White by Another Name? Can Anti-Christian Bias Claims Serve as a Racial Dog Whistle?

Abstract

Four preregistered experiments (N = 4,307) explored whether anti-Christian bias claims can discreetly signal White allyship among Christian American adults. In Experiments 1 and 2, reading about anti-Christian bias led White, but not Black, Christians to perceive more anti-White bias. Experiments 3 and 4 demonstrate the connection between Christian and White can be leveraged by politicians in the form of a racial dog whistle. In Experiment 3, White Christians perceived a politician concerned about anti-Christian bias as caring more about anti-White bias and more willing to fight for White people (relative to a control). This politician was also perceived as less offensive than a politician concerned about anti-White bias. In Experiment 4, Black Christians perceived a politician concerned about anti-Christian bias as less offensive than one concerned about anti-White bias yet still unlikely to fight for Black people. Results suggest “anti-Christian bias” can provide a relatively palatable way to signal allegiance to White people.

Keywords

racial dog whistle anti-White bias anti-Christian bias open data open materials preregistered

Former U.S. President Donald Trump recently declared, “Christians and Americans of faith are being persecuted like nothing this nation has ever seen before” (Fichera, 2024). These remarks reflect growing discourse about anti-Christian discrimination in the United States (U.S.). Republicans report that Christians face more discrimination than Black people (Public Religion Research Institute [PRRI], 2021a), U.S. Christians increasingly perceive bias against their in-group (Wilkins et al., 2022), and young White evangelical Christians report that Christians experience more discrimination than minority groups (Vandermaas-Peeler et al., 2018). We contend that anti-Christian bias claims by elites signal more than mere concern about religious discrimination; they may serve as effective racial dog whistles (i.e., coded language that indirectly signals race; Haney-López, 2014) by providing a relatively socially acceptable way to signal White allyship without explicitly referencing race. We tested whether anti-Christian bias claims can be effectively leveraged in this manner.

Motivation for Racial Dog Whistles

Although we do not aim to address whether people (and if so, who chooses to) intentionally use claims of anti-Christian bias as a dog whistle and focus instead on the “ripeness” of such claims for dog whistling, our research is grounded in the recognition that political leaders—and constituents—who wish to signal White allyship are likely motivated to use race-neutral language. This is because individuals who openly support White causes are widely condemned as White supremacists. For example, anti-White bias claims are perceived unfavorably (Wilkins et al., 2013), and claimants who openly express concerns about anti-White discrimination are viewed as racist (Blodorn & O’Brien, 2013). Explicitly racist speech is also censored (Bhat & Klen, 2020), making it ineffective.

One way to avoid censorship is to discreetly reference race, allowing individuals to capitalize on racist attitudes while minimizing social costs. To do this, politicians often use racial dog whistles, which use coded language (e.g., “welfare queens,” “inner city”) to indirectly evoke race, activate racial resentment, and garner support from a specific group while minimizing opposition (Haney-López, 2014; Mendelberg, 2001). Dog whistles may also work as in-group code to signal “people like us.” Given that terms such as “American” imply “White” (Devos & Banaji, 2005), Sarah Palin’s references to “real America” on the campaign trail were understood as racially coded appeals to her disproportionately White audiences (Silver, 2008). They can even be symbolic, such as Donald Trump silently holding a Bible at the height of the George Floyd protests, conveying both cultural leadership and White racial solidarity (Perry, 2023). In contrast to basic racial priming, dog whistles cloak the racial nature of the appeals for those who might otherwise be put off by overt racial content, making them broadly effective because of their seeming race neutrality (Haney-López, 2014; Valentino et al., 2002; White, 2007), even across the political spectrum (Wetts & Willer, 2019; White, 2007).

Anti-Christian Bias Claims May Serve as Racial Dog Whistles

There are many reasons to believe that anti-Christian bias claims could signal concern for White Americans. For example, there is significant numerical overlap between White people and Christians in the U.S.; 75.8% of Americans are White (U.S. Census Bureau, 2021), 71% of White Americans are Christian (PRRI, 2021b), and about 63% of Christians are White (intermediate values used to calculate estimates were obtained from PRRI, 2021b). Given this large numerical overlap, people may conflate Christian with White. Further, White people are perceived as more American than racial minorities (Devos & Banaji, 2005; Zou & Cheryan, 2017), and Christianity is associated with perceived Americanness (Butz & Carvalho, 2015; Whitehead & Scheitle, 2018). Thus, Americanness is associated with White and Christian, perhaps contributing to their conflation.

Statement of Relevance

Social scientists are increasingly sounding alarm bells about White Christian nationalism—an ethnoracial ideology that links Christian and White nationalism in the U.S. Simultaneously, some politicians openly proclaim pro-Christian views and claim that Christians are victimized. We provide novel evidence that such rhetoric can serve as a racial dog whistle for White Christian American adults, who interpret anti-Christian bias claims as implicating anti-White bias. Further, White Christians believe politicians concerned about anti-Christian bias are concerned about White, but not Black, people. We demonstrate that this dog-whistle tactic may be effective because both White and Black Christians perceive it as more palatable than overtly racial language. Results have important implications for psychologists, sociologists, and political scientists interested in racial and religious cognition, coalition building, social movements, and political discourse, as well as for the media, political leaders, and everyday consumers of political language.

Additionally, political forces that center Whiteness and Christianity within American life work in tandem. There is a long history of White supremacy within conservative American Christianity (Butler, 2021; Jones, 2016). Among White Americans, support for explicitly Christian nationalist views, even without racial references, strongly predicts racial prejudice and perceptions of anti-White discrimination (Perry, 2022, 2023). It is unsurprising, then, that large portions of the American public are threatened by rising racial and religious diversity (Al-Kire et al., 2022; Craig et al., 2018; Jones, 2016) and perceive that White people and Christians face a great deal of discrimination (PRRI, 2021a). Given numerical and contextual links between Whiteness and Christianity, concerns about anti-Christian bias may communicate anti-White bias.

Anti-Christian Bias Claims May Be Effective Racial Dog Whistles

A core feature of racial dog whistles is their ability to signal race while minimizing social sanction (Haney-López, 2014). Anti-Christian bias claims may meet this criterion for two reasons. First, despite the link between Whiteness and Christianity in the U.S., Christianity is a racially diverse religion. For example, a large majority of Black (72%; PRRI, 2021b) and Latinx (77%) Americans are Christian (Pew Research Center, 2014), suggesting pro-Christian attitudes can be cast as race-neutral.

Second, Americans hold contrasting views about the societal roles of race and religion, which makes expressing pro-Christian attitudes relatively more socially acceptable. For example, in the U.S., White people react negatively to discussions about racial discrimination (Reny et al., 2020) and prefer color blindness (Apfelbaum et al., 2012). Talking directly about racial bias incurs social penalties (Kaiser & Miller, 2001; Mendelberg, 2008; Nteta et al., 2016; Reny et al., 2020). In contrast, reactions to religious-discrimination claims may be less contentious. The U.S. was founded with strong religious protections, and Americans agree that Christianity, in particular, should be protected (Pew Research Center, 2021). Consequently, anti-Christian bias claims may not evoke the same penalties as explicit race references (e.g., anti-White bias claims; Perry, 2023; Wilkins et al., 2013). Instead, defending Christians may be seen as upholding a democratic ideal (religious freedom) shared by racially diverse Americans (Fox, 2021).

Current Research

Four preregistered experiments tested whether Christian Americans perceive a connection between anti-Christian and anti-White bias (Experiments 1 and 2) and whether this connection could be leveraged as a racial dog whistle whereby anti-Christian bias concerns signal concerns about anti-White bias (Experiments 3 and 4). Experiment 1 tested whether reading about anti-Christian bias evoked greater perceptions of anti-White bias among White Christians. Experiment 2 replicated Experiment 1 with White and Black Christian American samples and tested whether White Christians uniquely perceive anti-Christian bias to imply anti-White bias. Experiment 3 tested whether politicians can effectively signal White allyship to White Christians by discussing concerns about anti-Christian bias and/or religious freedom. Finally, Experiment 4 tested whether Black Christians are sensitive to anti-Christian bias claims as a racial dog whistle.

Open Practices Statement

Preregistrations, data, code, materials, and manipulations for all studies are publicly accessible on the OSF at https://osf.io/qbdyh.

Experiments 1 and 2

Our core question was whether reading about anti-Christian bias would evoke greater perceptions of anti-White bias (relative to a control). Secondarily, we tested whether reading about anti-White bias would increase perceptions of anti-Christian bias and whether effects were unique to the Christian–White association, and not Christian–Black. We hypothesized White Christians would perceive greater anti-White bias when reading about anti-Christian bias and made no predictions about Black Christians.

Method

Participants

Experiment 1

We recruited 900 adult participants from Prime Panels (Chandler et al., 2019). We removed 119 participants who did not report being White Christians living in the U.S. We then removed 238 participants who either failed a manipulation comprehension check (e.g., “Which of the following did the article NOT discuss?”) or an attention check (“Please select 5 to prove you’re paying attention”). Below we report data for 543 White Christians living in the U.S. (58.57% women, 41.07% men; M_age = 57.55 years, SD_age = 17.38 years).

Monte Carlo simulations of data (N = 10,000) with the observed sample size and standardized mean differences (pairwise mean differences divided by pooled standard deviation, i.e., the parameter that Cohen’s d estimates) of 0.34 achieved 80% power (0.31 for 70% power and 0.41 for 90% power). See analysis code for simulation details.

Experiment 2

We recruited 3,040 White and Black Christian adults living in the U.S. and who passed an attention check from Qualtrics. We removed 445 participants who failed at least one of two manipulation comprehension checks. Below we report data for 1,344 White Christians living in the U.S. (62.28% women, 37.65% men; M_age = 60.51 years, SD_age = 13.87 years) and 1,251 Black Christians living in the U.S. (71.36% women, 27.84% men; M_age = 47.96 years, SD_age= 16.07 years).

Monte Carlo simulations (N = 10,000) of data with the observed sample size and standardized mean differences of 0.22 for White participants and 0.23 for Black participants achieved 80% power (0.20 and 0.20 for 70% power and 0.25 and 0.26 for 90% power, respectively).

Procedure and materials

These studies were administered online via a Qualtrics survey. After consenting, participants were randomly assigned to either a null control condition or to one of three experimental conditions in which they read a brief article about bias against either Christians, White people, or Black people in the U.S. All articles were structurally similar and provided several examples of bias toward the relevant group. For the articles used in Experiment 2, see Table 1 (articles used in Experiment 1 were comparable). For full materials and manipulations, including additional measures that are not reported in this article that we collected to address separate research questions (e.g., how perceptions of bias are associated with policy attitudes), see https://osf.io/2ey3p. The procedure adhered to the ethical guidelines set by the authors’ institutional review board.

Table 1.

Experiment 2 Manipulations

Condition	Manipulation
Anti-Christian bias	Christians Face Bias in the U.S. According to recent reports, the majority of Christians in the U.S. believe they are facing more discrimination than ever before. The majority of Christians perceive that some laws have impeded their rights and opened them up to discriminatory treatment. These perceptions may not be unfounded. American Christians today routinely report facing prejudice in the workplace for expressing their views. Some Christians have been fined or jailed for refusing to comply with laws implemented in the last decade that violate their moral values. Schools all over the nation have removed religious references from classrooms and are preventing students from praying there. Christians are also heavily underrepresented in well-paying scientific academic fields, and many think bias is the likely cause. While the U.S. was founded on the premise of religious freedom and tolerance, many Christians feel that the tides have turned against them.
Anti-White bias	White People Face Bias in the U.S. According to recent reports, the majority of White people in the U.S. believe they are facing more discrimination than ever before. White people perceive new laws and policies as opening them up to discriminatory treatment. These perceptions may not be unfounded. White Americans today routinely report facing prejudice in hiring because employers are mostly interested in minority employees. Similarly, colleges all over the nation are favoring minority applicants over qualified White applicants, and the White students who are admitted report that their perspective is not appreciated on campuses. These days, many feel White people are expected to apologize for their race. While the U.S. aims for racial equality and tolerance, many White Americans feel that the tides have turned against them.
Anti-Black bias	Black People Face Bias in the U.S. According to recent reports, the majority of Black people in the U.S. believe they are facing more prejudice and discrimination than they have in decades. There is overwhelming objective evidence consistent with these perceptions. While some things have improved for Black people over the last few decades, discrimination against Black people is a persistent issue. Black Americans today routinely report facing racially unjust outcomes in domains like healthcare, the criminal justice system, and education. For example, Black Americans are far more likely to be incarcerated and serve longer sentences than White people for the same crimes. Moreover, Black people are routinely criminalized and have the cops called on them for things as benign as walking home or barbequing in public. Black Americans are more likely to be born into poverty and have fewer opportunities to overcome it. Despite the clear persistence of anti-Black racism, Black people are expected to ignore discrimination and to “move on” from centuries of racial subjugation. While the U.S. claims to value racial equality and tolerance, many Black Americans feel that they continue to be mistreated because of their race.
Control	You have been randomly assigned to not read an article. We will now ask you a number of questions about various topics that other participants read about. There are no right or wrong answers. We simply want to know your opinion.

After reading their respective blurb, participants completed three separate scales assessing their personal perceptions of how much bias is faced in the U.S. by Christians (six items in Experiment 1 and seven in Experiment 2; e.g., “Christians are victims of prejudice in the U.S.”; Wilkins et al., 2022), White people (eight items; e.g., “White people are victims of prejudice in the U.S.”; adapted from Wilkins & Kaiser, 2014), and Black people (six items; e.g., “Black people are victims of prejudice in the U.S.”). All items were assessed on a scale from 1 (strongly disagree) to 7 (strongly agree). For bivariate Pearson correlations, reliability estimates, and descriptive statistics, see Tables 2 through 4.

Table 2.

Experiment 1 Correlations and Descriptive Statistics

	Anti-Christian bias	Anti-White bias	Anti-Black bias	M	SD	α
Anti-Christian bias	1.00	—	—	4.37	1.49	.87
Anti-White bias	0.64	1.00	—	4.32	1.53	.91
Anti-Black bias	−0.26	−0.47	1.00	4.60	1.28	.81

Table 3.

Experiment 2 White Christian Correlations and Descriptive Statistics

	Anti-Christian bias	Anti-White bias	Anti-Black bias	M	SD	α
Anti-Christian bias	1.00	—	—	4.00	1.43	.89
Anti-White bias	0.62	1.00	—	4.06	1.56	.93
Anti-Black bias	−0.39	−0.51	1.00	4.30	1.32	.88

Table 4.

Experiment 2 Black Christian Correlations and Descriptive Statistics

	Anti-Christian bias	Anti-White bias	Anti-Black bias	M	SD	α
Anti-Christian bias	1.00	—	—	3.79	1.44	.85
Anti-White bias	0.26	1.00	—	2.35	1.22	.83
Anti-Black bias	−0.01	−0.46	1.00	5.96	1.03	.76

Results

In both experiments, we tested our hypotheses by separately regressing dependent variables of interest on condition (dummy coded with the control condition coded as the reference group except when two experimental conditions were directly compared). In Experiment 2, we added participant race (dummy coded) as a predictor along with the interaction between condition and race. We report simple effects tests for each racial group below. Because predictors were dummy coded, we report Cohen’s d as a measure of pairwise comparison effect size (pairwise mean difference divided by the square root of the pooled variances of each condition).

Given that many of our preregistered exclusions were based on posttreatment variables, across all studies we also conducted intent-to-treat analyses to ensure the robustness of effects (see Montgomery et al., 2018). Significant effects reported in this article retain significance and directionality even in intent-to-treat analyses unless otherwise noted.

Manipulation checks

Before testing the main predictions, we first ensured that our manipulations effectively increased perceptions of bias targeting the specified groups. To do so, we compared perceptions of anti-White, anti-Black, and anti-Christian bias between the experimental conditions and control conditions.

As shown in Figure 1, reading about anti-White bias increased White Christians’ perceptions of anti-White bias in Experiment 1, b = 0.79, d = 0.53, t(539) = 4.40, p < .001, 95% confidence interval (CI) = [0.44, 1.14], and Experiment 2, b = 0.60, d = 0.43, t(2587) = 5.82, p < .001, 95% CI = [0.40, 0.80], but did not influence Black Christians’ perceptions of anti-White bias, b = 0.19, d = 0.13, t(2587) = 1.77, p = .077, 95% CI = [−0.02, 0.39]. As shown in Figure 2, reading about anti-Christian bias increased perceptions of bias against Christians for White Christians in Experiment 1, b = 0.73, d = 0.51, t(539) = 4.25, p < .001, 95% CI = [0.39, 1.07], Experiment 2, b = 0.68, d = 0.49, t(2586) = 6.50, p < .001, 95% CI = [0.48, 0.89], and Black Christians in Experiment 2, b = 0.87, d = 0.62, t(2586) = 7.89, p < .001, 95% CI = [0.65, 1.08]. As shown in Figure 3, reading about anti-Black bias did not increase perceptions of anti-Black bias in Experiment 1 for White Christians, b = 0.08, d = 0.06, t(539) = 0.55, p = .586, 95% CI = [−0.21, 0.36]. After strengthening the manipulation for Experiment 2, White participants in the anti-Black bias condition perceived significantly more anti-Black bias than those in the control condition, b = 0.26, d = 0.22, t(2586) = 2.90, p = .004, 95% CI = [0.08, 0.44]. Reading about anti-Black bias did not increase Black Christians’ perceptions of anti-Black bias, b = 0.14, d = 0.12, t(2586) = 1.50, p = .134, 95% CI = [−0.04, 0.32], perhaps because mean levels were high even in the control condition (M = 5.92).

Fig. 1.

Density plot of perceptions of anti-White bias by condition and sample. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Fig. 2.

Density plot of perceptions of anti-Christian bias by condition and sample. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Fig. 3.

Density plot of perceptions of anti-Black bias by condition and sample. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Primary question: Did reading about anti-Christian bias increase perceptions of anti-White bias?

As predicted, White Christians who read about anti-Christian bias perceived more anti-White bias than those in the control condition in Experiment 1, b = 0.36, d = 0.24, t(539) = 2.00, p = .047, 95% CI = [0.01, 0.71], and Experiment 2, b = 0.22, d = 0.16, t(2587) = 2.07, p = .038, 95% CI = [0.01, 0.42]. The Experiment 2 effect became marginally significant when including participants who failed preregistered manipulation comprehension or attention checks, b = 0.19, d = 0.14, t(3031) = 1.92, p = .056, 95% CI = [−0.004, 0.38]. Reading about anti-Christian bias did not increase Black Christians’ perceptions of anti-White bias, b = 0.06, d = 0.04, t(2587) = 0.51, p = .608, 95% CI = [−0.16, 0.27]. See Figure 1.

Secondary questions

Did reading about anti-White bias also increase perceptions of anti-Christian bias?

Reading about anti-White bias increased White Christians’ perceptions of anti-Christian bias in Experiment 1, b = 0.34, d = 0.24, t(539) = 1.99, p = .047, 95% CI = [0.01, 0.68]. This effect became marginally significant when including participants who failed manipulation comprehension or attention checks, b = 0.29, d = 0.20, t(667) = 1.88, p = .060, 95% CI = [−0.01, 0.60]. We did not replicate this effect in Experiment 2 for either racial group—White Christians: b = 0.07, d = 0.05, t(2586) = .65, p = .518, 95% CI = [−0.14, 0.27]; Black Christians: b = −0.06, d = −0.04, t(2586) = −0.54, p = .589, 95% CI = [−0.27, 0.15]. Thus, the perceptual link between anti-White and anti-Christian bias appears to be primarily unidirectional. See Figure 2.

Did reading about anti-Christian bias increase perceptions of anti-Black bias?

Reading about anti-Christian bias did not increase perceptions of anti-Black bias in either study and for either racial group—White Christians (Experiment 1): b = 0.27, d = 0.21, t(539) = 1.75, p = .080, 95% CI = [−0.03, 0.56]; White Christians (Experiment 2): b = 0.06, d = 0.05, t(2586) = 0.62, p = .537, 95% CI = [−0.12, 0.23]; Black Christians: b = 0.02, d = 0.01, t(2586) = 0.18, p = .856, 95% CI = [−0.17, 0.20]—suggesting that the effects reported above cannot be explained by inferring that anti-Christian bias increases perceptions of bias in general. That is, there is a unique association between perceptions of anti-Christian and anti-White bias, particularly among White Christians. See Figure 3.

Discussion

As hypothesized, reading about anti-Christian bias increased White Christians’ perceptions of anti-White bias. There was no evidence that anti-Christian bias evoked greater perceptions of anti-Black bias or that anti-Christian bias influenced Black Christians’ perceptions of anti-White bias.

Experiment 3

Given the perceptual link between anti-Christian and anti-White biases, we theorized that politicians could subtly convey concern for White people and avoid social sanction by discussing concerns about anti-Christian bias. Experiment 3 examined whether White Christians perceive a politician worried about anti-Christian bias as also being concerned with anti-White bias and more likely to fight for White people (compared with a control). Experiment 3 also tested whether a politician expressing concern about anti-Christian bias is seen as less offensive than a politician expressing concern about anti-White bias.

Method

Participants

We recruited 1,344 adult participants from Prime Panels. We applied our preregistered exclusion criteria, removing 158 participants who did not report being White Christians living in the U.S. and 536 participants who either failed a multiple-choice manipulation comprehension check (e.g., “What did you just read?”) or an attention check (“If you are paying attention, select ‘strongly agree.’”). Below we report data for 650 White Christians living in the U.S. (39.29% men, 61.07% women; M_age = 55.92 years, SD_age = 16.58 years). Monte Carlo simulations (N = 10,000) of data with the observed sample size and standardized mean differences of 0.31 achieved 80% power (0.28 for 70% power and 0.36 for 90% power).

Procedure

Participants completed a brief series of demographic questions and then were randomly assigned to read one of four excerpts from a fictitious interview with a local politician who was asked what issue they felt was most pressing in their community (that they would focus on if elected). Participants read about a politician who either discussed (a) anti-White bias, (b) anti-Christian bias, (c) concerns about religious freedom, or (d) the economy (control condition). We included the religious freedom condition in addition to the anti-White and anti-Christian conditions because talking about religious freedom could conceivably imply concern about anti-Christian bias but may be less offensive because religious freedom, at face value, includes all religious groups. Indeed, prior work suggests people expect religious freedom arguments will be more effective in evading pushback than anti-Christian bias arguments (Danbold et al., 2022). At the same time, religious freedom arguments are generally made by conservative Christians (PRRI, 2021a) and may be readily perceived as coded language in their own right. See Table 5 for the manipulation text.

Table 5.

Experiment 3 Manipulation Text

Condition	Manipulation
Anti-Christian bias	Interviewer: “What do you perceive as an important issue in the U.S. that you would focus on if elected?” Politician: “I believe that one of the most pressing issues our community is facing is bias against Christians. Our country was founded upon Christian values, but now it’s a sin to even talk about our Christian heritage. Christians all around the country feel like they must shut up and keep silent, or else they will be labeled as a bigot. We need to take concrete actions now to stop these attacks on Christians.”
Anti-White bias	Interviewer: “What do you perceive as an important issue in the U.S. that you would focus on if elected?” Politician:” I believe that one of the most pressing issues our community is facing is bias against White people. Our country was founded upon European values, but now it’s taboo to even talk about our European heritage. White people all around the country feel like they must shut up and keep silent, or else they will be labeled as a racist. We need to take concrete actions now to stop these attacks on White people.”
Religious freedom	Interviewer: “What do you perceive as an important issue in the U.S. that you would focus on if elected?” Politician: “I believe that one of the most pressing issues our community is facing is the assault on religious freedom. Our country was founded upon principles of religious freedom, but now it’s a sin to even live life according to one’s religious beliefs. Religious people all around the country feel like they must abstain from practicing their faith in public, or else they will be punished by others. We need to take concrete actions now to stop these attacks on religious freedom.”
Control (economy)	Interviewer: “What do you perceive as an important issue in the U.S. that you would focus on if elected?” Politician: “I believe that one of the most pressing issues our community is facing is the assault on our economy. Our country was founded upon principles of economic opportunity and growth, but now many are struggling to make ends meet. Businesses are closing and unemployment is on the rise. People are worried about their financial future and their concerns are not heard by policy makers. We need to take concrete actions now to stop these attacks on Americans’ economic futures.”

After reading the excerpt, participants completed a manipulation comprehension check (in which they identified who was interviewed in the excerpt they read) and then responded to questions assessing their perceptions of the politician. Afterward, participants completed a questionnaire about their social and political attitudes.

The procedure adhered to ethical guidelines set by the authors’ institutional review board.

Measures

Unless otherwise noted, all measures were assessed on a 1 (strongly disagree) to 7 (strongly agree) scale. See Table 6 for bivariate Pearson correlations and descriptive statistics.

Table 6.

Experiment 3 Correlations and Descriptive Statistics

	Anti-Christian bias concern	Anti-White bias concern	Religious freedom concern	Offensiveness of speech	Politician’s controversiality	Fight for Christians	Fight for White people	Fight for Black people	Conservatism of politician	M	SD	α
Anti-Christian bias concern	1.00	—	—	—	—	—	—	—	—	4.50	2.03
Anti-White bias concern	0.32	1.00	—	—	—	—	—	—	—	4.25	2.07
Religious freedom concern	0.75	0.22	1.00	—	—	—	—	—	—	4.49	2.03
Offensiveness of speech	0.02	0.29	−0.07	1.00	—	—	—	—	—	3.72	1.53	.87
Politician’s controversiality	0.07	0.28	−0.03	0.79	1.00	—	—	—	—	4.21	1.19	.76
Fight for Christians	0.58	0.16	0.51	−0.18	−0.10	1.00	—	—	—	5.24	1.70
Fight for White people	0.29	0.49	0.21	0.00	0.05	0.59	1.00	—	—	5.32	1.56
Fight for Black people	0.21	−0.09	0.30	−0.39	−0.40	0.35	0.13	1.00	—	4.21	1.83
Conservatism of politician	0.18	0.10	0.10	0.02	0.12	0.28	0.26	−0.16	1.00	4.83	1.91

The politician’s perceived concern about bias against White people, bias against Christians, and concern about religious freedom were measured with a single item each (e.g., “To what extent do you think the politician is concerned about bias against White people in the U.S.?”) anchored at 1 (not at all concerned) and 7 (extremely concerned).

Palatability was assessed with two separate measures. We measured the offensiveness of the politician’s claim with four items (e.g., “This statement was offensive”). We assessed the politician’s controversiality with four items (e.g., “This politician sounds like a polarizing figure”).

Perceptions of the politician fighting for groups were also measured on a 7-point scale (1 = not at all, 7 = a great deal). Participants rated how likely they thought the politician was to fight for the rights of Christians, White people, and Black people. Each item was assessed separately.

Participants also answered questions about their societal attitudes and made additional evaluations of the politician. For example, we measured perceptions of the politician’s political ideology with two separate items: “What political group do you think the politician is affiliated with?” with Republican, Democrat, Independent, and Libertarian as response options; and “How liberal or conservative do you think the politician is?” anchored at 1 (very liberal) and 7 (very conservative).

Results

Manipulation checks

Participants in the anti-White bias condition perceived the politician as more concerned about anti-White bias than those in the control condition, b = 2.24, d = 1.18, t(646) = 10.53, p < .001, 95% CI = [1.82, 2.65]. Participants in the anti-Christian bias condition perceived the politician as more concerned about bias against Christians than those in the control condition, b = 0.77, d = 1.24, t(645) = 10.98, p < .001, 95% CI = [1.88, 2.70]. Similarly, participants in the religious freedom condition perceived the politician as more concerned with religious freedom than those in the control condition, b = 2.08, d = 1.14, t(646) = 9.86, p < .001, 95% CI = [1.67, 2.50]. Thus, the three experimental conditions were successfully manipulated. See Figure 4.

Fig. 4.

Density plot of politician’s perceived concern about group bias and religious freedom. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Examining perceived concern about group bias and religious freedom

As hypothesized, the politician who voiced concerns about anti-Christian bias was perceived as caring more about anti-White bias than the control politician, b = 0.56, d = 0.30, t(646) = 2.65, p = .008, 95% CI = [0.15, 0.98]. Additionally, the politician concerned with anti-White bias was perceived as more concerned with anti-Christian bias than the control politician, b = 0.77, d = 0.42, t(645) = 3.70, p < .001, 95% CI = [0.36, 1.18], consistent with Experiment 1. Thus, we found evidence of a perceived link between anti-Christian and anti-White bias claims.

As hypothesized, the politician concerned with anti-Christian bias was perceived as more concerned about religious freedom relative to the control politician, b = 1.94, d = 1.06, t(646) = 9.45, p < .001, 95% CI = [1.54, 2.35], and the religious freedom politician was perceived as more concerned about anti-Christian bias than the control politician, b = 1.55, d = 0.83, t(645) = 7.20, p < .001, 95% CI = [1.12, 1.97].

Interestingly, the anti-White bias politician was perceived as caring more about religious freedom, and the religious freedom politician as caring more about anti-White bias, each relative to the control politician, b = 0.54, d = 0.29, t(646) = 2.61, p = .009, 95% CI = [0.13, 0.94], and b = 0.50, d = 0.26, t(646) = 2.28, p = .023, 95% CI = [0.07, 0.93], respectively. See Figure 4.

Examining palatability of bias claims

As hypothesized, participants in the anti-Christian bias condition perceived the politician as less controversial and less offensive than those in the anti-White bias condition, b = −0.40, d = −0.36, t(646) = −3.37, p < .001, 95% CI = [−0.64, −0.17], and b = −0.45, d = −0.30, t(646) = −2.85, p = .005, 95% CI = [−0.76, −0.14], respectively. Nonetheless, participants in the anti-Christian bias condition perceived the politician as more controversial and the politician’s statement as more offensive than participants in the control condition, b = 0.72, d = 0.64, t(646) = 5.72, p < .001, 95% CI = [0.47, 0.97], and b = 0.66, d = 0.45, t(646) = 3.97, p < .001, 95% CI = [0.33, 0.99], respectively.

Although we hypothesized that participants in the religious freedom condition would perceive the politician as less controversial and the statement as less offensive than those in the anti-Christian bias condition, neither hypothesis was supported, b = −0.01, d = −0.01, t(646) = −0.06, p = .953, 95% CI = [−0.25, 0.23], and b = −0.04, d = −0.02, t(646) = −0.22, p = .823, 95% CI = [−0.36, 0.28], respectively.

Thus, participants perceived the politician concerned about anti-Christian bias as more palatable than the one concerned about anti-White bias, albeit less palatable than the control politician. And despite face-value neutrality, participants perceived a politician concerned about religious freedom quite similarly to a politician concerned about anti-Christian bias. See Figure 5.

Fig. 5.

Density plot of politician’s perceived palatability. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Perceived likelihood of fighting for groups

Both the anti-White and anti-Christian bias politicians were perceived as more likely to fight for White people than the control politician, b = 0.85, d = 0.56, t(646) = 4.94, p < .001, 95% CI = [0.51, 1.19], and b = 0.37, d = 0.25, t(646) = 2.18, p = .030, 95% CI = [0.04, 0.71], respectively. The anti-Christian bias politician was perceived as more likely to fight for Christians compared with the control politician, but the anti-White bias politician was not, b = 1.00, d = 0.61, t(646) = 5.38, p < .001, 95% CI = [0.63, 1.36], and b = −0.04, d = −0.03, t(646) = −0.22, p = .823, 95% CI = [−0.41, 0.32], respectively. The religious freedom politician was perceived as more likely to fight for Christians, but not White people, than the control politician, b = 0.58, d = 0.35, t(646) = 3.02, p = .003, 95% CI = [0.20, 0.95], and b = 0.16, d = 0.11, t(646) = 0.93, p = .353, 95% CI = [−0.18, 0.51], respectively.

We also explored condition effects on perceptions that the politician would fight for Black people. Likely a sign that the politician was perceived as racist, the anti-White bias politician was perceived as less likely to fight for Black people than the control politician, b = −0.83, d = −0.46, t(646) = −4.13, p < .001, 95% CI = [−1.23, −0.44]. Importantly, the religious freedom condition and anti-Christian bias condition did not differ from the control condition, b = −0.14, d = −0.08, t(646) = −0.67, p = .501, 95% CI = [−0.55, 0.27], and b = −0.07, d = −0.04, t(646) = −0.37, p = .712, 95% CI = [−0.47, 0.32], respectively. This suggests that racial dog whistles in the form of anti-Christian bias claims avoid the perceived anti-Black connotations associated with more direct pro-White advocacy. Participants in the anti-Christian bias condition perceived the politician as more willing to fight for Black people than those in the anti-White bias condition, b = 0.76, d = 0.42, t(646) = 3.97 p < .001, 95% CI = [0.38, 1.14], consistent with this theorizing. See Figure 6.

Fig. 6.

Density plot of the extent to which politicians will fight for Black people, Christians, and White people. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Exploratory analyses: Are effects explained by perceived politician conservativism?

Because each bias claim was associated with greater conservatism (see the Supplemental Material available online), it is plausible that the anti-Christian bias politician was perceived as being more concerned with bias against White people because they were perceived as being more conservative. However, the evidence does not support this claim. Even when controlling for the politician’s perceived conservatism and suspected political party (dummy coded), the anti-Christian bias politician was perceived as more concerned about anti-White bias than the control politician, b = 0.54, d = 0.28, t(641) = 2.53, p = .012, 95% CI = [0.12, 0.95]. Similarly, the anti-White bias politician was perceived as more concerned about anti-Christian bias than the control politician, b = 0.80, d = 0.43, t(640) = 3.86, p < .001, 95% CI = [0.39, 1.21]. Taken together, the results suggest that perceptions of the politicians’ political conservatism and political party affiliation do not explain our results.

Discussion

White Christians perceived a politician worried about anti-White bias as more controversial, more offensive, and less willing to fight for Black people than one concerned about anti-Christian bias. Despite being more palatable, the politician worried about anti-Christian bias still conveyed concern about anti-White bias. Thus, speaking about anti-Christian bias may communicate concern about anti-White bias without some cons of outright anti-White bias claims (e.g., being perceived as racist; Blodorn & O’Brien, 2013). Interestingly, politicians concerned about religious freedom and anti-Christian bias were perceived similarly by White Christians, despite previous work that would predict preference for the politician concerned about religious freedom (Danbold et al., 2022).

Experiment 4

We hypothesized that Black Christians would consider a politician worried about anti-White bias as less likely to fight for Black people and as more offensive relative to the control politician. We made no predictions comparing anti-White bias and anti-Christian bias conditions to each other, or about the anti-Christian bias politician’s concerns for different racial groups.

Perceptions of the politician’s willingness to fight for Black people and the offensiveness of the claim were intended as primary dependent variables; mistakenly, we failed to paste primary questions related to these variables into the preregistration’s “Hypothesis” section. Because we had no clear hypotheses regarding whether and in what direction these effects would emerge, we neglected to specify corresponding analyses on the preregistration.

Method

Participants

We recruited 593 Black Christians from Prolific. We applied our preregistered exclusion criteria and removed 68 participants who did not report being Black Christians living in the U.S. We then removed six participants who failed a manipulation comprehension check (“What did you just read?”) twice. Below we report data for 519 Black Christians living in the U.S. (57.42% women, 42.39% men; M_age = 38.03 years, SD_age = 12.98 years). Monte Carlo simulations (N = 10,000) of data with the observed sample size and standardized mean differences of 0.30 achieved 80% power (0.27 for 70% power and 0.35 for 90% power).

Procedure

The procedure was similar to Experiment 3. Participants first completed demographic questions, and then they were randomly assigned to read one of three excerpts from an interview with a local politician who was asked what issue they felt was most pressing in their community that they would focus on if elected. The politician discussed either anti-White or anti-Christian bias, or, in the control condition, the economy. After reading the excerpt, participants completed a manipulation comprehension check. They were allowed to reread materials after a first failure but were excluded after a second failure. Participants then responded to the questions described below.

The procedure adhered to ethical guidelines set by the authors’ institutional review board.

Measures

Unless otherwise noted, all measures were assessed on a scale from 1 (strongly disagree) to 7 (strongly agree).

Perceptions of the politician’s concerns and palatability of the claim

Participants responded to items regarding the extent to which they thought the politician was concerned about bias against White people, Christians, and Black people (e.g., “To what extent do you think the politician is concerned about bias against Black people in the U.S.?”) on a scale from 1 (not at all concerned) to 7 (very concerned). Each measure was assessed separately and included two items. Perceived offensiveness of the claim was measured with four items (e.g., “This statement was offensive”).

Perceptions of the politician fighting for groups

On a 7-point scale (1 = not at all, 7 = a great deal) participants rated the perceived likelihood that the politician would fight for and defend the rights of Christians, White people, and Black people.

Perceptions of the politician’s political attitudes and identity

Participants also answered questions about their societal attitudes and evaluations of the politician. For example, we measured perceptions of the politician’s political ideology with two items: “What political group do you think the politician is affiliated with?” with Republican, Democrat, Independent, and Libertarian being response options; and “How liberal or conservative do you think the politician is?” anchored at 1 (very liberal) and 7 (very conservative). See Table 7 for bivariate Pearson correlations and descriptive statistics.

Table 7.

Experiment 4 Correlations and Descriptive Statistics

	Anti-Christian bias concern	Anti-White bias concern	Offensiveness of speech	Fight for Christians	Fight for White people	Fight for Black people	Conservatism of politician	M	SD	α
Anti-Christian bias concern	1.00	—	—	—	—	—	—	4.39	1.90	.90
Anti-White bias concern	0.13	1.00	—	—	—	—	—	4.60	1.85	.85
Offensiveness of speech	0.07	0.58	1.00	—	—	—	—	3.67	1.82	.91
Fight for Christians	0.68	−0.04	−0.05	1.00	—	—	—	5.04	1.74
Fight for White people	0.00	0.55	0.43	0.09	1.00	—	—	5.61	1.51
Fight for Black people	0.12	−0.28	−0.51	0.27	−0.33	1.00	—	3.01	1.79
Conservatism of politician	0.17	0.38	0.45	0.14	0.39	−0.46	1.00	5.24	1.76

Results

Manipulation checks and perceptions of concern about bias against different groups

Black Christians in the anti-White bias condition perceived the politician as more concerned about anti-White bias than those in the control condition, b = 2.56, d = 1.69, t(513) = 15.54, p < .001, 95% CI = [2.24, 2.89]. Participants in the anti-Christian bias condition perceived the politician as more concerned about bias against Christians than those in the control condition, b = 2.69, d = 1.83, t(513) = 17.09, p < .001, 95% CI = [2.38, 3.00]. Thus, both experimental manipulations were effective. See Figure 8.

We also explored whether Black Christians perceived a link between anti-White and anti-Christian bias claims. Black Christians who read about the anti-Christian bias politician perceived the politician as significantly more concerned about bias against White people relative to control participants, b = 0.80, d = 0.52, t(513) = 4.90, p < .001, 95% CI = [0.48, 1.11]. Similarly, participants in the anti-White bias condition perceived the politician as more concerned about anti-Christian bias compared with control participants, b = 0.40, d = 0.27, t(513) = 2.48, p = .013, 95% CI = [0.08, 0.71]. Thus, Black Christians perceived a link in messaging about anti-White and anti-Christian bias. See Figure 7.

Fig. 7.

Density plot of politician’s perceived concern about group bias. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Examining perceived likelihood of fighting for groups

As predicted, the anti-White bias politician was perceived as less likely to fight for Black people than the control politician, b = −1.91, d = −1.19, t(511) = −10.93, p < .001, 95% CI = [−2.26, −1.57]. Black Christians perceived the anti-Christian bias politician as less likely to fight for Black people than the control politician, b = −0.50, d = −0.31, t(511) = −2.92, p = .004, 95% CI = [−0.84, −0.16]. Importantly, participants in the anti-Christian bias condition perceived the politician as more willing to fight for Black people than those in the anti-White bias condition, b = 1.41, d = 0.88, t(511) = 8.15, p < .001, 95% CI = [1.07, 1.75]. This suggests that Black Christians perceive both anti-Christian and, to a greater extent, anti-White bias claims as implying anti-Blackness, but anti-Christian bias claims are relatively preferred. See Figure 8.

Fig. 8.

Density plot of the extent to which the politician will fight for Black people. The mean is indicated by a black dot surrounded by a 99% confidence interval. Gray distributions indicate a significant difference (α = .05) between that condition’s mean and the analogous control condition.

Examining palatability of bias claims

The anti-Christian bias claim was perceived as less offensive than the anti-White bias claim, b = −1.73, d = −1.28, t(511) = −11.88, p < .001, 95% CI = [−2.01, −1.44]. Nonetheless, Black Christians perceived both the anti-White and anti-Christian claims as more offensive than the control claim, b = 3.02, d = 2.24, t(511) = 20.55, p < .001, 95% CI = [2.73, 3.31], and b = 1.30, d = 0.96, t(511) = 8.96, p < .001, 95% CI = [1.01, 1.58], respectively.

Thus, the politician who voiced concerns about anti-Christian bias was perceived as more palatable than the one concerned about anti-White bias, albeit less palatable than the control politician. See Figure 9.

Fig. 9.

Exploratory analyses: Are effects explained by perceived politician conservativism?

Because each claim of bias was associated with greater conservatism (than the control; see the Supplemental Material), we tested whether the anti-Christian bias politician was only perceived as being more concerned about bias against White people (relative to the control politician) because they were perceived as being more conservative. Consistent with Experiment 3, findings held even when controlling for the politician’s perceived conservatism and suspected political party (dummy coded); the anti-Christian bias politician was perceived as more concerned with anti-White bias than the control politician, b = 0.56, d = 0.37, t(506) = 3.41, p < .001, 95% CI = [0.24, 0.88].

It is also plausible that participants perceived the anti-White bias politician as less palatable and less willing to fight for Black people relative to the anti-Christian bias politician only because they perceived the anti-White bias politician as more conservative. However, even when controlling for the politician’s perceived conservatism and suspected political party, the anti-White bias politician’s claim was still perceived as more offensive than the anti-Christian bias politician’s claim, b = 1.58, d = 1.17, t(506) = 11.24, p < .001, 95% CI = [1.31, 1.86]. Similarly, the anti-White bias politician was still perceived as less willing to fight for Black people than the anti-Christian bias politician, b = −1.16, d = −0.72, t(506) = −7.25, p < .001, 95% CI = [−1.48, −0.85]. Taken together, our findings suggest that the relative palatability of claiming anti-Christian bias (which may make such claims less subject to censorship and thus more effective) is not driven by these claimants being viewed as less conservative.

General Discussion

Four studies tested whether people perceive anti-Christian bias as signaling anti-White bias (Experiments 1 and 2) and whether this signal can work as a racial dog whistle whereby people perceive anti-Christian bias claims as communicating concern for White people (Experiments 3 and 4). White Christians inferred that bias against Christians reflected bias against White people, but Black Christians did not. Across racial groups, reading about anti-Christian bias did not influence perceptions of anti-Black bias. Black and White Christians evaluated a politician concerned about anti-Christian bias as more concerned about anti-White bias than a politician concerned about a control topic.

Expressing concern about anti-Christian bias can effectively signal concern for White people, making anti-Christian bias claims ripe for dog whistling. This is noteworthy given that a similar percentage of Black Americans (72%) and White Americans (71%) identify as Christian (PRRI, 2021b).

Alternative explanations, limitations, and future directions

We explored two alternative explanations. First, we examined whether perceiving greater anti-White bias in response to anti-Christian bias resulted from White Christians overestimating the percentages of White people who are Christian and Christians who are White. White Christians underestimated both of these figures; thus, White Christians’ perceptual overlap is not simply overestimation. As described above, we also ruled out the alternative that perceiving the politician as conservative accounted for effects.

What types of policy beliefs do dog whistles signal? Because politicians claiming anti-Christian bias are perceived as more likely to fight for White people, we might expect them to endorse more pro-White policies. However, because politicians concerned about anti-Christian bias were not seen as being less likely to fight for Black people, they may not be perceived as opposed to policies aimed at aiding Black Americans. Exploratory analyses support this latter proposition (see the Supplemental Material). However, because we did not include direct measures of pro-White policies, further investigation is warranted.

These results are specific to White and Black Christians in the U.S. Thus, we urge caution in generalizing our findings to other religious or racial groups in the U.S. or to other cultural/political contexts. Future work may test whether bias claims against other religions can serve as racial dog whistles in countries in which ethnoreligious nationalism is prevalent (e.g., Hindu nationalism in India; George, 2022). Because anti-Christian bias rhetoric has become frequent in U.S. political discourse, participants’ awareness of and attitudes toward coded language might also fluctuate with political discourse.

We focused on “Christian” as a racial dog whistle for “White” because religion is particularly effective given its racial deniability, Americans generally agree religious freedom should be protected (Fox, 2021), and religious identities remain demonstrably racialized (Perry, 2023). By contrast, other conservative issues such as proheteronormativity and protraditional gender roles are not racialized to the same extent because racial minorities often support heteronormative (Sherkat, 2017) and patriarchal views (Scarborough et al., 2019). Thus, compared with anti-Christian bias, these issues likely would not communicate White race/racial concerns or may do so to a much lesser extent. Nonetheless, it is an empirical question whether bias against other racialized identities could be leveraged as racial dog whistles.

This research does not provide insight as to who uses coded language or whether this is done intentionally. Future work could examine whether leveraging religion as racially coded language is purposeful and to what end (e.g., for political benefit). Also, is it used by political elites or the broader populace?

Future research could also examine for whom anti-Christian bias claims work as a dog whistle. For example, those most concerned about anti-Christian bias may also be more concerned about anti-White bias, and thus more susceptible. Although we were unable to directly test this, we examined whether White Christians who endorsed Christian nationalism—an ideology highly associated with both (a) perceived threat to Christianity (Al-Kire et al., 2022) and (b) pro-Whiteness (Davis & Perry, 2021)—reported greater perceptions of anti-White bias after reading about anti-Christian bias (see the Supplemental Material). Effects did not differ on the basis of Christian nationalism, suggesting those who are more concerned about anti-White bias are not simply more affected by anti-Christian bias claims. However, future work could test this using more direct measures of group bias concerns.

Implications

Our results have important implications for religious and racial discourse in the U.S. Because “Christian” can stand in for “White,” political elites may take advantage of this coded language. Additionally, given that many websites are moderated to prevent hate speech, users may be particularly motivated to adopt coded language (Bhat & Klen, 2020). Our results suggest one way people could evade restrictions is by discussing bias against Christians.

This work makes important theoretical contributions to the dog-whistle literature. In contrast to previous findings, which suggest White people react to indirect racial messages and Black people react to direct racial messages (White, 2007), our work suggests White and Black Americans react to both. Importantly, although Black Christians did not perceive anti-Christian bias to reflect greater anti-White bias, they did infer a politician concerned about anti-Christian bias was more concerned about anti-White bias. Thus, Black Christians perceive indirect racial messaging, even when they do not associate anti-Christian and anti-White biases themselves.

Conclusion

For Christian Americans, anti-Christian bias claims can serve as a racial dog whistle conveying concern about White people. Thus, when former U.S. President Trump said, “No president has ever fought for Christians as hard as I have” (Vigdor, 2023), he likely signaled concern for White Americans (and lower concern for Black Americans) without ever mentioning race. Listeners likely understood as much.

Supplemental Material

sj-docx-1-pss-10.1177_09567976241236162 – Supplemental material for White by Another Name? Can Anti-Christian Bias Claims Serve as a Racial Dog Whistle?

Supplemental material, sj-docx-1-pss-10.1177_09567976241236162 for White by Another Name? Can Anti-Christian Bias Claims Serve as a Racial Dog Whistle? by Rosemary L. Al-Kire, Chad A. Miller, Michael H. Pasek, Samuel L. Perry and Clara L. Wilkins in Psychological Science

Footnotes

Transparency

Action Editor: Mark Brandt

Editor: Patricia J. Bauer

Author Contributions

Rosemary L. Al-Kire: Conceptualization; Data curation; Investigation; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Chad A. Miller: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Visualization; Writing – original draft; Writing – review & editing.

Michael H. Pasek: Conceptualization; Methodology; Supervision; Writing – original draft; Writing – review & editing.

Samuel L. Perry: Conceptualization; Writing – review & editing.

Clara L. Wilkins: Conceptualization; Investigation; Methodology; Project administration; Resources; Supervision; Writing – original draft; Writing – review & editing.

ORCID iDs

Rosemary L. Al-Kire

Chad A. Miller

Michael H. Pasek

Samuel L. Perry

Clara L. Wilkins

Supplemental Material

Additional supporting information can be found at

References

Al-Kire

Pasek

Tsang

J.-A.

Leman

Rowatt

(2022). Protecting America’s borders: Christian nationalism, threat, and attitudes toward immigrants in the United States. Group Processes & Intergroup Relations, 25(2), 354–378. https://doi.org/10.1177/1368430220978291

Apfelbaum

E. P.

Norton

M. I.

Sommers

S. R.

(2012). Racial color blindness: Emergence, practice, and implications. Current Directions in Psychological Science, 21(3), 205–209. https://doi.org/10.1177/0963721411434980

Bhat

Klen

(2020). Covert hate speech: White nationalists and dog whistle communication on Twitter. In Bouvier

Rosenbaum

J. E.

(Eds.), Twitter, the public sphere, and the chaos of online deliberation (pp. 151–172). Palgrave Macmillan. https://doi.org/10.1007/978-3-030-41421-4_7

Blodorn

O’Brien

L. T.

(2013). Evaluations of White American versus Black American discrimination claimants’ political views and prejudicial attitudes. Journal of Experimental Social Psychology, 49(2), 211–216. https://doi.org/10.1016/j.jesp.2012.11.004

Butler

(2021). White evangelical racism: The politics of morality in America. University of North Carolina Press.

Butz

D. A.

Carvalho

J. D.

(2015). Examining the psychological separation of church and state: The American—Christian effect. Psychology of Religion and Spirituality, 7(2), 109–119. https://doi.org/10.1037/a0037706

Chandler

Rosenzweig

Moss

A. J.

Robinson

Litman

(2019). Online panels in social science research: Expanding sampling methods beyond Mechanical Turk. Behavior Research Methods, 51(5), 2022–2038.

Craig

M. A.

Rucker

J. M.

Richeson

J. A.

(2018). Racial and political dynamics of an approaching “majority-minority” United States. The ANNALS of the American Academy of Political and Social Science, 677(1), 204–214. https://doi.org/10.1177/0002716218766269

Danbold

Onyeador

I. N.

Unzueta

M. M.

(2022). Dominant groups support digressive victimhood claims to counter accusations of discrimination. Journal of Experimental Social Psychology, 98, Article 104233. https://doi.org/10.1016/j.jesp.2021.104233

10.

Davis

J. T.

Perry

S. L.

(2021). White Christian nationalism and relative political tolerance for racists. Social Problems, 68(3), 513–534. https://doi.org/10.1093/socpro/spaa002

11.

Devos

Banaji

M. R.

(2005). American = White? Journal of Personality and Social Psychology, 88(3), 447–466. https://doi.org/10.1037/0022-3514.88.3.447

12.

Fichera

(2024, January). Biden’s Christian ‘persecution’? We assess Trump’s recent claims. The New York Times. https://www.nytimes.com/2023/12/29/us/politics/trump-biden-christianity.html

13.

Fox

(2021). What is religious freedom and who has it? Social Compass, 68(3), 321–341. https://doi.org/10.1177/00377686211012368

14.

George

(2022, February 28). The rise of Hindu nationalism. The MIT Press Reader. https://thereader.mitpress.mit.edu/the-rise-of-hindu-nationalism

15.

Haney-López

(2014). Dog whistle politics: How coded racial appeals have reinvented racism and wrecked the middle class. Oxford University Press.

16.

Jones

R. P.

(2016). The end of White Christian America. Simon & Schuster.

17.

Kaiser

C. R.

Miller

C. T.

(2001). Stop complaining! The social costs of making attributions to discrimination. Personality and Social Psychology Bulletin, 27(2), 254–263. https://doi.org/10.1177/0146167201272010

18.

Mendelberg

(2001). The race card: Campaign strategy, implicit messages, and the norm of equality. Princeton University Press.

19.

Mendelberg

(2008). Racial priming revived. Perspectives on Politics, 6(1), 109–123. https://doi.org/10.1017/S1537592708080092

20.

Montgomery

J. M.

Nyhan

Torres

(2018). How conditioning on posttreatment variables can ruin your experiment and what to do about it. American Journal of Political Science, 62(3), 760–775. https://doi.org/10.1111/ajps.12357

21.

Nteta

T. M.

Lisi

Tarsi

M. R.

(2016). Rendering the implicit explicit: Political advertisements, partisan cues, race, and white public opinion in the 2012 presidential election. Politics, Groups, and Identities, 4(1), 1–29. https://doi.org/10.1080/21565503.2015.1050407

22.

Perry

S. L.

(2022). American religion in the era of increasing polarization. Annual Review of Sociology, 48(1), 87–107. https://doi.org/10.1146/annurev-soc-031021-114239

23.

Perry

S. L.

(2023). Mating call, dog whistle, trigger: Asymmetric alignments, race, and the use of reactionary religious rhetoric in American politics. Sociological Theory, 41(1), 56–82. https://doi.org/10.1177/073527512311536

24.

Pew Research Center. (2014). Religious landscape study. https://www.pewresearch.org/religion/religious-landscape-study

25.

Pew Research Center. (2021). In U.S., far more support than oppose separation of church and state. https://www.pewresearch.org/religion/2021/10/28/in-u-s-far-more-support-than-oppose-separation-of-church-and-state

26.

Public Religion Research Institute. (2021a). Is religious liberty a shield or a sword? https://www.prri.org/research/is-religious-liberty-a-shield-or-a-sword

27.

Public Religion Research Institute. (2021b). The 2020 PRRI census of American religion. https://www.prri.org/research/2020-census-of-american-religion

28.

Reny

T. T.

Valenzuela

A. A.

Collingwood

(2020). “No, you’re playing the race card”: Testing the effects of anti-Black, anti-Latino, and anti-immigrant appeals in the post-Obama era. Political Psychology, 41(2), 283–302. https://doi.org/10.1111/pops.12614

29.

Scarborough

W. J.

Sin

Risman

(2019). Attitudes and the stalled gender revolution: Egalitarianism, traditionalism, and ambivalence from 1977 through 2016. Gender & Society, 33(2), 173–200. https://doi.org/10.1177/08912432188096

30.

Sherkat

D. E.

(2017). Intersecting identities and support for same-sex marriage in the United States. Social Currents, 4(4), 380–400. https://doi.org/10.1177/2329496516663221

31.

Silver

(2008, October 18). “Real” America looks different to Palin, Obama. FiveThirtyEight. https://fivethirtyeight.com/features/real-america-looks-different-to-palin/

32.

U.S. Census Bureau. (2021). QuickFacts: United States. https://www.census.gov/quickfacts/fact/table/US/PST045221

33.

Valentino

N. A.

Hutchings

V. L.

White

I. K.

(2002). Cues that matter: How political ads prime racial attitudes during campaigns. American Political Science Review, 96(1), 75–90. https://doi.org/10.1017/S0003055402004240

34.

Vandermaas-Peeler

Cox

Fisch-Friendman

Jones

(2018). Diversity, division, discrimination: The state of young America. Public Religion Research Institute. https://www.prri.org/research/mtv-culture-and-religion

35.

Vigdor

(2023, June 24). In pitch to evangelicals, Trump casts himself as Christian crusader who helped end Roe v. Wade. The New York Times. https://www.nytimes.com/2023/06/24/us/politics/trump-abortion-evangelicals.html

36.

Wetts

Willer

(2019). Who is called by the dog whistle? Experimental evidence that racial resentment and political ideology condition responses to racially encoded messages. Socius, 5, 2378023119866268. https://doi.org/10.1177/2378023119866268

37.

White

I. K.

(2007). When race matters and when it doesn’t: Racial group differences in response to racial cues. American Political Science Review, 101(2), 339–354. https://doi.org/10.1017/S0003055407070177

38.

Whitehead

A. L.

Scheitle

C. P.

(2018). We the (Christian) people: Christianity and American identity from 1996 to 2014. Social Currents, 5(2), 157–172. https://doi.org/10.1177/2329496517725333

39.

Wilkins

C. L.

Kaiser

C. R.

(2014). Racial progress as threat to the status hierarchy: Implications for perceptions of anti-White bias. Psychological Science, 25(2), 439–446. https://doi.org/10.1177/0956797613508412

40.

Wilkins

C. L.

Wellman

J. D.

Kaiser

C. R.

(2013). Status legitimizing beliefs predict positivity toward Whites who claim anti-White bias. Journal of Experimental Social Psychology, 49(6), 1114–1119. https://doi.org/10.1016/j.jesp.2013.05.017

41.

Wilkins

C. L.

Wellman

J. D.

Toosi

N. R.

Miller

C. A.

Lisnek

J. A.

Martin

L. A.

(2022). Is LGBT progress seen as an attack on Christians? Examining Christian/sexual orientation zero-sum beliefs. Journal of Personality and Social Psychology, 122(1), 73–101. https://doi.org/10.1037/pspi0000363

42.

Zou

L. X.

Cheryan

(2017). Two axes of subordination: A new model of racial position. Journal of Personality and Social Psychology, 112(5), 696–717. https://doi.org/10.1037/pspa0000080

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

0.24 MB