans are only slightly more likely to affiliate with the Republican (51%) versus Democratic Party (43%; Pew Research Center, 2018b), to ensure the credibility of our manipulation, we used less extreme values (i.e., 30%–40% impeachment approval rate in the Whites Support Trump condition and 60%–70% impeachment approval rate in the Whites Oppose Trump condition).

Measures

We measured egalitarian behaviors by offering participants an opportunity to donate a research bonus to nonprofit organizations focused on promoting racial equity (Sharps & Schroeder, 2019). Participants read,

Congratulations! You have received a bonus of $1.00. You can keep your entire bonus or share some or all of your bonus with one or more of the following non-profit organizations. These organizations all work within the field of racial equity, and on a variety of issues and topics. The donations will really go to the organization(s).

Then, participants read about seven organizations (e.g., the National Association for the Advancement of Colored People and the American Indian College Fund) and their mission statements. Participants then indicated whether they wished to keep their bonus or donate part or all of the bonus to one or more of the nonprofit organizations listed (randomized). We examined two indicators of behavioral advocacy for racial equity: (a) decision to donate part or all of the bonus versus keep the bonus (76.6% kept all bonus; see Table 4 for decision by condition and political ideology), and (b) total amount of donation regardless of organization(s) (M = US$0.16, SD = US$0.34).

Table 4.

Percentage of Participants Who Donated Some or All Bonus, by Condition and Political Ideology.

Political ideology	Control condition (%)	Whites Oppose Trump condition (%)	Whites Support Trump condition (%)
Extremely conservative	5.3	16.7	13.3
Conservative	23.8	7.0	25
Somewhat conservative	20.0	25.0	17.4
Slightly conservative	17.9	15.2	11.1
Slightly liberal	29.5	24.3	9.1
Somewhat liberal	33.3	20.0	44.0
Liberal	32.8	21.7	23.1
Extremely liberal	25.5	34.1	35.9

Results

Perceived credibility of manipulation

Ratings of the manipulation’s credibility (M = 4.16, SD = 1.59) were significantly above the scale midpoint (4), t(599) = 2.38, d = .10, p = .017, suggesting participants found the information relatively credible. Unlike previous studies, perceived credibility did not differ across conditions, Whites Support Trump: M = 4.23, SD = 1.54; Whites Oppose Trump: M = 4.07, SD = 1.65; t(598) = 1.17, d = .10, p = .241.

Does group-image threat increase racial disidentification among liberal-leaning and conservative-leaning White Americans?

We first examined: (a) whether exposure to information suggesting that White Americans support Trump’s continuation in office would increase disidentification among liberal-leaning White Americans, and (b) whether exposure to information suggesting that White Americans oppose Trump’s continuation in office would increase disidentification among conservative-leaning White participants. Following the analytic strategy outlined in Studies 1a and 1b, and consistent with prior studies, the interaction between political ideology and the Whites Oppose Trump versus White Support Trump conditions predicted racial disidentification (see Table 2 for model results; Figure 2). J-N analysis revealed that, as hypothesized, liberal-leaning White participants disidentified more in the Whites Support Trump condition than in the Whites Oppose Trump condition (region of significance: above 4.63 on the political ideology scale). We also found shifts in disidentification among conservative-leaning White Americans. As hypothesized, conservative-leaning White participants disidentified more in the Whites Oppose Trump condition than in the Whites Support Trump condition (region of significance: below 2.29).

Furthermore, as in Studies 1a and 2, there was a significant interaction between political ideology and the Whites Support Trump versus Control conditions. J-N analysis revealed that liberal-leaning and moderately conservative-leaning White participants disidentified more in the Whites Support Trump condition than in the Control condition (region of significance: above 3.66 on the political ideology scale).

Diverging from Studies 1a and 2, political ideology also significantly interacted with the Whites Oppose Trump versus Control conditions. J-N analysis revealed that conservative-leaning White participants disidentified more in the Whites Oppose Trump condition than in the Control condition (region of significance: below 3.71 on the political ideology scale).

Together these results replicate those of Studies 1a to 2: learning that the majority of White Americans support Trump’s continuation in office increased racial disidentification among liberal-leaning participants relative to learning that the majority oppose and relative to the control condition. These results also extend Study 1a to 2 findings regarding conservative-leaning White Americans by demonstrating that learning that the majority of White Americans oppose Trump’s continuation in office increased racial disidentification among conservative-leaning participants relative to learning that the majority support and relative to the control condition.

Does racial disidentification increase egalitarian behaviors among liberal-leaning White Americans?

To examine whether racial disidentification in response to Trump-related group-image threat would lead liberal-leaning White Americans to engage in egalitarian behaviors, we used the same analytic strategy as Study 2 (see Figure 3 for the diagram of the conceptual model). Replicating Study 2 findings, greater racial disidentification predicted greater egalitarian behavior in terms of both whether and how much individuals donated to equity-focused organizations (all ps < .001; see Table 5 for model results; also see Table S7 in the Supplemental file for similar results from a more saturated model). Moreover, the indirect effect of political ideology on donations through racial disidentification was moderated by condition. Consistent with our hypothesis, racial disidentification in response to information that the majority of White Americans support Trump’s continuation in office increased liberal-leaning (vs. conservative-leaning) White participants’ likelihood of donating and donation amounts to organizations that promote racial equity.

Table 5.

Regression Coefficients for Moderated Mediation Models Predicting Donation Decision (Model 1) and Amount of Donation (Model 2), With the Whites Support Trump Condition as the Reference Category.

Predictor	Racial disidentification (mediator variable model)				95% CI
	B	SE	t	p
Intercept	3.38	.07	51.69	<.001	[3.25, 3.50]
X1 (Whites Support Trump vs. Whites Oppose Trump)	−0.26	.09	2.75	.006	[−0.44, −0.07]
X2 (Whites Support Trump vs. Control)	−0.33	.09	3.55	<.001	[−0.51, −0.15]
Political ideology	0.31	.03	10.75	<.001	[0.26, 0.37]
X1 × Political ideology	−0.18	.04	4.38	<.001	[−0.27, −0.10]
X2 × Political ideology	−0.07	.04	1.56	.121	[−0.15, 0.02]
Predictor	Donation decision (dependent variable Model 1)				95% CI
	B	SE	z	p
Intercept	−2.69	.28	9.50	<.001	[−3.26, −2.14]
X1	−0.01	.21	0.03	.980	[−0.41, 0.40]
X2	0.34	.20	1.73	.084	[−0.05, 0.72]
Racial disidentification	0.42	.07	6.35	<.001	[0.29, 0.55]
Political ideology	Conditional indirect effects of X1 (Whites Support Trump vs. Whites Oppose Trump condition)			Bootstrap 95% CI
	Bootstrap indirect effect		Bootstrap SE
−1SD	.06		.06	[−.04, .18]
M	−.11		.04	[−.20, −.03]
+1SD	−.28		.07	[−.44, −.15]
Index of moderated mediation	−.08		.02	[−.13, −.04]
Political ideology	Conditional indirect effects of X2 (Whites Support Trump vs. Control condition)			Bootstrap 95% CI
	Bootstrap indirect effect		Bootstrap SE
−1SD	−.08		.06	[−.19, .03]
M	−.14		.04	[−.23, −.06]
+1SD	−.20		.07	[−.34, −.08]
Index of moderated mediation	−.03		.02	[−.07, .01]
Predictor	Amount of donation (dependent variable Model 2)				95% CI
	b	SE	t	P
Intercept	−.05	.03	1.29	.196	[−.11, .02]
X1	−.01	.03	0.27	.786	[−.06, .05]
X2	.02	.03	0.86	.393	[−.03, .08]
Racial disidentification	.06	.01	7.21	<.001	[.05, .08]
Political ideology	Conditional indirect effects of X1 (Whites Support Trump vs. Whites Oppose Trump condition)			Bootstrap 95% CI
	Bootstrap indirect effect		Bootstrap SE
−1SD	.01		.01	[−.01, .03]
M	−.02		.01	[−.03, −.005]
+1SD	−.04		.01	[−.06, −.02]
Index of moderated mediation	−.01		.003	[−.02, −.01]
Political ideology	Conditional indirect effects of X2 (Whites Support Trump vs. Control condition)			Bootstrap 95% CI
	Bootstrap indirect effect		Bootstrap SE
−1SD	−.01		.01	[−.03, .005]
M	−.02		.01	[−.03, −.01]
+1SD	−.03		.01	[−.05, −.01]
Index of moderated mediation	−.004		.003	[−.01, .001]

Note. Whites Support Trump n = 314, Whites Oppose Trump n = 293, Control n = 307. Political ideology and racial disidentification were mean-centered; condition was dummy coded such that Whites Support Trump condition was the reference category. Unstandardized coefficients are reported because standardized coefficients are not available for moderated mediation models in the PROCESS macro (Hayes, 2018). Log-odds coefficients are reported for the model predicting donation decision. Bootstrap sample size = 5,000. SE = standard error; CI = confidence interval; SD = standard deviation.

Discussion

Study 3 expanded our understanding of Trump-related group-image threat in three ways. First, among liberal-leaning White Americans, results demonstrated that Trump-related group-image threat extends beyond associations with Trump’s anti-egalitarianism: White Americans’ support for Trump’s continuation in office elicited racial disidentification among liberal-leaning White participants. Notably, although the White American–Trump association in Study 3 did not center on Trump’s anti-egalitarianism, the resultant racial disidentification still increased expressions of egalitarianism among liberal-leaning White participants as in Study 2. This suggests that even when Trump’s anti-egalitarianism is not explicitly activated, liberal-leaning White Americans view their racial group’s association with Trump as implicating their egalitarian beliefs.

Second, while Studies 1 to 2 did not find consistent support for group-image threat among conservative-leaning White Americans, Study 3 found that White Americans’ lack of support for Trump’s continuation in office elicited group-image threat among conservative-leaning White participants. One explanation for this finding is that White Americans’ opposition to Trump’s presidential power threatens the political power of the Republican party, with which many conservative-leaning White Americans affiliate (Saad, 2020). Conflict between their racial ingroup’s political preferences and their desire to maintain political power may create psychological tension that leads to racial disidentification. Further research is needed to understand whether this group reliably disidentifies from their racial group when racial and political identities conflict.

Finally, Study 3 demonstrated that, beyond expressing egalitarian intentions, liberal-leaning White Americans also engaged in more egalitarian behaviors (i.e., were more likely to donate and to donate larger amounts of money to racial equity-focused organizations) as a result of racial disidentification. These findings suggest that, although psychological threats are often associated with adverse outcomes, Trump-related group-image threat can galvanize liberal-leaning White Americans to act on behalf of racial equity, even at a personal cost. However, for conservative-leaning White Americans, racial disidentification stemming from group-image threat did not increase egalitarianism. It is possible that group-image threat elicits ideologically consistent behavior among conservative-leaning White Americans (e.g., increasing support for conservative political action), which our behavioral outcome in Study 3 does not capture. Future research should examine consequences of conservative-leaning White Americans’ racial disidentification.

General Discussion

Four studies demonstrated that group-image threat arising from the White American-Trump association, specifically White Americans’ perceived support for Trump’s anti-egalitarianism (Studies 1a–2) and his continuation in office (Study 3), increased racial disidentification among liberal-leaning White Americans. Furthermore, racial disidentification increased liberal-leaning White Americans’ egalitarian intentions (Study 2) and egalitarian behaviors (Study 3). Notably, the effect of group-image threat on racial disidentification among liberal-leaning White Americans persisted even after major political wins for liberals (e.g., Democrats winning control of the House, Trump’s associates pleading guilty to Mueller investigation charges, and the first Trump impeachment trial; see Figure 5), suggesting that liberal-leaning White Americans’ racial disidentification was not simply a response to a loss of political power, but instead stemmed from the persistent portrayal of their racial ingroup as being supportive of Trump. For conservative-leaning White Americans, however, their racial ingroup’s opposition to Trump’s political power, but not his anti-egalitarianism, increased racial disidentification. These results suggest that White Americans’ racial identification is responsive to the sociopolitical context during Trump’s presidency in which White Americans as a group are associated with Trump.

Figure 5.

Timeline of the major political events happening immediately before and during data collection for Studies 1 to 3.

Our findings are both consistent with and extend past literature. Specifically, research on White identity largely focuses on White Americans’ suppression of their racial identity in response to group image threat (see Knowles et al., 2014). Our research expands this literature by demonstrating that group-image threat also elicits disidentification among White Americans. This finding speaks to the generalizability of disidentification theorizing, which has largely focused on low-status groups (e.g., racially/ethnically minoritized groups and women; Branscombe et al., 2012), by demonstrating that this psychological process also occurs among a high-status racial group. In doing so, this research highlights the importance of studying White identity as a dynamic socially constructed identity rather than a static or presumably “normative” comparison identity (McDermott & Samson, 2005).

In addition, although we conceptualize the association between Trump’s anti-egalitarianism and White Americans as group-image threat, this association can also be considered through the lens of literature on stereotypes (i.e., consensually held social representations; Stangor & Lange, 1994). From this perspective, our research demonstrated that contemporary stereotypes of White Americans (i.e., as anti-egalitarian and supportive of Trump) produce both individual psychological tension and positive intergroup consequences. While stereotyping research often focuses on negative intergroup consequences, particularly for racially minoritized groups (e.g., prejudice, discrimination, and disparate police violence against Black Americans; Kahn & McMahon, 2015), our results suggest that when White Americans are stereotyped as anti-egalitarian, liberal-leaning White Americans compensate by working to promote racial equity. Future research may consider unpacking the extent to which this stereotype about White Americans is widely held, the long-term implications of this stereotype for White Americans’ sense of self, and how this stereotype shapes the ways in which people from other racial groups respond to White individuals.

Studies 2 and 3 also have implications for efforts to promote racial equity in the U.S. Past research offers mixed results regarding the effects of White racial disengagement; however, our findings consistently demonstrate that disengagement among liberal-leaning White Americans has positive implications for racial equity. While these findings may be specific to the sociopolitical context during Trump’s presidency, they might also help to identify parameters of the theorized effects of White racial disengagement. For example, intergroup consequences may depend upon the underlying motive of White racial disengagement. Indeed, Knowles and colleagues (2014) theorized that racial disengagement allows White Americans to deny the existence of White privilege, thereby satisfying their motivation to see their racial group’s high status as legitimate. This denial of privilege in turn predicted greater opposition to affirmative action (i.e., policies intended to create racial equality). In our research, however, liberal-leaning White Americans disengaged from their racial group to resolve the tension between anti-egalitarian portrayals of their racial ingroup and their own deeply held egalitarian beliefs. As a result, they expressed greater egalitarian intentions and engaged in more egalitarian behaviors. Rather than viewing our findings as oppositional to previous research demonstrating negative intergroup consequences of White disengagement, we suggest that motivational and contextual precursors may explain why White disengagement can have both positive and negative intergroup consequences.

Limitations

Although our studies demonstrate a consistent relation between Trump-related group-image threat, racial disidentification, and egalitarian outcomes among liberal-leaning White Americans, we report our findings with several limitations in mind. First, we used a single-item conservative-liberal scale to assess political ideology. Research shows that the single-item political ideology measure is a strong predictor of social views, including preference for social equality (Kroh, 2007; Rockey, 2014). However, political ideology is multidimensional (Duckitt et al., 2002), and individuals may simultaneously hold liberal and conservative views regarding different issues (Jost, 2006). Research using a multidimensional measure of political ideology may further illuminate the nuances and parameters of relations between group-image threat, White identity, and intergroup beliefs and behaviors.

In addition, despite our consistent findings that racial disidentification among liberal-leaning White Americans predicted both greater expressions of egalitarian intentions and engagement in egalitarian behavior, two key questions remain. First, the majority (76.6%) of participants kept their entire bonus instead of donating. One possibility is that financial strain influenced participants’ donation decisions. Although we did not assess participants’ financial status, almost half of the MTurk workers in our samples had not earned a bachelor’s degree, meaning that they were likely to work lower-paying jobs (U.S. Census Bureau, 2012) and perhaps supplement their income through participating in MTurk studies. Another possibility is that participants viewed the bonus amount (US$1) as too small to be impactful, which deterred them from donating (see Duncan, 2004). Finally, our measure of egalitarian behavior was limited to monetary donations. However, egalitarianism can manifest in other ways, such as volunteering or attending protests. These actions can be more personally costly than donating money, but they also affirm individuals’ moral identities more so than donating money (Reed et al., 2016). Further research should examine whether the positive intergroup effects of White racial disengagement extend to other egalitarian behaviors.

Constraints on Generality

While this work adds to a growing body of literature examining White racial identity, the use of MTurk participants and the temporal specificity of this research may limit the extent to which our findings generalize to other high-status groups or sociopolitical contexts (Simons et al., 2017). First, compared with the general American population, MTurk workers are younger, more liberal, and more highly educated, but make less income (Buhrmester et al., 2011; Levay et al., 2016). Further research is needed to determine whether the model generalizes to a sample that reflects the national population. Second, because we designed the stimuli to reflect a particular political moment (i.e., Trump’s presidency), the effects of our manipulations on racial disidentification are likely to change over time as the portrayals of White Americans change with the sociopolitical climate. We have no reason to believe that the results depend on other characteristics of the participants, materials, or context.

Conclusion

In the political climate during Trump’s presidency, where White Americans are credited with electing and continuing to empower a president who consistently attacks, dehumanizes, and devalues marginalized groups (Conway et al., 2017), liberal-leaning Whites are forced to navigate the psychological tension between their racial identity and egalitarian beliefs. Our findings suggest that racial disidentification provides one means of alleviating this tension by motivating White Americans to reaffirm their egalitarian beliefs. However, racial disidentification is not simply a matter of individual psychological protection for liberal-leaning Whites; it is a matter of grappling with one’s role in a system that discriminates against marginalized groups. The question moving forward is how to leverage this psychological process to encourage liberal-leaning Whites to ally themselves with communities fighting to build a more inclusive, equitable society.

Supplemental Material

sj-docx-1-psp-10.1177_0146167220987988 – Supplemental material for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity

Supplemental material, sj-docx-1-psp-10.1177_0146167220987988 for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity by J. Doris Dai, Arianne E. Eason, Laura M. Brady and Stephanie A. Fryberg in Personality and Social Psychology Bulletin

Supplemental Material

sj-docx-2-psp-10.1177_0146167220987988 – Supplemental material for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity

Supplemental material, sj-docx-2-psp-10.1177_0146167220987988 for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity by J. Doris Dai, Arianne E. Eason, Laura M. Brady and Stephanie A. Fryberg in Personality and Social Psychology Bulletin

Supplemental Material

sj-docx-3-psp-10.1177_0146167220987988 – Supplemental material for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity

Supplemental material, sj-docx-3-psp-10.1177_0146167220987988 for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity by J. Doris Dai, Arianne E. Eason, Laura M. Brady and Stephanie A. Fryberg in Personality and Social Psychology Bulletin

Supplemental Material

sj-docx-4-psp-10.1177_0146167220987988 – Supplemental material for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity

Supplemental material, sj-docx-4-psp-10.1177_0146167220987988 for #NotAllWhites: Liberal-Leaning White Americans Racially Disidentify and Increase Support for Racial Equity by J. Doris Dai, Arianne E. Eason, Laura M. Brady and Stephanie A. Fryberg in Personality and Social Psychology Bulletin

Footnotes

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: A Ford Foundation Predoctoral Fellowship and a National Institute on Aging (NIA) Aging and Development Training Grant (T32 AG0000030) supported A.E.E.’s contributions to this manuscript.

ORCID iD

J. Doris Dai

Open Practices Statement

Study 1a was not preregistered. The preregistration for Study 1b can be accessed at https://osf.io/2gbfk/?view_only=75b4841612ae4f7b8f9f07fad5bb2cd7; Study 2: https://osf.io/wep3d/?view_only=05419b536bba403c8a4daaad2120dc4b; Study 3: https://osf.io/pn2em/?view_only=d855813b06c8455da748793c7af66751. De-identified data and SPSS syntax are posted at . Access to data is limited to qualified researchers.

Supplemental Material

Supplemental material is available online with this article.

Notes

References

Andreychik

M. R.

Gill

M. J.

(2009). Ingroup identity moderates the impact of social explanations on intergroup attitudes: External explanations are not inherently prosocial. Personality and Social Psychology Bulletin, 35(12), 1632–1645. https://doi.org/10.1177/0146167209345285

Arndt

Greenberg

Schimel

Pyszczynski

Solomon

(2002). To belong or not to belong, that is the question: Terror management and identification with gender and ethnicity. Journal of Personality and Social Psychology, 83(1), 26–43. https://doi.org/10.1037/0022-3514.83.1.26

Barrie

R. E.

Langrehr

Jerémie-Brink

Alder

Hewitt

Thomas

(2016). Stereotypical beliefs and psychological well-being of African American adolescent girls: Collective self-esteem as a moderator. Counselling Psychology Quarterly, 29(4), 423–442. https://doi.org/10.1080/09515070.2015.1129494

Becker

J. C.

Tausch

(2014). When group memberships are negative: The concept, measurement, and behavioral implications of psychological disidentification. Self and Identity, 13(3), 294–321. https://doi.org/10.1080/15298868.2013.819991

Branscombe

N. R.

Fernández

Gómez

Cronin

(2012). Moving toward or away from a group identity: Different strategies for coping with pervasive discrimination. In Jetten

Haslam

S. A.

(Eds.), The social cure: Identity, health and well-being (pp. 115–131). Psychology Press.

Branscombe

N. R.

Schmitt

M. T.

Schiffhauer

(2007). Racial attitudes in response to thoughts of White privilege. European Journal of Social Psychology, 37(2), 203–215. https://doi.org/10.1002/ejsp.348

Buhrmester

Kwang

Gosling

S. D.

(2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6(1), 3–5. https://doi.org/10.1177/1745691610393980

Chow

R. M.

Lowery

B. S.

Knowles

E. D.

(2008). The two faces of dominance: The differential effect of ingroup superiority and outgroup inferiority on dominant-group identity and group esteem. Journal of Experimental Social Psychology, 44(4), 1073–1081. https://doi.org/10.1016/j.jesp.2007.11.002

Conway

I. I. I.

Repke

M. A.

Houck

S. C.

(2017). Donald Trump as a cultural revolt against perceived communication restriction: Priming political correctness norms causes more Trump support. Journal of Social and Political Psychology, 5(1), 244–259. https://doi.org/10.5964/jspp.v5i1.732

10.

Crocker

Luhtanen

Blaine

Broadnax

(1994). Collective self-esteem and psychological well-being among White, Black, and Asian college students. Personality and Social Psychology Bulletin, 20(5), 503–513. https://doi.org/10.1177/0146167294205007

11.

Doosje

Branscombe

N. R.

Spears

Manstead

A. S. R.

(1998). Guilty by association: When one’s group has a negative history. Journal of Personality and Social Psychology, 75(4), 872–886. https://doi.org/10.1037/0022-3514.75.4.872

12.

Duckitt

Wagner

Du Plessis

Birum

(2002). The psychological bases of ideology and prejudice: Testing a dual process model. Journal of Personality and Social Psychology, 83(1), 75–93. https://doi.org/10.1037/0022-3514.83.1.75

13.

Duncan

(2004). A theory of impact philanthropy. Journal of Public Economics, 88(9–10), 2159–2180. https://doi.org/10.1016/S0047-2727(03)00037-9

14.

Fandos

Shear

M. D.

(2019, December 18). Trump impeached for abuse of power and obstruction of congress. The New York Times. https://www.nytimes.com/2019/12/18/us/politics/trump-impeached.html

15.

Faul

Erdfelder

Buchner

Lang

A.-G.

(2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149–1160. https://doi.org/10.3758/BRM.41.4.1149

16.

FiveThirtyEight. (n.d.). How (un)popular is Trump? https://projects.fivethirtyeight.com/trump-approval-ratings/

17.

FiveThirtyEight. (2020, February 12). Did Americans support removing Trump from office? https://projects.fivethirtyeight.com/impeachment-polls/

18.

Gallup. (2018, June 25). Trump job approval slips back to 41%. https://news.gallup.com/poll/235955/trump-job-approval-slips-back.aspx

19.

Gallup. (2019, March 29). Subgroup differences in Trump approval mostly party-based. https://news.gallup.com/poll/248135/subgroup-differences-trump-approval-mostly-party-based.aspx

20.

Goren

M. J.

Plaut

V. C.

(2012). Identity form matters: White racial identity and attitudes toward diversity. Self and Identity, 11(2), 237–254. https://doi-org.offcampus.lib.washington.edu/10.1080/15298868.2011.556804

21.

Harmon

Tavernise

(2020, June 12). One big difference about George Floyd protests: Many white faces. The New York Times. https://www.nytimes.com/2020/06/12/us/george-floyd-white-protesters.html

22.

Hayes

A. F.

(2018). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (2nd ed.). Guilford Press.

23.

Hayes

A. F.

Montoya

A. K.

(2017). A tutorial on testing, visualizing, and probing an interaction involving a multicategorical variable in linear regression analysis. Communication Methods and Measurements, 11(1), 1–30. https://doi.org/10.1080/19312458.2016.1271116

24.

Hodson

Esses

V. M.

(2002). Distancing oneself from negative attributes and the personal/group discrimination discrepancy. Journal of Experimental Social Psychology, 38(5), 500–507. https://doi.org/10.1016/S0022-1031(02)00012-4

25.

Hogg

M. A.

(2003). Social identity. In Leary

M. R.

Tangney

J. P.

(Eds.), Handbook of self and identity (pp. 462–479). Guilford Press.

26.

Iyer

Leach

C. W.

Crosby

F. J.

(2003). White guilt and racial compensation: The benefits and limits of self-focus. Personality and Social Psychology Bulletin, 29(1), 117–129. https://doi.org/10.1177/0146167202238377

27.

Iyer

Schmader

Lickel

(2007). Why individuals protest the perceived transgressions of their country: The role of anger, shame, and guilt. Personality and Social Psychology Bulletin, 33(4), 572–587. https://doi.org/10.1177/0146167206297402

28.

Jacobson

G. C.

(2016). Polarization, gridlock, and presidential campaign politics in 2016. The ANNALS of the American Academy of Political and Social Science, 667(1), 226–246. https://doi.org/10.1177/0002716216658921

29.

Jardina

(2019). White identity politics. Cambridge University Press.

30.

Jardina

(2020). In-group love and out-group hate: White racial attitudes in contemporary U.S. elections. Political Behavior. https://doi.org/10.1007/s11109-020-09600-x

31.

Jasinskaja-Lahti

Liebkind

Solheim

(2009). To identify or not to identify? National disidentification as an alternative reaction to perceived ethnic discrimination. Applied Psychology: An International Review, 58(1), 105–128. https://doi.org/10.1111/j.1464-0597.2008.00384.x

32.

Johnson

P. O.

Neyman

(1936). Tests of certain linear hypotheses and their application to some educational problems. Statistical Research Memoirs, 1, 57–93.

33.

Jost

J. T.

(2006). The end of the end of ideology. American Psychologist, 61(7), 651–670. https://doi.org/10.1037/0003-066X.61.7.651

34.

Jost

J. T.

Napier

J. L.

Thorisdottir

Gosling

S. D.

Palfai

T. P.

Ostafin

(2007). Are needs to manage uncertainty and threat associated with political conservatism or ideological extremity? Personality and Social Psychology Bulletin, 33(7), 989–1007. https://doi.org/10.1177/0146167207301028

35.

Kahn

K. B.

McMahon

J. M.

(2015). Shooting deaths of unarmed racial minorities: Understanding the role of racial stereotypes on decisions to shoot. Translational Issues in Psychological Science, 1(4), 310–320. https://doi.org/10.1037/tps0000047

36.

Kiley

(2017, October 23). In polarized era, fewer Americans hold a mix of conservative and liberal views. https://www.pewresearch.org/fact-tank/2017/10/23/in-polarized-era-fewer-americans-hold-a-mix-of-conservative-and-liberal-views/

37.

Knowles

E. D.

Lowery

B. S.

Chow

R. M.

Unzueta

M. M.

(2014). Deny, distance, or dismantle? How white Americans manage a privileged identity. Perspectives on Psychological Science, 9(6), 594–609. https://doi.org/10.1177/174569161455465

38.

Knowles

E. D.

Peng

(2005). White selves: Conceptualizing and measuring a dominant-group identity. Journal of Personality and Social Psychology, 89(2), 223–241. https://doi.org/10.1037/0022-3514.89.2.223

39.

Kroh

(2007). Measuring left–right political orientation: The choice of response format. Public Opinion Quarterly, 71(2), 204–220. https://doi.org/10.1093/poq/nfm009

40.

Levay

K. E.

Freese

Druckman

J. N.

(2016). The demographic and political composition of Mechanical Turk samples. SAGE Open, 6(1), 215824401663643. https://doi.org/10.1177/2158244016636433

41.

Levin

Sidanius

Rabinowitz

J. L.

Federico

(1998). Ethnic identity, legitimizing ideologies, and social status: A matter of ideological asymmetry. Political Psychology, 19(2), 373–404. https://doi.org/10.1111/0162-895X.00109

42.

Linnehan

Chrobot-Mason

Konrad

A. M.

(2006). Diversity attitudes and norms: The role of ethnic identity and relational demography. Journal of Organizational Behavior, 27(4), 419–442. https://doi.org/10.1002/job.382

43.

Marshburn

C. K.

Knowles

E. D.

(2018). White out of mind: Identity suppression as a coping strategy among Whites anticipating racially charged interactions. Group Processes & Intergroup Relations, 21(6), 874–892. https://doi.org/10.1177/1368430216681178

44.

Matschke

Fehr

(2017). Does identity incompatibility lead to disidentification? Internal motivation to be a group member acts as buffer for sojourners from independent cultures. Frontiers in Psychology, 8, Article 335. https://doi.org/10.3389/fpsyg.2017.00335

45.

McDermott

Samson

F. L.

(2005). White racial and ethnic identity in the United States. Annual Review of Sociology, 31, 245–261. https://doi.org/10.1146/annurev.soc.31.041304.122322

46.

Mussweiler

Gabriel

Bodenhausen

G. V.

(2000). Shifting social identities as a strategy for deflecting threatening social comparisons. Journal of Personality and Social Psychology, 79(3), 398–409. https://doi.org/10.1037/0022-3514.79.3.398

47.

Parker

Nakamura

(2017, August 15). Trump again blames “both sides” for violence at white supremacist rally in Charlottesville. The Washington Post. https://www.washingtonpost.com/politics/trump-defends-organizers-of-white-supremacist-rally-in-charlottesville/2017/08/15/de01ff66-81f9-11e7-902a-2a9f2d808496_story.html

48.

Pew Research Center. (2018a, August 9). An examination of the 2016 electorate, based on validated voters. https://www.people-press.org/2018/08/09/an-examination-of-the-2016-electorate-based-on-validated-voters/

49.

Pew Research Center. (2018b, March 20). Wide gender gap, growing educational divide in voters’ party identification. https://www.people-press.org/2018/03/20/wide-gender-gap-growing-educational-divide-in-voters-party-identification/

50.

Pew Research Center. (2020, June 24). Recent protest attendees are more racially and ethnically diverse, younger than Americans overall. https://www.pewresearch.org/fact-tank/2020/06/24/recent-protest-attendees-are-more-racially-and-ethnically-diverse-younger-than-americans-overall/

51.

Pieterse

A. L.

Utsey

S. O.

Miller

M. J.

(2016). Development and initial validation of the anti-racism behavioral inventory (ARBI). Counselling Psychology Quarterly, 29(4), 356–381. https://doi.org/10.1080/09515070.2015.1101534

52.

Powell

A. A.

Branscombe

N. R.

Schmitt

M. T.

(2005). Inequality as ingroup privilege or outgroup disadvantage: The impact of group focus on collective guilt and interracial attitudes. Personality and Social Psychology Bulletin, 31(4), 508–521. https://doi.org/10.1177/0146167204271713

53.

Rabinovich

Morton

T. A.

(2010). Who says we are bad people? The impact of criticism source and attributional content on responses to group-based criticism. Personality and Social Psychology Bulletin, 36(4), 524–536. https://doi.org/10.1177/0146167210362980

54.

Rabinovich

Morton

T. A.

Landon

Neill

Mason

B. S.

Burdett

(2014). The password is praise: Content of feedback affects categorization of feedback sources. British Journal of Social Psychology, 53(3), 484–500. https://doi.org/10.1111/bjso.12041

55.

Reed

I. I.

Kay

Finnel

Aquino

Levy

(2016). I don’t want the money, I just want your time: How moral identity overcomes the aversion to giving time to prosocial causes. Journal of Personality and Social Psychology, 110(3), 435–457. https://doi.org/10.1037/pspp0000058

56.

Rockey

(2014). Who is Left-Wing, and Who Just Thinks They Are? University of Leicester, UK. https://www.le.ac.uk/economics/research/RePEc/lec/leecon/dp09-23.pdf

57.

Rosnow

R. L.

Rosenthal

(1989). Statistical procedures and the justification of knowledge in psychological science. American Psychologist, 44(10), 1276–1284. https://doi.org/10.1037/0003-066X.44.10.1276

58.

Saad

(2020, January 9). The U.S. remained center-right, ideologically, in 2019. Gallup. https://news.gallup.com/poll/275792/remained-center-right-ideologically-2019.aspx

59.

Sedikides

Gregg

(2003). Portraits of the self. In Hogg

M. A.

Cooper

(Eds.), Sage handbook of social psychology (pp. 110–138). SAGE.

60.

Sharps

D. L.

Schroeder

(2019). The preference for distributed helping. Journal of Personality and Social Psychology, 117(5), 954–977. https://doi.org/10.1037/pspi0000179

61.

Sheeran

(2002). Intention–behavior relations: A conceptual and empirical review. European Review of Social Psychology, 12(1), 1–36. https://doi.org/10.1002/0470013478.ch1

62.

Shuman

Johnson

Saguy

Halperin

(2018). Threat to the group’s image can motivate high identifiers to take action against in-group transgressions. Personality and Social Psychology Bulletin, 44(11), 1523–1544. https://doi.org/10.1177/0146167218768800

63.

Sidanius

Van Laar

Levin

Sinclair

(2004). Ethnic enclaves and the dynamics of social identity on the college campus: The good, the bad, and the ugly. Journal of Personality and Social Psychology, 87(1), 96–110. https://doi.org/10.1037/0022-3514.87.1.96

64.

Sidorsky

(2016). Chapter 2: An interpretation of American conservative thought: Political issues, conceptual differences, and attitudinal disjunctions. In Levinson

Williams

Parker

(Eds.), American conservatism (pp. 55–139). New York University Press.

65.

Simons

D. J.

Shoda

Lindsay

D. S.

(2017). Constraints on generality (COG): A proposed addition to all empirical papers. Perspectives on Psychological Science, 12(6), 1123–1128. https://doi.org/10.1177/1745691617708630

66.

Southern Poverty Law Center. (2020, June 17). Family separation under the Trump administration—A timeline. https://www.splcenter.org/news/2020/06/17/family-separation-under-trump-administration-timeline

67.

Stangor

Lange

J. E.

(1994). Mental representations of social groups: Advances in understanding stereotypes and stereotyping. In Zanna

M. P.

(Ed.), Advances in experimental social psychology (Vol. 26, pp. 357–416). Academic Press. https://doi.org/10.1016/S0065-2601(08)60157-4

68.

Swim

J. K.

Miller

D. L.

(1999). White guilt: Its antecedents and consequences for attitudes toward affirmative action. Personality and Social Psychology Bulletin, 25(4), 500–514. https://doi.org/10.1177/0146167299025004008

69.

Tajfel

Turner

(1979). An integrative theory of intergroup conflict. In Austin

W. G.

Worchel

(Eds.), The social psychology of intergroup relations (pp. 33–48). Brooks/Cole.

70.

Tajfel

Turner

J. C.

(1986). The social identity theory of intergroup behavior. In Worchel

Austin

(Eds.), Psychology of intergroup relations (pp. 7–24). Nelson Hall.

71.

Turner

J. C.

Hogg

M. A.

Oakes

P. J.

Reicher

S. D.

Wetherell

M. S.

(1987). Rediscovering the social group. Wiley-Blackwell.

72.

U.S. Census Bureau. (2012). Statistical abstract of the United States: 2012. https://www.census.gov/library/publications/2011/compendia/statab/131ed.html

73.

Ward

L. M.

(2004). Wading through the stereotypes: Positive and negative associations between media use and black adolescents’ conceptions of self. Developmental Psychology, 40(2), 284–294. https://doi.org/10.1037/0012-1649.40.2.284

74.

Wong

Cho

G. E.

(2005). Two-headed coins or Kandinskys: White racial identification. Political Psychology, 26(5), 699–720. https://doi.org/10.1111/j.1467-9221.2005.00440.x

75.

Yip

(2016). To be or not to be: How ethnic/racial stereotypes influence ethnic/racial disidentification and psychological mood. Cultural Diversity and Ethnic Minority Psychology, 22(1), 38–46. https://doi.org/10.1037/cdp0000046

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB

0.04 MB

0.06 MB

0.12 MB