THE GREAT REPLACEMENT THEORY: CONTEXT, CONTENT, AND ELITE ENDORSEMENT

Social-identity theory posits that people gain considerable benefit from belonging to a social group, including acceptance, social support, and a system of shared values and norms (Tajfel and Turner Reference Tajfel, Turner, Worchel and Austin1986). According to this theory, due to the psychological and material benefits of group membership, individuals “tend to favor [their] own group and exhibit hostility toward other groups, especially during dangerous or contentious times” (Stephan, Ybarra, and Morrison Reference Stephan, Ybarra, Morrison and Nelson2015). Building on these insights, proponents of intergroup-conflict theory argue that when members of a group perceive threats from another group to their economic interests, political power, and social values, they take decisive economic, political, and cultural steps to reassert their interests (Renfro et al. Reference Renfro, Duran, Stephan and Clason2006; Stephan and Stephan Reference Stephan, Stephan and Oskamp2000). These efforts often are destructive, which intensifies intergroup tensions (Stephan, Ybarra, and Morrison Reference Stephan, Ybarra, Morrison and Nelson2015).

These theoretical insights provide a lens for understanding the emergence of the GRT. During the past several decades, large-scale immigration—particularly from Latin America—has profoundly altered the demography, culture, and politics of the United States (Kramer and Passel Reference Kramer and Passel2025). From the perspective of intergroup-conflict theory, we would expect these transformations to generate resistance from other groups in American society—especially whites, the nation’s traditionally hegemonic group. From the perspective of intergroup-conflict theory, the GRT is an ideational framework that identifies the “threats” posed by immigration to white economic and political interests and cultural values; provides an explanation of the origins of this threat; and implicitly endorses an agenda (i.e., draconian immigration policies) that whites can rally behind to reassert their dominance. The GRT thus provides a template for collective action by aggrieved whites to attempt to reclaim political dominance.

The GRT is a belief system, which originated in right-wing French political thought, that asserts that native (white) residents of the United States are being replaced demographically, culturally, and politically by foreigners due to permissive immigration policies supported by political and economic elites (Bauder Reference Bauder2022; Obaidi et al. Reference Obaidi, Kunst, Ozer and Kimel2022; Rose Reference Rose2022). As noted by scholars, the GRT comprises five core beliefs: (1) the United States is facing an invasion of immigrants from other nations (particularly non-European nations); (2) this influx is producing the demographic replacement (and potential elimination) of native (white) citizens by those from other nations and cultures; (3) the rapid growth of the immigrant population threatens native (white) culture; (4) the expansion of the immigrant population threatens the political power of native (white) citizens; and (5) permissive immigration policies represent an elite conspiracy to replace native (white) citizens with a dependent caste of immigrant voters who will maintain elites in political power (Becker and Jipson Reference Becker and Jipson2022; Bracke and Hernández Aguilar Reference Bracke, Aguilar, Bracke and Aguilar2023; Camus Reference Camus2012; Dennison and Kustov Reference Dennison and Kustov2025; Ekman Reference Ekman2022; Feola Reference Feola2021; Obaidi et al. Reference Obaidi, Kunst, Ozer and Kimel2022).

Of course, nativist and negative racial attitudes, fears of status threat, and concerns about the impact of demographic change have long been present in the United States and Europe (Craig and Richeson Reference Craig and Richeson2014; Mutz Reference Mutz2018; Parker and Barretto Reference Parker and Barreto2014; Schuman Reference Schuman1997; Sidanius et al. Reference Sidanius, Pratto, Van Laar and Levin2004). The GRT clearly is related to xenophobia—defined as an “attitudinal, affective, and behavioral prejudice toward immigrants and those perceived as foreign” (Reynolds and Vine Reference Reynolds and Vine1987; van der Veeret al. Reference van Der Veer, Ommundsen, Yakushko and Higler2011, 27; Yakushko Reference Yakushko2009). The GRT also is related to out-group prejudice, which often is measured as the difference between indicators for liking the in-group and for liking the out-group (Christ et al. Reference Christ, Schmid, Lolliot, Swart, Stolle, Tausch, Al Ramiah, Wagner, Vertovec and Hewstone2014).Footnote ¹

However, unlike xenophobia or out-group prejudice, the GRT emphasizes the existential threat posed by people of color and continued immigration to the very existence of whites (Camus Reference Camus2012). More pointedly, whereas xenophobia and out-group prejudice comprise negative attitudes toward immigrants and out-groups, respectively, the GRT combines a set of negative attitudes toward immigrants with an “explanatory” political theory about the causes of increased immigration: namely, the conspiracy theory that political elites advance permissive immigration policies to replace independent-minded whites with politically obedient immigrants who uncritically will support elites’ plans. As discussed previously, these fears have motivated some white Americans to act on these views and to use violence against the groups that are viewed as the main threats to their status (Obaidi et al. Reference Obaidi, Kunst, Ozer and Kimel2022).

In recent years, prominent Republican figures, including President Donald Trump, Vice President JD Vance, and House Speaker Mike Johnson, have endorsed ideas associated with the GRT (Milligan Reference Milligan2022; Ragland Reference Ragland2022; Rashid Reference Rashid2024). The conservative media ecosystem has had a critical role in circulating and amplifying these ideas associated with the GRT. At the center of this development is Fox News, the most important conservative media outlet and the single-most-utilized source of political news in the United States today (Hoewe, Brownell, and Weimer Reference Hoewe, Brownell and Wiemer2021). GRT ideas have been endorsed by popular Fox News hosts, including Sean Hannity, Jesse Watters, and Laura Ingraham—all of whom have top-five cable news programs (Hodil and Simon Reference Hodil and Simon2024). Until his firing by Fox News, Tucker Carlson hosted the most-watched cable news program on television, providing extensive coverage that endorsed GRT ideas (Confessore Reference Confessore2022). According to a New York Times analysis, from 2016 to 2022, Carlson reported on the demographic replacement of native citizens by undocumented immigrants in more than 400 episodes of Tucker Carlson Tonight (Confessore and Yourish Reference Confessore and Yourish2022).

The GRT is politically and socially consequential in contemporary American politics. As described in the online appendix, support for the GRT is correlated with important political outcomes, including increased warmth toward President Donald Trump, holding constant factors such as party affiliation, ideology, education, income, and racial attitudes. It is troubling that in an age of increasing political violence, our study also reveals that endorsement of the GRT is correlated with increased support for violence as a tactic in politics. Consequently, assessing whether exposure to Fox News is associated with increased support for the GRT will shed light on the influence of this news source on important dynamics in contemporary American politics.

HYPOTHESIS

What impact, if any, does exposure to elite rhetoric about the GRT have on white opinion? Scholarly explorations of public opinion largely concluded that, given the “rational ignorance” of the mass public, it is in significant part an “echo” of elite rhetoric on pertinent political issues (Carmines and Stimson Reference Carmines and Stimson1980; Zaller Reference Zaller1992). Rather than expand the cognitive energy necessary for researching and understanding political issues, ordinary Americans use the actions, statements, and interpretations of political elites as a heuristic to help them form their opinions. Copartisan elites and elites who are viewed as legitimate, trustworthy, and knowledgeable emerge as particularly influential in shaping mass opinion (Lenz Reference Lenz2012; Sears and Kosterman Reference Sears, Kosterman, Shavitt and Brock1994; Zaller Reference Zaller1992). More recent research exploring the boundaries of elite influence finds that elite opinion is most important when the issue is remote, complex, and abstract relative to issues for which the mass public has more extensive experience, interest, and knowledge (Bullock Reference Bullock2011; Druckman, Peterson, and Slothuus Reference Druckman, Peterson and Slothuus2013; Feldman, Huddy, and Marcus Reference Feldman, Huddy and Marcus2012; Zaller Reference Zaller1992). Of course, elite opinion does not determine the content of public opinion, but it can influence attitudes on important issues.

Recent research finds that exposure to conservative elite rhetoric via Fox News has a significant impact on mass opinion (Ash et al. Reference Ash, Galletta, Hangartner, Margalit and Pinna2024a, Reference Ash, Galletta, Pinna and Warshaw2024b; Ash and Poyker Reference Ash and Poyker2024; Broockman and Kalla Reference Broockman and Kalla2025; DellaVigna and Kaplan Reference DellaVigna and Kaplan2007; Earle and Hodson Reference Earle and Hodson2022; Hoewe et al. Reference Hoewe, Peacock, Kim and Barnidge2020; Martin and Yurukoglu Reference Martin and Yurukoglu2017). Using a wide array of methods, scholars have found that Fox News viewership has moved the electorate rightward (Ash et al. Reference Ash, Galletta, Pinna and Warshaw2024b; Earle and Hodson Reference Earle and Hodson2022); reinforced conservative predispositions among Republicans (Hopkins Reference Hopkins2014); and increased Republican vote share in elections (Ash et al. Reference Ash, Galletta, Pinna and Warshaw2024b; DellaVigna and Kaplan Reference DellaVigna and Kaplan2007; Martin and Yurukoglu Reference Martin and Yurukoglu2017). Likewise, watching Fox News influenced attitudes about the COVID-19 pandemic (Ash et al. Reference Ash, Galletta, Hangartner, Margalit and Pinna2024a); led to more punitive views regarding drug crime and sentencing decisions (Ash and Poyker Reference Ash and Poyker2024); and elicited support for restrictive immigration policies (Hoewe et al. Reference Hoewe, Peacock, Kim and Barnidge2020).

These studies suggest that exposure to Fox News is associated with increased support for conservative candidates, policies, and positions. However, does this influence extend to support for the GRT? We believe that it does. As noted previously, ordinary Americans are more likely to depend on elites for opinion guidance when the issue in question is abstract, remote, and complex (Bullock Reference Bullock2011; Druckman, Peterson, and Slothuus Reference Druckman, Peterson and Slothuus2013; Feldman, Huddy, and Marcus Reference Feldman, Huddy and Marcus2012; Zaller Reference Zaller1992). Given the multifaceted causes of increases in undocumented border crossings and the lack of knowledge concerning immigrants and the immigration system (Blinder and Schaffner Reference Blinder and Schaffner2020; Lutz and Bitschnau Reference Lutz and Bitschnau2023; Sidler, Knotz, and Ruedin Reference Sidler, Knotz and Ruedin2024), we expected that the mass public would rely on the opinions of political elites deemed trustworthy, legitimate, and knowledgeable to help them interpret immigration issues. Because Fox News is a pillar of the contemporary media environment (Broockman and Kalla Reference Broockman and Kalla2025), platforms hosts who publicize key tenets of the GRT, and invites prominent proponents of the GRT on its programs, we expected that exposure to it would be associated with greater support for the GRT among white respondents. Therefore, we hypothesized the following:

DATA AND MEASURES

To explore white support for the GRT and to test our hypothesis, we undertook both cross-sectional and panel analyses of Wave 3 and Wave 4 of the AMPS (Goldman et al. Reference Goldman, Huo, Nteta, Pérez and Tropp2025). Wave 3 was fielded online by YouGov from July 31 to August 23, 2024, with a total of 4,335 respondents, including 1,000 respondents who identified only as white. Wave 4 was fielded from June 11 to July 10, 2025, and interviewed 4,611 respondents, of which 1,030 identified exclusively as white. As described in our panel analyses, 533 white respondents in Wave 3 also participated in Wave 4 (Rhodes et al. Reference Rhodes, Goldman, Nteta, Eichen, Lapcheske, Tropp, Pérez and Huo2026). YouGov’s methodology uses propensity scores to create weights that adjust its samples to be nationally representative of American adults with respect to age, gender, race, years of education, and region.

Our dependent variable measures support for key tenets of the GRT. As described previously, the GRT is composed of beliefs concerning the invasion of the nation by immigrants, the loss of native-born (white) political and cultural power and status to these groups, the replacement (and potential) elimination of native-born Americans by these groups, and the presence of an elite cabal responsible for the replacement of native-born Americans by new immigrant groups. To capture these core sentiments, we asked respondents to indicate their agreement with the following statements: (1) “Immigrants invade and colonize the United States”; (2) “Native-born Americans are losing their economic, political, and cultural influence in this country because of the growing population of immigrants”; and (3) “There are people who secretly work to make sure immigrants will eventually replace real Americans.” The first statement measures respondents’ beliefs about “unchecked” immigration; the second assesses their beliefs about threats to the influence of “natives”; and the third measures the sentiment that elites are engaged in a conspiracy to exploit immigration policy to undermine the power of “natives.” Each statement was measured on a 5-point scale of “strongly disagree” to “strongly agree.” These statements were averaged into a single scale, with higher values indicating greater support for the GRT. An exploratory factor analysis revealed that responses to these questions all load on a single factor in both waves of the AMPS; moreover, the statements comprising this variable have a high reliability on Wave 3 (alpha=0.87) and Wave 4 (alpha=0.88).

We measured consumption of Fox News using two indicators: a dichotomous measure of whether respondents watch Fox News Channel and an interval-level measure of the number of Fox News Channel programs viewed. The dichotomous measure relied on a single question that asked respondents: “Which of the following television sources do you watch regularly?” and prompted them to select “any that you watch at least once a month” from a list of major news networks (e.g., Fox News Channel, MSNBC, and CNN). To facilitate recall, the question prompt for each network included a list of several prominent programs. Respondents who selected Fox News were coded as 1; those who did not select Fox News were coded as 0.

Respondents who stated that they watched Fox News were presented with a list of the network’s nine most-watched programs and asked to indicate which ones they watched regularly (defined as at least once a month): “Fox & Friends,” “America’s Newsroom,” “Outnumbered,” “The Five,” “Special Report with Bret Baier,” “Jesse Watters Primetime,” “Fox News at Night,” “Hannity,” and “The Ingraham Angle.” The interval-level measure of Fox News viewership is a count of the number of programs viewed, coded from 0 to 1. Political television exposure measured through the program-list technique demonstrated high true-score reliability and strong predictive, construct, and convergent validity, with widespread use, including on the American National Election Studies (Dilliplane, Goldman, and Mutz Reference Dilliplane, Goldman and Mutz2013; Goldman, Mutz, and Dilliplane Reference Goldman, Mutz and Dilliplane2013; Goldman and Warren Reference Goldman, Warren, Grofman, Suhay and Trechsel2020).

Our statistical models included the standard measures of party identification and ideology. We measured racial attitudes using three items from the Fear, Institutionalized Racism, and Empathy (FIRE) Scale (DeSante and Smith Reference DeSante and Smith2020). Each item was measured on a 5-point scale from “strongly disagree” to “strongly agree.” We also included controls for age, gender, educational attainment, and income. Question wording for all variables in the following models is included in the online appendix. All control variables were coded to vary between 0 and 1.

RESULTS: PREVALENCE OF BELIEFS ABOUT THE GRT AMONG WHITES

How prevalent are beliefs about the GRT among white Americans? Table 1 presents the weighted share of white respondents by their level of agreement with each of the statements in Wave 3 (July–August 2024) of the AMPS. As shown in table 1, there is substantial support for the major tenets of the GRT among white American adults: 41% strongly or somewhat agreed that “immigrants invade and colonize the United States”; 53% strongly or somewhat agreed that “native-born Americans are losing their economic, political, and cultural influence in this country because of the growing population of immigrants”; and 38% strongly or somewhat agreed with the conspiratorial belief that “there are people who secretly work to make sure immigrants will eventually replace real Americans.”

Table 1 Agreement with Tenets of the Great Replacement Theory Among All Whites and White Fox News Viewers, AMPS Wave 3

Whereas pluralities of white respondents expressed support for some tenets of the GRT, we found that among white viewers of Fox News, strong majorities exhibited agreement with these same key tenets of the conspiracy theory. Almost two thirds (64%) of white Fox News viewers agreed with the belief that “immigrants invade and colonize the United States,” and a similar percentage (62%) agreed with the statement that people are working to “eventually replace real Americans.” Finally, more than three quarters (76%) of Fox News viewers believed that “native-born Americans are losing their economic, political, and cultural influence in this country because of the growing population of immigrants.”

In Wave 4 (June–July 2025) of the AMPS, we found that white viewers of Fox News expressed stronger support for tenets of the GRT than whites overall. For example, whereas 35% of whites believe that “there are people who secretly work to make sure immigrants will eventually replace real Americans,” 63% of white Fox News viewers agreed with this statement.

Table 2 Agreement with Tenets of the Great Replacement Theory Among All Whites and White Fox News Viewers, AMPS Wave 4

RESULTS: CROSS-SECTIONAL ANALYSES OF WAVE 3 AND WAVE 4 OF THE AMPS

To assess more rigorously the relationship between Fox News viewership and support for the GRT, table 3 presents cross-sectional analyses of Wave 3 (July–August 2024) and Wave 4 (June–July 2025) of the AMPS. We estimated four ordinary least squares (OLS) regression models:

• • Model 1 includes demographic controls, measures of partisanship and ideology, racial attitudes, and our dichotomous measure of Fox News viewership in Wave 3.

• • Model 2 is identical to model 1 except that it substitutes our more fine-grained interval-level measure of Fox News viewership, called Fox News Total.

• • Models 3 and 4 present the same analyses as the first two models but use data from Wave 4.

Table 3 presents remarkably consistent findings across the four models. Individuals with more education were significantly less likely to support the GRT (p<0.01 in Wave 3; p<0.001 in Wave 4). Likewise, higher income was associated with significantly lower support for the GRT (p<0.01 in Wave 3; p<0.05 in Wave 4). Conservative ideology, by contrast, was correlated with greater support for the GRT (p<0.001 in both waves). Moving from the most liberal to the most conservative ideology was associated with a 0.22–0.23 increase in support for the GRT in Wave 3 and a 0.32 increase in Wave 4 in both model specifications. Similarly, Republican partisanship was associated with significantly greater support for the GRT in both waves. In Wave 3, moving from Strong Democrat to Strong Republican party affiliation was associated with a 0.18 increase in support for the GRT in both models 1 and 2 (p<0.001). In Wave 4, it was associated with a 0.09 (model 3, p<0.01) to 0.10 (model 4, p<0.01) increase.

Table 3 OLS Regression Models Predicting Support for the Great Replacement Theory Among White Adults, 2024 AMPS (Wave 3 and Wave 4)

Notes: These are weighted unstandardized coefficients. Standard errors are in parenthesis. *p<0.05, **p<0.01, ***p<0.001.

Negative racial attitudes, as measured by the FIRE Scale, were strongly associated with greater support for the GRT among whites in both waves (p<0.001 in all four models). Moving from the most positive to the most negative racial attitudes was associated with a substantial increase (0.44–0.48) in support for the GRT among white respondents—holding the other variables at their means—depending on the wave and model specification.

Was exposure to Fox News associated with greater support for the GRT beyond the many factors discussed previously? Across the two waves and different model specifications, Fox News was associated consistently and significantly with increased support for the GRT among whites. Model 1 (Wave 3) and model 3 (Wave 4) show that our dichotomous measure of exposure to Fox News was associated with a 0.07–0.10 increase in support for the GRT (p<0.001 in both models). Model 2 (Wave 3) and model 4 (Wave 4) reveal that our more fine-grained interval-level measure of Fox News viewership was associated even more strongly with increased support for the GRT. Indeed, the predicted effect of the number of Fox News programs viewed ranged from 0.14 (Wave 4, p<0.001) to 0.22 (Wave 3, p<0.001).

The online appendix (SI 4, SI 5, and SI 6) includes several robustness tests of our cross-sectional finding that viewing Fox News was associated with greater support for the GRT. These tests show that the cross-sectional results are robust to a wide range of model specifications; they hold when we used hierarchical entropy balancing to balance the data on the observables; and they are boundedly robust to unobserved confounding. These results strengthen our confidence that Fox News viewership is associated with increased support for the GRT net of other predictors.

RESULTS: PANEL DATA ANALYSIS

A limitation of our cross-sectional models is that they do not account for unobserved confounders and thus might produce biased estimates of the effect of Fox News viewership on support for the GRT. To address this problem, we leveraged the panel features of the AMPS and estimated fixed-effects models of within-person change from Wave 3 (July–August 2024) to Wave 4 (June–July 2025). Fixed-effects regression assesses within-person changes by comparing every respondent to themselves at an earlier point in time, thereby controlling for all unchanging individual-level characteristics (Allison Reference Allison1990, Reference Allison2009). We accounted for factors that might have changed over time with indicators of change in the control variables from Wave 3 to Wave 4. In addition, we accounted for the changing impact of baseline levels of the control variables by computing interactions between the Wave 3 control variables and the wave variable. Although the main effects of stable indicators automatically drop out with fixed effects, interactions between stable indicators and the wave variable evaluate the changing impact of these indicators (e.g., whether the impact of baseline education levels varies over time). Moreover, including the wave variable to represent the passage of time controls for the average total effects of all other factors that changed over time (i.e., period effects) (Halaby Reference Halaby2004). Because fixed-effects regression uses only within-person variance, accompanied by robust standard errors, these models provide highly conservative tests of our hypothesis and unusually strong causal evidence.

Within-person change in the independent variable provides the strongest control for spurious confounders (Allison Reference Allison2009); therefore, we focused our tests on examining the impact of within-person change in Fox News viewership. Not surprisingly, we found a positive but nonsignificant relationship between change in our dichotomous measure of Fox News and change in support for the GRT due to the limited variance in this exposure measure (see the online appendix, SI 7). However, we found consistent evidence using the interval-level measure of exposure. Models 1 through 4 in table 4 show the significant relationship between change in Fox News viewership and support for the GRT across increasingly saturated model specifications. Model 1 shows the effect of increases in Fox News viewership on increases in support for the GRT, controlling for period effects with the wave variable (0.13, p<0.05). Model 2 shows a similar effect of Fox News viewership with added controls for change in the viewership of six other television networks (0.14, p<0.05). Model 3 adds controls for change in levels of demographic variables and the changing impact of baseline demographic variables in 2024 (Wave 3), as well as 2024 support for the GRT to account for regression to the mean. The effect of Fox News is unchanged in Model 3 (0.14, p<0.05), which also shows that respondents who identified as Republican in Wave 3, men, and older respondents increased in their support for the GRT, whereas those with the highest levels of support for the GRT in Wave 3 decreased in their support, as expected. Finally, Model 4 adds controls for change in levels of FIRE and the changing impact of 2024 FIRE, and shows that respondents who became more negative over time in their racial attitudes or who initially had more negative racial attitudes increased in their support for the GRT; again, the effects of Fox News remain unchanged (0.14, p<0.05). In summary, these stringent fixed-effects models provide remarkably consistent evidence that watching Fox News contributed to increases over time in support for the GRT from 2024 to 2025.

Table 4 Effects of Within-Person Change in the Number of Fox News Channel Programs Viewed on Within-Person Change in Support for the Great Replacement Theory, AMPS Wave 3 and Wave 4 (OLS Fixed-Effects Panel Models)

Notes: Unstandardized fixed-effects regression coefficients, with standard errors in parentheses. All variables range from 0 to 1. #p<0.10, *p<0.05, **p<0.01, ***p<0.001.