Over the last decade, scholars have found convincing evidence that Christian nationalism is linked to more xenophobic views than those who reject that worldview. These views have manifested in a variety of ways, including prejudice against immigrants, Muslims, Blacks, Latinos, Asians, LGBT people. The consensus is that Christian nationalists do not like those outside their group. A concern for these xenophobic views is whether or not their prejudice would shape their behavioral intentions: if afforded the power, would Christian nationalists unequally apply the law in order to benefit the ingroup and disadvantage outgroups?
In this paper, we provide an answer to that question set in the context of the prospective discriminatory enforcement of the rule of law. While existing work demonstrates Christian nationalism linked to preferential treatment of the ingroup and denigration in various ways of outgroups, we argue that there are two separate mechanisms driving them. Christian nationalism, as measured, is the driver of ingroup love while belief in Christian persecution signals a threat that triggers outgroup hate. Our results signal the importance of capturing them both to fully understand the public face of Christian nationalism as well as to understand dynamics that could allow Christian nationalism to wax and wane as a driver of extremist politics. In this way, we join a growing chorus that emphasizes threat as a critical resource fueling (Christian nationalist) group politics (e.g., Al-Kire et al., Reference Al-Kire, Pasek, Rowatt, Tsang and Leman2021a, Reference Al-Kire, Pasek, Tsang and Rowatt2021b; Djupe, Reference Djupe and Rudolph2022; Djupe et al., Reference Djupe, Lewis and Sokhey2023; Perry, Reference Perry2023).
Prejudice and policy implementation
A cornerstone of democratic governance is the rule of law. Beyond the equal access to ballot and counting the votes as cast, administrations must treat everyone equally and groups must not face discriminatory implementation of public policy. Violations of these standards without legal recourse may weaken trust in governmental institutions, which may undermine political involvement, and lead to the eventual collapse of the state in civil conflict as the only means left to resolve differences (e.g., Tyler, Reference Tyler2003), at least if the aggrieved minorities are strong enough to encourage egalitarian enforcement (e.g., Maravall and Przeworski, Reference Maravall and Przeworski2003).
Therefore, it is important to understand the default orientation of Americans to treat certain groups with prejudice. Average citizens have long been found to dislike others and on that basis deny them basic civil rights protections (e.g., Stouffer, Reference Stouffer1955; Sullivan et al., Reference Sullivan, Piereson and Marcus1982). The particular outgroups shift across time and political tolerance may have increased (contingent on which groups are considered—Sullivan et al., Reference Sullivan, Piereson and Marcus1979), but either way it is clear that intolerance has not disappeared even as civil rights protections have expanded (Gibson, Reference Gibson2006). Some level of intolerance may, in fact, be pro-democratic as there continue to be groups that violate democratic norms and pursue anti-democratic ends (Petersen et al., Reference Petersen, Slothuus, Stubager and Togeby2011; Djupe and Neiheisel, Reference Djupe and Neiheisel2022).
It is possible that civil servants function differently, especially since there are laws that govern their behavior to prohibit discrimination. For example, in the context of election administration where laws often dictate implementation of vote security requirements, like requesting voters show an ID, those requirements are implemented unfairly (Atkeson et al., Reference Atkeson, Bryant, Hall, Saunders and Alvarez2010; White et al., Reference White, Nathan and Faller2015). Even in jurisdictions and electoral contests in which racial interests were unlikely to feature, which otherwise might incentivize discrimination, election administrators still asked non-whites for identification at higher rates than whites (Cobb et al., Reference Cobb, James Greiner and Quinn2012).
Of course, the literature on disproportionate attention from the police and outright discrimination against racial minorities is extensive and the findings consistent: “Black men are more likely than any other group in the United States to be profiled, targeted, accused, investigated, wrongfully committed, harshly sentenced, and incarcerated for crimes” (see also Pettit and Western, Reference Pettit and Western2004; Pew Center on the States, 2008; English et al., Reference English, Bowleg, del Río-González, Tschann, Agans and Malebranche2017, 186). Beyond the essential finding of rampant bias in policing, we are also interested in the contextual variables that affect policing decisions, such as ingroup bias—including increased responsiveness to white victims (e.g., Smith et al., Reference Smith, Visher and Davidson1984). But police also appear to be especially responsive to growing minority populations in otherwise white neighborhoods (e.g., Smith, Reference Smith1986; Stewart et al., Reference Stewart, Baumer, Brunson and Simons2009)—outgroup threat. Though there are many throughlines in this vast area of work, we can find policing decisions that appear to be shaped by multiple dimensions of intergroup relations that reveal structure beyond personal prejudice. If we can take inspiration from this literature, the extensive work about mass intolerance as well as widespread disparate implementation of public-facing laws, we at least expect that enabling survey participants to implement the law can reasonably provide evidence of disparate treatment.
Christian nationalism as ingroup protection
Christian nationalism has been defined as an ideology that seeks to fuse American civic life with a particular type of Christianity and Christian culture (Gorski, Reference Gorski2017; Whitehead and Perry, Reference Whitehead and Perry2020a). Despite being a fusion of two very broad identities, Christian nationalism as a worldview holds one ideal very close that applies universally: a desire to maintain their way of life from a position of dominance (Djupe et al., Reference Djupe, Lewis and Sokhey2023). Specifically, Christian nationalists see the United States as a Christian nation, founded by Christians for Christians (Straughn and Feld, Reference Straughn and Feld2010). As a result, Christian nationalism embraces beliefs, public policies, and politicians that have the potential to exclude those from outside of their group from entering the United States, entering civic life, and gaining power since those outside their group will potentially upset their preferred established order.
Christian nationalism appears to be linked to rejection of outsiders along two principal lines: race/ethnicity and religion. In limiting those seen as apart from Christian society, Muslims and immigrants are classic examples. Not only is Christian nationalism linked to stronger anti-immigrant animus (McDaniel et al., Reference McDaniel, Nooruddin and Shortle2011; Sherkat and Lehman, Reference Sherkat and Lehman2018), it also is linked to restricting the rights of worship for Muslims (Shortle and Gaddie, Reference Shortle and Gaddie2015). Moreover, Al-Kire et al. (Reference Al-Kire, Pasek, Rowatt, Tsang and Leman2021a) find a causal association between Christian nationalism and anti-immigrant stereotyping. With this in mind, it is not surprising that much of the rhetoric of the Christian right has centered on excluding Muslims (Lajevardi, Reference Lajevardi2020) since there is no greater threat to America being a Christian nation than if there are more Americans who not only are not Christian, but have a very low chance of becoming Christian, and have been portrayed as desiring the end of Christianity (Bail, Reference Bail2016). However, Dennen and Djupe (Reference Dennen and Djupe2023, Hobbs et al., Reference Hobbs, Lajevardi, Li and Lucas2023) show that Christian nationalism fuels prejudice to a wide variety of groups not limited to Muslims or Jews, perhaps because Christian nationalism itself appears to generate from angst about demographic change (Al-Kire et al., Reference Al-Kire, Pasek, Tsang and Rowatt2021b; Walker and Haider-Markel, Reference Walker and Haider-Markelforthcoming).
Christian nationalism, at least among whites, hold negative views about domestic groups that compete for rights and power, such as Black Americans. Perry et al. (Reference Perry, Whitehead and Davis2019) discovered that Christian nationalists are significantly more likely to believe not only that White and Black Americans are treated the same by police, but that Black Americans are shot more because they commit more crime. Moreover, white Christian nationalism is linked to greater support of voter ID laws, a move that could disenfranchise Black voters (Perry et al., Reference Perry, Whitehead and Grubbs2022). On top of supporting policies that would be detrimental to African Americans, Christian nationalists tend to support stronger punishments for crimes committed in general (Davis, Reference Davis2018). This does not mean Christian nationalists hate rights altogether; they tend to support rights such as gun rights while disavowing others like freedom of the press or rights to unlawful search and seizure (Davis et al., Reference Davis, Perry and Grubbs2023). While these policy preferences could potentially be chalked up to ideological concerns, Christian nationalists are not only more likely to dislike African Americans, they are also more likely than non-Christian whites to show support for racist whites (Perry and Whitehead, Reference Perry and Whitehead2015; Davis and Perry, Reference Davis and Perry2021). Overall, Christian nationalists' views toward Black Americans range from unfavorable to outright hostile, in part triggered by a barrage of elite cues (Perry, Reference Perry2023).
Another strategy to maintain Christian nationalists' dominant position is to reject alternate sources of information. A well-explored example is the rejection of science. Baker et al. (Reference Baker, Perry and Whitehead2020a) build on previous research that links anti-science beliefs with literalist Christianity and conservatism to find that Christian nationalism is another predictor of anti-science beliefs. A recent manifestation is that Christian nationalists are less trusting of vaccines (Whitehead and Perry, Reference Whitehead and Perry2020b; see also Vegter et al., Reference Vegter, Lewis and Bolin2023) and democracy itself (Djupe et al., Reference Djupe, Lewis and Sokhey2023).
The Christian nationalism literature has shown ingroup/outgroups dynamics in the context of elections. Marsh (Reference Marsh2021) documents that outgroup hatred has been a motivating factor for White evangelicals in presidential elections from 2004 to 2016; white evangelicals are strongly Christian nationalist even if they do not constitute the entire population with that worldview. Christian nationalism was at the forefront in the 2016 US Presidential election, when Whitehead et al. (Reference Whitehead, Perry and Baker2018) demonstrated that it was the strongest predictor of voting for Donald Trump (see also Gorski, Reference Gorski2017). But outgroup attitudes that are strongly correlated with Christian nationalism, such as a fear of outsiders, was another strong predictor of voting for Trump in 2016 (Major et al., Reference Major, Blodorn and Major2018). While the strongest predictor of voting for Trump shifted to xenophobia in 2020, Christian nationalism was still a strong predictor (Baker et al., Reference Baker, Perry and Whitehead2020b).
Despite the numerous articles shedding consistent light on our understanding of Christian nationalism, we believe that the literature can be improved in at least two ways. First, the one piece of evidence that we do not yet have is the controlled comparison of how Christian nationalists would treat the ingroup versus the outgroup. The weight of the evidence is clear and consistent across a wide variety of groups and policies that Christian nationalism is linked to discriminatory views of outgroups. But the questions often put groups on different informational footings in survey questions; for example, African Americans are treated differently because they proportionally commit more crimes (Perry et al., Reference Perry, Whitehead and Davis2019). While we expect results that parallel those from this rapidly growing literature, causal evidence of discriminatory treatment under equal information conditions would only increase the validity of its claims.
Second, we believe that our understanding of Christian nationalism needs a dynamic element rather than assuming dispositional influence. Most of the literature suggests Christian nationalist associations with public policy attitudes are enduring, with some indicating roots in a deep story that may even be hundreds of years in the making (Gorski and Perry, Reference Gorski and Perry2021; Jones, Reference Jones2023). In this view, Christian nationalism captures a stable, historical ideology covering a wide swath of American politics. Though elements may be enduring, we believe that, in the main, Christian nationalism needs to be activated by identified threats to take particular positions that would shore up their power (Djupe et al., Reference Djupe, Lewis and Sokhey2023). In this way, events and elites may work to gin up threats, but those threats may not be everpresent in history and can come and go.
Ingroup–outgroup dynamics
This helps us understand that Christian nationalist effects in existing research may be the result of two distinguishable processes: (1) the Baylor Christian nationalism scale captures a preference for the ingroup's dominance, while (2) threat may trigger outgroup animus. What is particularly important about this framework is that, in general, the two—ingroup love and outgroup hate—have been found to move independently (Brewer, Reference Brewer1999). Growing from the minimal group paradigm that argued that even arbitrary distinctions could motivate strong group identities (e.g., Tajfel, Reference Tajfel1970; Tajfel and Turner, Reference Tajfel, Turner, Worchel and Austin1986), Brewer argues that ingroup love is not a “necessary precursor” of outgroup hate. Many times discriminatory practices, which are often viewed as attacks on the outgroup, are rather ways of supporting and protecting the ingroup (Brewer, Reference Brewer1999), reflections of a perceived moral superiority of the ingroup.
That is not to say that ingroup love and outgroup hate are completely independent of each other. Rather, a situation where ingroup support is likely to be found also provides an environment for outgroup hate to occur through the need to justify moral superiority, sensitivity to threat, anticipation of interdependence, social comparison processes, and power politics (Brewer, Reference Brewer1999). In sum, the five things that make the ingroup landscape prime for outgroup hate all lend themselves to potential threat from those outside their perceived ingroup.
While these dynamics play out in many areas of life where intergroup relations are present, we have seen similar patterns in Americans' attachments to the parties and consequent affective polarization (Iyengar et al., Reference Iyengar, Lelkes, Levendusky, Malhotra and Westwood2019). At this point, affective polarization has been growing, whether captured through feeling thermometers or otherwise (Iyengar, et al., Reference Iyengar, Sood and Lelkes2012), because of augmented dislike of the other party while love for the ingroup party has remained stable (e.g., Iyengar et al., Reference Iyengar, Lelkes, Levendusky, Malhotra and Westwood2019; Finkel et al., Reference Finkel, Bail, Cikar, Ditto, Iyengar, Klar, Mason, McGrath, Nyhan, Rand, Skitka, Tucker, Van Bavel, Wang and Druckman2020). Abramowitz and Webster (Reference Abramowitz and Webster2016) simply refer to “negative partisanship” as the motivating force rather than mere partisanship. And there may be instances of policy agreement that cannot make the leap across interparty animus (Mason, Reference Mason2015).
Of course, a considerable number of forces has contributed to this state of American politics where outgroup hate is the binding force. Most prominent among them are social and political sorting. Ideological consistency has grown with affective polarization, suggesting that such sorting entails hearing more negative cues about the outparty (Levendusky and Malhotra, Reference Levendusky and Malhotra2016). Moreover, Americans now have fewer cross-cutting identities, which may also entail fewer inducements to see the other side favorably (Mason, Reference Mason2018). Of course, we can see such dynamics in American religion, where political disagreement in various forms has motivated people to leave congregations (Djupe et al., Reference Djupe, Neiheisel and Sokhey2018) or even leave religion entirely (Hout and Fischer, Reference Hout and Fischer2002; Margolis, Reference Margolis2018).
Just as ingroup love does not actively work alongside outgroup hate all the time, threat is not a constant in all situations. Threat is not ever present and can be measured explicitly, which is a matter we turn to next.
Victimhood as a tool of Christian Nationalism
Victimhood, or the belief that one is treated unfairly in life, has seeped into every corner of politics. This has led Horwitz (Reference Horwitz2018) to claim that, “The victim has become among the most important identity positions in American politics” (533). Victimization in American politics works when a minority group, or even a virtuous majority, is the recipient of some sort of (perceived) evil (Horwitz, Reference Horwitz2018). Perry (Reference Perry2023) applies a theoretical angle to this in regards to Christian nationalism, seeing that victimhood of Christian nationalism is not a static force in a vacuum, but rather activated by interested parties. He sees that there are three stages of the mobilization process, the mating call, the dog whistle, and the trigger. These revolve around turning the victimhood beliefs into actual mobilization, which religion and politics scholars have also theorized and found evidence to support (Armaly et al., Reference Armaly, Buckley and Enders2022; Djupe, Reference Djupe and Rudolph2022; Djupe et al., Reference Djupe, Lewis and Sokhey2023).
The concept of victimhood regularly occurs in religion and politics scholarship, centering on ingroup and outgroup dynamics. Scholars studying Christian nationalism have remarked on how Christian nationalists feel threatened by those outside their group, thus creating their own victimhood. This can be seen in several corners of Christian nationalist research, such as the perceived threat posed by immigrants and scientists (e.g., McDaniel et al., Reference McDaniel, Nooruddin and Shortle2011; Baker et al., Reference Baker, Perry and Whitehead2020b). This perceived victimhood has also stretched into current global developments, such that Christian nationalists felt threatened and a sense of victimhood from COVID-19 lockdown policies (Lewis and Bennett, Reference Lewis, Bennett, Djupe and Friesen2023; Wilson, Reference Wilson, Djupe and Friesen2023). Some went so far as to suggest that COVID restrictions showed the “real agenda of our opponents: to keep our churches shut down indefinitely and attack religious freedom in America” (First Liberty, 2020). While not every Christian nationalist study touches on victimhood or threat, there are at least a handful that have utilized it to help provide mechanisms driving the results.
While victimhood has helped explain the results of some Christian nationalist research, there has been little research explicitly measuring victimhood. One of the only papers that has sought to actively measure victimhood along with Christian nationalism was Armaly et al. (Reference Armaly, Buckley and Enders2022; see also Armaly and Enders, Reference Armaly and Enders2022), where victimhood helped explain support for the January 6th insurrection and political violence. Djupe et al. (Reference Djupe, Lewis and Sokhey2023) draw on a different measure that we draw on here capturing anticipated political persecution that helped to mobilize Christian nationalists in recent elections.
Despite this small pool of literature, it does provide some useful insights into how most scholars believe Christian nationalism operates. Generally speaking, scholars have noted that Christian nationalism is a stable disposition that has been especially prominent over the course of the last two decades (e.g., Gorski, Reference Gorski2017; Djupe et al., Reference Djupe, Lewis and Sokhey2023). However, this is not to say that threat is also stable and omnipresent; rather, depending on the messaging from elites and media, the perceived threat level and the source of the threat can change. For example, during the 2016 Presidential campaign, Donald Trump alluded to a threat of Muslim immigrants, going so far as to call for a “total and complete shutdown of Muslims entering the United States” (Johnson, Reference Johnson2015). By May 2017, that language was gone from the Trump campaign website (Barbash, Reference Barbash2017).
Along these lines, we expect that when given the opportunity to treat groups in a discriminatory fashion, Christian nationalism as typically measured will be linked to favorable treatment of the ingroup (e.g., religious conservatives) and does not necessarily entail maltreatment of outgroups. It takes the presence of threat against Christians to trigger discrimination against outgroups, though it may not entail better treatment of ingroups.
Data and methods
The data utilized in this study result from 2,020 American adults surveyed through Qualtrics Panels in mid-September 2021. We used quotas for age, region, and sex to return a sample that looks like the population relative to Census estimates (for descriptive results the remaining imbalances were corrected with raking weights). This is not a probability sample. Nevertheless, research using data from panel suppliers has found relationships that approximate those obtained from data obtained from probability samples (Berinsky et al., Reference Berinsky, Huber and Lenz2012; Lewis et al., Reference Lewis, Djupe, Mockabee and Wu2015; Mullinix et al., Reference Mullinix, Leeper, Druckman and Freese2015; Coppock et al., Reference Coppock, Leeper and Mullinix2018).
We employ an experiment with four treatments to assess whether various groups in society are treated equivalently. One of the four was randomly assigned to each respondent. In each scenario, a man is pulled over for speeding. When confronted, he explains why he was speeding, which was linked to a group activity, two of which were religious and two secular. Each is introduced by, “A man in his car is pulled over for speeding. He explains to the officer that…
• he was speeding because he is late for worship at his Baptist church [Baptist].
• kickoff for a huge college football rivalry is soon and he needs to buy a 6 pack before it begins [Football].
• he was speeding because he is late for Jummah prayers at his Mosque [Muslim].
• he was speeding because he is running late for a lecture on veganism he has been looking forward to [Vegan].”
Respondents were then asked, “Should he be given a warning or a ticket?” The survey respondents were given four choices for how to respond: definitely a ticket, probably a ticket, probably a warning, and definitely a warning. We decided to dichotomize the options to reduce definitely a ticket and probably a ticket (coded as a 0) and definitely a warning and probably a warning (coded as a 1). Neither warnings nor tickets are evidence of violations of the rule of law themselves. That evidence comes with inconsistency—treating some groups differently than others.
The Baptist, Football, Muslim, and Vegan treatments were chosen to test how those who lean toward Christian nationalism would respond to those perceived inside their group and those outside their group. We considered Baptists and perhaps the football case to be ingroup members. The last two treatments, Muslim and Vegan, were added to provide outgroups. Like Baptists, Christian nationalism's negative attitudes toward Muslims are clearly established (McDaniel et al., Reference McDaniel, Nooruddin and Shortle2011; Shortle and Gaddie, Reference Shortle and Gaddie2015; Sherkat and Lehman, Reference Sherkat and Lehman2018). We chose vegans as they are another likely liberal outgroup but without obvious religious ties.
Moderator variables
There were two moderator variables of interest in our experiment: the first was the Christian nationalist index. Following the second wave of the Baylor Religion Survey (see Whitehead and Perry, Reference Whitehead and Perry2020b), we utilized six questions to gauge how Christian nationalist a survey respondent was:
(1) The federal government should declare the United States a Christian nation
(2) The federal government should advocate Christian values
(3) The federal government should enforce strict separation of church and state [reverse coded]
(4) The success of the United States is part of God's plan
(5) The federal government should allow prayer in public schools
(6) The federal government should allow the display of religious symbols in public spaces
Each question had the options of strongly agree, agree, neutral, disagree, and strongly disagree. The questions were coded on a 1 to 5 scale (strongly agree = 5), which were added and then reduced to a 0–1 scale.
While the vast majority of scholars utilize this six-question scale to create the Christian nationalist index, there is some pushback in the literature regarding the effectiveness of these six questions. Following Davis (Reference Davis2023), we decided to examine the model utilizing a four-question (questions #1–4) and two-question (questions #1–2) Christian nationalism index.Footnote 1 There were few substantive differences in our results between the three versions (see Figure A1), which prompts us to follow precedent and continue utilizing the six-question Christian nationalist index.Footnote 2
The second moderator variable utilized was belief in Christian persecution, for which we utilized four questions:
Christians will lose their religious freedom if Democrats control the federal government.
A Democratic President is likely to ban the Bible.
A Democratic President will force you to pay for abortions.
A Democratic President will take your guns.
Each question had the options of strongly agree, agree, neutral, disagree, and strongly disagree. The questions were coded on a 1–5 scale (strongly agree = 5), which were added and then reduced to a 0–1 scale. We used these items because they capture threat, but also the particular form of it that was in wide airing in the run up to the 2020 election. Democrats were specifically identified as the source of threat to Christians by conservative elites, including the president himself.Footnote 3
While the Christian nationalist index and the belief in Christian persecution are used independently of each other in this study, they are highly correlated (r = 0.61). Figure 1 displays the correlation between the Christian nationalism index and the Christian persecution index, with histogram displays of each variable on the top and right side.
Relationship between Christian nationalism index and Christian persecution index.
Source: September 2021 Study.
Randomization
The treatments were effectively randomized. First, roughly equal numbers received each treatment (Baptist = 411, Football = 396, Muslim = 404, Vegan = 401). Second, we assessed whether equivalent distributions of the sample were presented with each treatment. ANOVA tests were run using five demographic variables: gender, education, race, region, and age. All five tests were insignificant (see Table A1), suggesting that the treatments were randomized properly. Still, we use eight controls in the model (education, gender, age, race, region, religious attendance, religious tradition, partisanship, and Christian nationalism for belief in Christian persecution and vice versa—see the Appendix for variable coding) and their inclusion does not change the estimated treatment effects.Footnote 4
Results
Our focus will be on the conditional effects of the treatments, but it is worth quickly reviewing the sample-level findings. Survey respondents were more likely to give the speeder a ticket for all four treatments with the sample treating other groups differently with a sharp drop off in likelihood to give a warning. The highest was the Muslim who was late for prayers—34.4% of respondents gave him a warning. This continued with the Baptist at 24.3%, Vegan at 20.3%, and the fellow out for a beer run before the big game—only 12.6% of respondents said he should receive a warning.
These results suggest that people prioritize certain actions over others. Understandably, the football control received the lowest amount of support since the person who was speeding had no legitimate justification in doing so. On the other hand, the Baptist and Muslim treatments feature a moral obligation that many could relate with or at least understand why they would be in a rush.
When looking at the two religious treatments, Baptist and Muslim, it appears that the speeding Muslim treatment received more leeway than the Baptist in a hurry. Respondents who received the Muslim treatment selected warning 34.4% of the time, compared to the Baptist treatment where people were more harsh and only selected warning 24.3% of the time. These results show some contrast with the conventional understanding of how the American population views Evangelical Christians and Muslims. For instance, research from the Pew Research Center in 2019 finds that out of all religious groups, Americans view Evangelical Christians at a warmth of 56 out of 100, while Muslims are viewed the most negatively, at 49 out of 100 (Figure 2).
Sample-level responses to treatment.
Source: September 2021 Survey.
Conditional on Christian Nationalism?
In order to see if Christian Nationalism affects the equal application of the rule of law, we begin our analysis with an OLS regression model that interacts the treatments with the Christian nationalism index, while controlling for partisanship, gender, education, age, race, and region (Table A4 presents the full model results). Figure 3 represents the marginal effects of the Christian nationalist index—the change in probability of providing a warning from the least Christian nationalist to the most Christian nationalist. A positive effect means those who hold Christian nationalist beliefs are more likely to provide a warning. Two of the treatments—Muslim and Vegan—did not yield any significant differences at the 90 or 95% confidence level, while respondents who sway toward Christian nationalism were more likely to give warnings in the Football (24.4% more often) and Baptist cases (30.1% more often). We regard these two as the most likely ingroups for Christian nationalists and they are treated with greater lenience.
The marginal effect of Christian nationalism by treatment.
Source: September 2021 Survey.
Note: Wider errorbar indicates 95% confidence intervals and narrower errorbar indicates 90% confidence intervals. The model also included race, partisanship, gender, education, age, region, religious tradition, and belief in Christian persecution.
These results raise questions regarding the Muslim treatment. Past research has found hostility of Christian nationalists toward those outside their group, specifically Muslims (Sherkat and Lehman, Reference Sherkat and Lehman2018). However, our initial results do not align with these findings—those who hold Christian nationalist beliefs are no more likely than those who do not to support giving a ticket to them. Of course, this is in context of controlling for beliefs in Christian persecution, which we believe will trigger outgroup intolerance, a possibility we examine next.
Conditional on belief in Christian persecution?
We ran the same model again utilizing the treatments and all controlling variables, but in this model, we interacted the treatments with our Christian persecution measure, with Figure 4 showing the marginal effects. We found that the Baptist and Football treatments did not produce any significant differences when comparing the behavior of those with the lowest belief in Christian persecution versus those with the highest belief in Christian persecution. However, the Muslim and Vegan treatments showed that as belief in Christian persecution increased, they were more likely to support giving this person a ticket by 12–21%.
The marginal effect of the belief in Christian persecution by treatment.
Source: September 2021 Survey.
Note: Wider errorbars indicate 95% confidence intervals and narrower errorbars indicate 90% confidence intervals. The model also included race, partisanship, gender, education, age, region, religious tradition, and Christian nationalism.
These results indicate that adding belief in Christian persecution provides considerable refinement in our understanding of how Christian nationalism is linked to the equal application of the rule of law. When Christian persecution is controlled, Christian nationalism is an expression of how the ingroup should be treated—it boosts ingroup love as seen in the increase in warnings given to Baptists and the Football character. Christian persecution, on the other hand, is a belief in the threat posed by outgroups and hence fuels outgroup hate—increases in tickets to Muslims and Vegans. Together, Christian nationalism and the perceived threat of forthcoming Christian persecution form a powerful package that undermines the rule of law.Footnote 5
Discussion
The results of the Muslim treatment, especially within the interaction with Christian nationalism, are worth examining in more detail. The literature predicts that we would see higher antipathy toward Muslims in the general population (Pew Research Center, 2019), and especially from Christian nationalists (McDaniel et al., Reference McDaniel, Nooruddin and Shortle2011; Shortle and Gaddie, Reference Shortle and Gaddie2015; Sherkat and Lehman, Reference Sherkat and Lehman2018). However, a higher percentage of survey respondents wanted to give the Muslim driver a warning than the Baptist, and there were no significant differences in warning/ticket choice between those who hold Christian nationalist beliefs and those who don't in the Muslim case.
To check the robustness of these results and ensure it was not an anomaly, we looked at the relationship between Christian nationalism and a feeling thermometer toward Muslim in the same survey (asked before the treatments). Controlling for variables that could potentially influence feeling toward Muslims (including # of Muslim friends, belief that religious freedom is threatened in the US, and opinion on the freedom to build a Mosque in your community), we see in Figure 5 that those who subscribe to Christian nationalism have a much cooler feeling toward Muslims in all four treatment groups. The marginal effect of Christian nationalism was actually a bit greater (i.e., more negative) in the Muslim treatment group, though not significantly. So, the treatment effect is not the result of more positive views that happened to accrue to the Muslim treatment group.
The marginal effect of Christian Nationalism on feelings towards Muslims, by treatment.
Source: September 2021 Survey.
Note: Wider errorbar indicates 95% confidence interval and narrower errorbar indicates 90% confidence interval. The model also included race, partisanship, gender, education, age, region, religious tradition, number of Muslim friends, belief that religious freedom is in danger, and opinion on the freedom to build a Mosque in your community.
In order to explain this break between the treatment effects and feeling thermometer data, we propose three possible explanations. First, those who hold Christian nationalist beliefs are trying to become more socially desirable. As we know from Figure 1, many Christian nationalists believe that they are being persecuted by those outside their group. In order for them to win over those who are persecuting them and stop a perceived attack on their beliefs, they may alter some of their actions to become more socially desirable in their opponents' eyes. That may be what is happening here—those with Christian nationalist beliefs are less harsh to Muslims than expected in order to save face.
Second, there may be a flaw in the wording of our question. While the failure could happen at any point, we used the phrase “Jummah prayer” in the Muslim treatment and just “worship” in the Baptist treatment. This may cause respondents to assign higher importance to the “Jummah prayer,” believing it may be more important than a standard worship session and thus giving more leniency to the Muslim treatment.
Third, those who believe in Christian nationalism may not react negatively to Muslims unless actively threatened (Bail, Reference Bail2016). In line with our observations regarding the ingroup and outgroup responses to Christian nationalism and belief in Christian persecution, the explanation for the difference may be because those with Christian nationalist beliefs will not actively seek to harm Muslims unless they are threatened. This could help to explain why we saw the higher marginal rate of ticketing Muslims when belief in forthcoming Christian persecution was higher. A final dynamic to consider is that forces that make Christian nationalism salient (e.g., through priming) can encourage adherents to consider religious minorities as allies (Lewis and McDaniel, Reference Lewis and McDaniel2023), especially if threats are generally concentrated among the non-religious.
Conclusion
A long line of literature has demonstrated that those who hold Christian nationalist beliefs are highly group oriented, with a strong preference for the ingroup and animosity for those outside their group. That is, Christian nationalists are willing to express prejudice and take prejudiced policy stands negatively affecting those outside their groups. Furthermore, previous work has expressed an interest regarding the interaction between Christian nationalism and support for rule of law (Davis, Reference Davis2018; Perry et al., Reference Perry, Whitehead and Davis2019), even if it is not tested explicitly.
We extend these studies by looking at Christian nationalism in an experimental setting that makes the application to the rule of law explicit in a controlled setting where all groups are effectively in the same car. In doing so, we find support for the general thrust of the Christian nationalism literature. Our results show that Christian nationalism is a predictor of ingroup love and not outgroup hate, while belief in Christian persecution presents a threat which activates outgroup hate but does not clearly signal ingroup love. In nature, of course, these two are strongly correlated so that most Christian nationalists perceive that Christians are (or will be) persecuted. However, that relationship is not near perfect, it reflects a widespread campaign that will not always be in evidence, and an independent measure of victimhood/persecution is worth capturing.
The core finding of our experiment is that Christian nationalism and belief in Christian persecution are complementary. Qualitatively, it makes conventional sense that those who believe that Christians are going to lose their religious freedom if Democrats take over are also willing to believe the federal government should declare the United States a Christian Nation. Quantitatively, this is supported with a high Pearson's correlation between the Christian persecution index and the Christian nationalist index (r = 0.61). With that in mind, the potential separability of Christian nationalism and belief in Christian persecution becomes important when looking long term at attitudes toward those outside and inside their group.
This reinforces our argument that capturing belief in Christian persecution (or victimhood or threat) is valuable to the literature exploring Christian nationalism (see also Armaly et al., Reference Armaly, Buckley and Enders2022; Djupe, Reference Djupe and Rudolph2022). It introduces, if not a new theoretical dimension, a new empirical dimension to the Christian nationalism literature: threat. Threat can constitute a dynamic element to help explain the surge and decline of Christian nationalism as a political force, reflecting political will and mobilization of voters.
That is, Christian nationalism and belief in Christian persecution are likely not equally stable. We suspect that Christian nationalist's preference for the ingroup is a disposition that will remain across time (Gorski, Reference Gorski2017; Djupe et al., Reference Djupe, Lewis and Sokhey2023). However, perceived threat (i.e., belief in Christian persecution) is not a stable worldview and persecution beliefs are mobilized, lately as part of a campaign to drum up support for the GOP. In this way, we suggest that future work continue to monitor threat levels since they are likely the motive force that activates Christian nationalist views into politics, or toward any domain from which threat is perceived.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1755048323000305.
Competing interests
None.
Zachary D. Broeren is a PhD student in the political science department at Vanderbilt University. His research lies in the American Political Behavior subfield, with a focus on political information, political psychology, public opinion, participation, and religion and politics.
Paul A. Djupe is a professor and director of the Data for Political Research program at Denison University. He edits the Religious Engagement and Democratic Politics series at Temple UP, is the coauthor of The Full Armor of God: The Mobilization of Christian Nationalism in American Politics (2023, Cambridge), and the coeditor of “An Epidemic Among My People:” Religion, Politics, and COVID-19 in the United States (2023, Temple).
Open access
