In the United States, the COVID-19 pandemic was marked by the creation and spread of misinformation (Motta, Stecula, & Farhart, 2020). At the same time, another American misinformation crisis was unfolding – this one related to the 2020 Presidential Election, featuring the “Big Lie,” President Trump’s false claim that the election had been riddled with fraud and that he, not President Joseph Biden, had won the election (Jacobson, 2021). In the context of both COVID-19 and the 2020 Election, belief in misinformation was higher on the political right than the left (Pennycook & Rand, 2021; Uscinski et al., 2020), with devastating consequences. While there are certainly political reasons to explain why Republicans – and Trump supporters in particular – would be more likely to believe misinformation related to both COVID and the Election, it is also possible that belief in misinformation may be generally higher among conservatives than liberals – across topics and across contexts (Brulle et al., 2012; Garrett & Bond, 2021), a divide that may stem from the very different ways of knowing that are valued by the right and the left (Deppe et al., 2015; Ruisch & Stern, 2021). The goal of the current project is to untangle the relationships between political preferences (ideology, party, and support for President Trump), epistemic motivations (placing value in intuition and emotions, or evidence and data) and belief in misinformation related to COVID-19 and the 2020 election.
Using survey data obtained in November-December 2020, (ages 18 – 48 [N = 475], ages 50 + [N = 1304]), we first test the relationships between political preferences and belief in misinformation related to COVID and the election. We expect misperceptions about COVID and the 2020 Election to be positively associated with: (a) Republicanism, (b) conservatism, and (c) Trump favorability. Then, we build upon work of Garrett and Weeks (2017) to assess to what extent belief in COVID and Election misinformation was related to epistemic motivations. We finally propose how and why these epistemic motivations should correlate with ideology and partisanship and consider how Donald Trump’s populist appeal may reflect a distinct manifestation of epistemic motivations rooted in intuition and emotions, and a rejection of evidence and data. Through regression analysis, we demonstrate that individuals who valued intuition and emotions were more favorable towards President Trump, while those who valued empirical evidence and data were less favorable towards Trump and more likely to be liberal and Democratic. Finally, we examine how conservatism, Republicanism, and Trump favorability might help account for the relationships between epistemic motivations and misperceptions of COVID and the 2020 Election.
Misinformation and Political Preferences
Scholars have identified partisan gaps in crucial preventative health behaviors related to COVID, with Republicans less likely than Democrats to engage in masking, social distancing, and limited travel (Allcott et al., 2020; Gadarian, Goodman, & Pepinsky, 2021; Grossman et al., 2020; Rothwell & Desai, 2020). Republicans were also less likely to receive the COVID vaccine than Democrats (Kirzinger et al., 2021), a gap that likely contributed to the significantly higher COVID death rates within Republican-leaning counties in the later phases of the pandemic (Chen & Karim, 2021). Perhaps unsurprisingly, belief in the “Big Lie” was also higher among Republicans than Democrats, even long after the insurrection attempt on the Capitol on January 6, 2021 (Arceneaux & Truex, 2021). Two months after President Biden won the election, 75% of Trump voters reported believing that Donald Trump had “probably” or “definitely” won the election, a claim that was demonstrably false (Pew Research Center, 2021). Our first hypothesis, thus concerns the relationships between political preferences and belief in misinformation related to COVID and the election:
H1: Misperceptions about COVID and the 2020 Election will be positively associated with: (a) Republicanism, (b) conservatism, and (c) Trump favorability.
Misinformation and Epistemic Motivations
Belief in and sharing of misinformation is related to numerous individual characteristics and psychological traits, including need for cognition (Leding & Antonio, 2019) and “cognitive reflection” (Frederick, 2005; Pennycook et al., 2016). Also related to belief in misinformation is the concept of epistemic beliefs, or beliefs about the origins of knowledge (Garrett & Weeks, 2017; Swami et al., 2014). Epistemic beliefs include the extent to which an individual trusts intuition as a pathway to truth, and how important it is that one’s beliefs comport with empirical evidence (Epstein et al., 1996). As conceptualized by Epstein et al. (1996), these “two kinds of processing are not opposite equivalents but represent two kinds of information processing that are independent” (p. 401). Individuals can be high in one or the other, both, or neither. Belief in misinformation and conspiracy theories is more likely among those who tend to trust their intuition and gut instinct, and less likely among those who adjust their beliefs based on new evidence and data (Garrett & Weeks, 2017). This theoretical framework is consistent with recent interventions that have successfully reduced belief in misinformation by incentivizing accuracy (and reflection) over more reflexive – and hence intuitive – judgments (Pennycook & Rand, 2019; Pennycook et al., 2021). Hence we hypothesize that:
H2: Misperceptions about COVID and the 2020 Election will be (a) positively associated with epistemic motivations based in feelings and intuition, and (b) negatively associated with epistemic motivations based in evidence and data.
Epistemic Motivations, Political Preferences, and Populism
According to the motivated social cognition framework (Jost et al., 2003), for reasons related to biology, physiology and genetics, people have interpersonal threat-monitoring systems on varying levels of alert, which shape psychological needs that contribute to our social and cultural attitudes and beliefs (Hibbing, Smith, & Alford, 2013). This work highlights asymmetries in the needs and urges that shape how liberals and conservatives think and behave, both in terms of existential needs relating to threat management, and epistemic needs related to how individuals come to know the things they know (Federico & Ekstrom, 2018; Jost, 2017; Jost, Federico, & Napier, 2013; Jost, Sterling, & Stern, 2017). Social and cultural conservatives are high threat monitors, so are motivated by efficiency and more likely to form judgments based on emotions and intuition (Deppe et al., 2015; Oliver & Wood, 2018). In contrast, those who are less acutely aware of physical threats, have higher tolerance for ambiguity and need for cognition. These needs contribute to more socially liberal attitudes, while also resulting in greater reflection when faced with empirical evidence (Deppe et al., 2015; Ruisch & Stern, 2021).
Crucially, these relationships are most pronounced in the context of social and cultural ideology – beliefs related to gender, sexuality, race and crime (Deppe et al., 2015; Saribay & Yilmaz, 2017; Yilmaz & Saribay, 2017). In the American political context, the two major parties have experienced significant polarization along this social/cultural dimension (Hare & Poole, 2014), with the Republican party holding more conservative positions on social issues and Democrats holding more liberal ones. At the same time, the parties have also become “socially sorted,” or increasingly distinct in terms of party members’ race, religion, geography, and culture (Mason, 2018). Given the documented associations between social ideology and epistemic traits, we would expect the American trends of polarization and social sorting to be coupled with a growing epistemic divide as well. Indeed, as the Republican party became the dominant home to white evangelical Protestants starting in the 1960’s (Layman, 2001), scholars have noted a widening gulf between the epistemic beliefs of Democrats and Republicans, with Republicans more likely to be guided by intuition and faith, and less likely to be guided by evidence and data (Oliver & Wood, 2018). Additionally, since the 2016 election of Donald Trump, scholars have observed that his support was driven by social and cultural conservatism (Mason, Wronski, & Kane, 2021; Mutz, 2018; Norris & Inglehart, 2019; Sides, Tesler, & Vavreck, 2019). Considering the link between social ideology and epistemic beliefs, the socially conservative underpinnings of Trump support would suggest a voting base that is especially prone to valuing intuition and feelings while downplaying the importance of data and evidence.
Trump’s support also exhibited the characteristics of a populist movement, as it championed the authenticity of “the pure people” while criticizing the “corrupt elites” (Fieschi, 2019; Mudde, 2004). In lieu of policy, it is the rhetoric of populism, with its adulation of “the people,” their homogeneity, and their moral distance from the immoral elites, that is its defining characteristic (Jagers & Walgrave, 2007). In solidarity with “the people,” populists often appeal to “common sense” (Betz & Johnson, 2004) or gut instinct (Mede & Schäfer, 2020). Through its rejection of intellectualism and elevation of group affiliation and symbolism over rational discourse, Green (2019) argues that populism serves as an efficient vehicle for “bullshit” – defined by Frankfurt (2009) as statements uttered with no concern as to their truth value. Trump’s rhetoric exemplified this populist epistemology, frequently highlighting his faith in instinct coupled with a suspicion of experts (Keith, 2020; Nichols, 2019). In 2018, he told the Washington Post, “my gut tells me more sometimes than anybody else’s brain can ever tell me” (Rucker, Dawsey, & Paletta, 2018, p. 27). We anticipate that this combination of Trump’s populist appeal with the trends of social sorting and social/cultural polarization results in the following:
H3: Republicanism, conservatism and Trump favorability will be (a) positively associated with epistemic motivations based in feelings/intuition, and (b) negatively associated with epistemic motivations based in evidence/data.
Political Preferences, Epistemic Motivations, and Misinformation
These relationships have important implications for how liberals and conservatives engage with misinformation (Baptista & Gradim, 2020; Guess, Nagler, & Tucker, 2019). While social conservatives’ reliance on instinct might make them quick decision-makers, it would likely reduce the cognitive reflection that would safeguard them from misinformation (Pennycook et al., 2021). And while social liberals’ reconsideration of beliefs in the face of evidence might delay decision-making (Ruisch & Stern, 2021), these processes likely increase analytical thought (Talhelm et al., 2015), thereby reducing susceptibility to misinformation (Pennycook & Rand, 2019). There is also a potential synergy between the anti-intellectualism inherent in Trump’s populist appeal and the intuition-based epistemic motivations of his socially conservative supporters. In fact, long before COVID or the 2020 Election, scholars found higher conspiracy theory beliefs among Trump supporters (Enders & Uscinski, 2021), and observed that support for Trump did not waver even when people learned – and believed – that his statements were untrue (Swire et al., 2017).
Political psychologists tend to conceptualize psychological and epistemic traits as underlying characteristics that shape how we engage with the world, thus influencing political preferences (Hibbing, Smith, & Alford, 2013). Conceptualized this way, political preferences are indirect consequences of valuing instinct and rejecting evidence. If this were true, then political preferences would help account for (e.g., mediate) some of the statistical associations between epistemic motivations and belief in misinformation.
H4: (a) Republicanism, (b) conservatism, and (c) Trump favorability mediate the relationships between epistemic motivations and misperceptions of COVID and the election.
These relationships were examined within a dataset collected from a national survey of U.S. adults 18 and over administered by SSRS. Two separate samples were pulled from the population to serve broader project goals, one sample ages 18 – 49 and one ages 50 and older. Data were collected at two time points: from Oct 20 – Nov 2, 2020 (Time 1) and Nov 26 – Dec 4, 2020 (Time 2) for each group. 52% of Wave 1 respondents also completed the survey at Wave 2 [18-49 (Unweighted N at Wave 1 = 1007, Unweighted N at Wave 2 = 475); 50+ (Wave 1 N = 2411, Wave 2 N = 1304)]. Within each age group, oversamples from racial and ethnic minority populations were drawn to facilitate comparisons. For the present study, data are weighted (within each age group) to match general population demographics and are treated as cross sectional. The measures section reports weighted descriptive statistics for each sample. Approval was obtained by the Institutional Review Board of the sponsoring institution. This dataset was also used in a separate study of the relationship between political ideology, conflict orientation, reactance, and self-reported mask-wearing behaviors (Young et al., 2022).
Unweighted sample characteristics: Age: Under 50 M = 38.32, SD = 7.33; 50+ M = 64.78, SD = 7.63). Gender: Under 50: 34.5% male, 50+: 46% male). Race: Under 50: 29.1% Non-Hispanic White, 30.9% Non-Hispanic Black, 29.9% Hispanic/Latinx, 10.1% other; 50+: 36.3% Non-Hispanic White, 28.9% Non-Hispanic Black, 30% Hispanic/Latinx, 4.8% other. Education: Under 50: 56.6% hold college degree; 50+: 54.4% hold college degree. Income: Under 50, 55.7% make over $60k; 50+: 58.6% make over $60k.
Epistemic Motivations Rooted in Feelings and Intuition (Wave 2)
Respondents were asked to what extent they agreed or disagreed (on a scale of 1 to 7) with the following statements (from Garrett & Weeks, 2017): “I trust my gut to tell me what’s true and what’s not.” “I trust my initial feelings about the facts.” “My initial impressions are almost always right.” “I can usually feel when a claim is true or false even if I can’t explain how I know.” Items were averaged to create an “epistemic feelings” score. (Under 50 M = 4.62, SD = 1.11, α = .80; 50+ M = 4.80, SD = 1.03, α = .79).1
Epistemic Motivations Rooted in Evidence and Data (Wave 2)
Respondents were asked to what extent they agreed or disagreed (on a scale of 1 to 7) with the following statements (from Garrett & Weeks, 2017): “Evidence is more important than whether something feels true.” “A hunch needs to be confirmed with data.” “I trust the facts, not my instincts, to tell me what is true.” “I need to be able to justify my beliefs with evidence.” Items were averaged to create an “epistemic evidence” score. (Under 50 M = 4.96, SD = 1.15, α = .77; 50+ M = 5.50, SD = .98, α = .71).2
Trump Favorability (Wave 2)
Respondents were asked, “Please rate the following public officials on a 0-10. A score of 5 means you feel equally favorable and unfavorable.” With one of the prompts being Donald Trump. (Under 50 M = 2.89, SD = 3.43; 50+ M = 2.72, SD = 3.84).
Ideology (Wave 1)
Respondents were asked, “How would you describe your political ideology?” very liberal (1), liberal, (2) somewhat liberal (3), middle of the road (4), somewhat conservative (5), conservative (6), very conservative (7). Under 50 M = 3.38, SD = 1.76; 50+ M = 3.95, SD = 1.69).
Party Identification (Wave 1)
Respondents were asked, “Generally speaking, do you think of yourself as a Democrat, Republican, Independent, or something else?” Those responding “Democrat” or “Republican” were asked would you describe yourself as a “strong” or “not very strong” Democrat/Republican. Those responding Independent or other were asked, “Would you say you are closer to the Democratic or Republican Party?” Responses were recoded into a seven-point scale from Strong Democrat (coded 1) to Strong Republican (coded 7). Under 50 M = 3.17, SD = 2.19; 50+ M = 3.17, SD = 2.29).
COVID Misperceptions (Wave 2: Nov 26 – Dec 4)
Respondents were asked to what extent they agreed or disagreed with the following statements on a 7-point scale, with higher values indicating higher misperceptions3: “Coronavirus (COVID-19) has affected most countries more negatively than the United States;” “Coronavirus (COVID-19 is a hoax;” “Asian American people are more likely to carry the virus than other people;” “Coronavirus (COVID-19) test results are often not valid indicators of whether or not a person is infected with the virus;” “The coronavirus (COVID-19) vaccine will be used to implant people with microchips;” “The flu is more lethal than coronavirus (COVID-19);” “Herd immunity is the best strategy for managing the coronavirus (COVID-19) in the US;” and “Coronavirus (COVID-19) affects communities of color more negatively than it affects other communities” (rev.). (Under 50 M = 3.38, SD = 1.34, α = .85; 50+ M = 2.77, SD = .94, α = .76).
Election Misperceptions (Wave 2: Nov 26 – Dec 4)
Respondents were asked to what extent they agreed (coded 7) or disagreed (coded 1) with the following statements on a 7-point scale, with higher values indicating higher misperceptions4: “The fact that Trump lost means that the election was rigged;” “There was widespread voter fraud in the 2020 Presidential election;” “We can never be sure that Biden’s win was legitimate;” “Joe Biden is the winner of the 2020 Presidential election (rev);” “Mail-in ballots are a safe and secure method of voting (rev);” and “The COVID pandemic made it necessary for states to expand mail-in voting to keep citizens safe (rev).” Items were averaged to calculate an Election misperception score for each respondent. (Under 50 M = 2.96, SD = 1.63, α = .89; 50+ M = 2.47, SD = 1.92, α = .95).
For analyses, data were weighted to be nationally representative on age, education, gender, and region. The parameters used in the post stratification in the 18-49 sample were age (18-34, 35-49), gender, education (High school or less, some college/associate’s degree, bachelor’s degree or higher), and census region (Northeast, Midwest, South, West). The parameters used in the post stratification in the 50+ sample were age (50-64, 65+), gender, education (less than college, bachelor’s degree or higher), and census region. The demographic benchmarks were obtained from the 2018 American Community Survey (ACS). Post-stratification weighting was accomplished using SPSSINC RAKE, an SPSS extension module that balances variable distributions using the GENLOG procedure.
First, multiple linear regression analyses were run examining associations between the three political preference variables – party, ideology, and Trump favorability – and misperceptions about COVID-19 (H1), controlling for sociodemographic variables (see Table 1). Models were significant for both age samples (50+: F(8, 1,206) = 33.60, p < .001, R2 = .24, 18-49: F(8, 423) = 3.24, p = .001, R2 = .13). Data show some support for H1, as Republicanism, conservatism, and Trump favorability were all positively related to misperceptions about COVID-19 in the 50 and older sample only. In the 18 – 49 sample, none of the three political variables were significantly related to misperceptions about COVID-19.
|Predictor||Predicting COVID Misperceptions
||Predicting Election Misperceptions
|b (SE)||β||b (SE)||β||b (SE)||β||b (SE)||β|
|Constant||4.20 (.45)***||3.03 (.26)***||3.40 (.46)***||.99 (.38)*|
|Trump favorability||.02 (.01)||.12||.01 (.00)**||.16||.08 (.04)||.29||.02 (.01)***||.14|
|Republicanism||.08 (.05)||.14||.12 (.02)***||.30||.22 (.05)***||.30||.51 (.03)***||.60|
|Conservatism||.03 (.06)||.04||.08 (.03)**||.15||.09 (.06)||.10||.16 (.04)***||.14|
|Black Race||.00 (.19)||.00||.09 (.08)||.04||-.74 (.18)***||-.21||-.04 (.08)||-.01|
|Male Gender||.35 (.17)*||.14||-.09 (.06)||-.05||.39 (.17)*||.12||-.08 (.09)||-.02|
|Education||-.01 (.01)||.00||-.01 (.00)**||-.06||.00 (.01)||.02||.00 (.00)||-.01|
|Income||.00 (.01)||-.02||-.02 (.01)*||-.11||-.02 (.01)**||-.10||.00 (.00)||-.01|
|Age||-.04 (.01)***||-.25||-.01 (.00)***||-.11||-.05 (.01)***||-.23||-.01 (.01)*||-.05|
*p < .05. **p < .01. ***p < .001.
Next, multiple linear regression analyses were run to examine the associations between the three political preference variables and misperceptions about the 2020 U.S. presidential election (H1) (see Table 1). Models were significant for both the 50+ sample, F(8, 1,206) = 104.88, p < .001, R2 = .56, and the under 50 sample, F(8, 421) = 28.04, p < .001, R2 = .44. For participants 50 and older, Republicanism, conservatism, and Trump favorability were all positively related to misperceptions about the 2020 U.S. presidential election. Among participants under the age of 50, Republicanism was a positive correlate and Trump favorability was marginally significant, but ideology was not a significant correlate of misperceptions about the 2020 U.S. presidential election. Therefore, data are consistent with H1 in the age 50 and older sample, as all three political constructs were significant positive predictors of election misperceptions. However, this only received mixed support in the 18-49 sample, as only political party (Republicanism) was a significant (p < .05) predictor of election misperceptions.
Turning to H2, multiple linear regression analyses were run to examine the relationships between the two types of epistemic motivations – those based on feelings and intuition (i.e., epistemic feelings) and those based on evidence and data (i.e., epistemic evidence) – and misperceptions about COVID-19 and the 2020 election (see Table 2). Sociodemographic covariates were included in all models. The models predicting COVID misperceptions were significant for both the 50 and older sample, F(7, 1,266) = 28.14, p < .001, R2 = .17, and the under 50 sample, F(7, 464) = 8.02, p < .001, R2 = .19). Consistent with H2a, epistemic feelings were a significant positive correlate of misperceptions about COVID-19 both among participants 50 and older and among those under 50. Consistent with H2b, Epistemic evidence was a significant negative correlate of misperceptions about COVID-19 among participants 50 and older and among those under 50.
|Predictor||Predicting COVID Misperceptions
||Predicting Election Misperceptions
|b (SE)||β||b (SE)||β||b (SE)||β||b (SE)||β|
|Constant||3.60 (.52)***||3.95 (.32)***||5.44 (.68)***||4.18 (.65)***|
|Epistemic Feelings||.42 (.08)***||.37||.22 (.03)***||.24||.32 (.09)***||.22||.29 (.06)***||.16|
|Epistemic Evidence||-.16 (.07)*||-.14||-.28 (.03)***||-.29||-.45 (.08)***||-.32||-.48 (.06)***||-.25|
|Black Race||-.18 (.17)||-.07||-.18 (.06)**||-.09||-.95 (.19)***||-.27||-1.18 (.11)***||-.28|
|Male Gender||.44 (.16)**||.17||-.03 (.07)||-.02||.62 (.20)**||.19||.15 (.13)||.04|
|Education||.00 (.01)||.01||-.00 (.00)||-.02||.01 (.01)||.03||.01 (.01)*||.03|
|Income||-.04 (.01)||.00||-.01 (.00)*||-.07||-.01 (.01)||-.04||.01 (.01)||.02|
|Age||-.04 (.01)***||-.29||-.01 (.00)*||-.07||-.05 (.01)***||-.24||.00 (.01)||-.02|
*p < .05. **p < .01. ***p < .001.
Multiple linear regression models were also run to assess the relationships between the two types of epistemic motivations and misperceptions about the 2020 U.S. presidential election controlling for demographic predictors. Models were significant for participants 50 and older, F(7, 1,266) = 30.71, p < .001, R2 = .17, and those under 50, F(7, 464) = 12.91, p < .001, R2 = .25. Again, consistent with H2a, epistemic feelings were significant positive correlates of misperceptions about the 2020 U.S. presidential election in both age groups. Consistent with H2b, epistemic evidence was a significant negative predictor of misperceptions about the 2020 U.S. presidential election in both age groups. In sum, data across both age samples were consistent with H2. Valuing intuition and feelings as pathways to truth was positively associated with belief in misinformation (about both COVID and the 2020 election), and valuing evidence and data was negatively associated with these beliefs, controlling for sociodemographic constructs.5
||50 and older
†p = .05. *p < .05. **p < .01. ***p < .001.
H3 concerns the relationships between epistemic motivations and political preferences. Multivariate multiple regression analysis was performed to assess how epistemic motivations were associated with (a) Republicanism, (b) conservatism, and (c) Trump favorability (see Table 3). Demographics were included as covariates. Results (Table 3) show some support for H3a; among respondents 50 and older, feelings-based motivations were positively associated with all three political constructs. In the 18–49-year-old sample, feelings-based motivations were associated with Trump favorability, but not with either Republicanism or conservatism. Turning to H3b, epistemic evidence was a significant negative correlate of Republicanism and conservatism in the 50+ sample, and was a significant negative correlate of Republicanism and Trump favorability in the 18-49 year old sample, offering some support for H3b.5 The strength of the standardized betas for epistemic motivations predicting Trump favorability in the 18-49 sample (epistemic feelings: b = .25, epistemic evidence b = -.24) suggest that support for Trump was especially associated with the value young people placed on different ways of knowing.
H4 proposed that these political constructs might mediate the relationships between epistemic motivations and belief in misinformation. Given the significant associations between political constructs and belief in misinformation (H1), epistemic motivations and belief in misinformation (H2), and epistemic motivations and political constructs (H3), there is reason to believe that political preferences may account for some of the relationships between epistemic motivations and belief in misinformation.
H4 was tested using the SEM command in STATA 16. Table 4 reports the indirect effects of epistemic motivations on misperceptions of COVID and the election through Republicanism, conservatism, and Trump favorability. Models controlled for race, gender, education, income, and age. Republicanism was a significant negative mediator of the relationship between epistemic evidence and misperceptions of both COVID and the election in both age groups. This is consistent with the proposition that Republicans lower likelihood of valuing evidence helped account for some of the negative association between evidence-based motivations and both forms of misperceptions. In the 50+ age group (but not the younger sample), we find the same pattern replicated with conservatism as a mediator. Additionally, Trump favorability was a significant negative mediator of the association between epistemic evidence and COVID misperceptions in the younger sample only. Finally, the only significant mediation effects found in the context of epistemic feelings were in the 50+ sample with Trump favorability. For this age group, feelings-based motivations were positively associated with Trump favorability, which helped account for some of the association between epistemic feelings and misperceptions of COVID and the election. We should note that several of these patterns were more pronounced in the 50+ sample than the 18-49 sample, a finding that may be attributable both to sample size and to the higher variance sometimes found in political constructs among younger populations (Fieldhouse et al., 2007).
|Predictor||Sample age||Through Trump Favorability
|COVID Misperceptions||Election Misperceptions||COVID Misperceptions||Election Misperceptions||COVID Misperceptions||Election Misperceptions|
|b (SE)||b (SE)||b (SE)||b (SE)||b (SE)||b (SE)|
|Epistemic Feelings||18-49||.01 (.01)||.14 (.13)||.00 (.01)||.02 (.04)||.01 (.01)||.03 (.03)|
|50 +||.02 (.01)*||.04 (.02)*||.02 (.01)||.06 (.05)||.02 (.01)||.07 (.04)|
|Epistemic Evidence||18-49||-.03 (.01)**||-.19 (.15)||-.05 (.02)*||-.13 (.04)**||-.01 (.01)||-.04 (.03)|
|50 +||-.01 (.01)||-.02 (.02)||-.07 (.01)***||-.26 (.05)***||-.05 (.01)***||-.16 (.04)***|
*p < .05. **p < .01. ***p < .001.
This study examined the link between epistemic motivations and belief in misinformation about COVID-19 and the 2020 U.S. election. Drawing on a motivated social cognition framework (Jost et al., 2003; Jost & Krochik, 2014; Jost, 2017), and the social and epistemic sorting of America’s political parties (Mason, 2018; Oliver & Wood, 2018), we outline how the epistemology of today’s socially conservative Republican party places more value on feelings and instinct over evidence and data. We then assess how current political preferences are associated with underlying epistemic motivations, and how these dynamics may contribute to greater belief in misinformation on the right than the left. Analyses confirm that belief in misinformation – about both COVID and the 2020 election – was more prevalent among Republicans, conservatives, and those more favorable towards President Trump. Our findings are consistent with prior work illustrating that valuing intuition and emotion over evidence and data is associated with greater belief in misinformation (Garrett & Weeks, 2017; Swami et al., 2014). But our work expands upon that research by integrating political preferences and proposing a mechanism through which political preferences may contribute to the epistemology-misperception link. Together these results speak to a political divide not just in Americans’ beliefs about COVID and the 2020 election, but in how they come to understand their worlds.
Our findings are consistent with a framework that conceptualizes political preferences as conduits for the construction of distinct partisan realities – particularly on the right. In the 50+ sample, Republicans and conservatives were less likely to value evidence, and conservatives and Trump supporters were more likely to value feelings. Similarly, in the 18-49 sample, Republicans and Trump supporters were less likely to value evidence and Trump supporters were more likely to value feelings. Mediation results also support the proposition that disregarding evidence and believing misinformation are associated in part due to political party and ideology; and that valuing feelings and believing misinformation may be associated in part through Trump favorability.
Admittedly, while our literature review proposes several possible mechanisms to account for these relationships, because of the cross-sectional nature of our data, we can’t say which of these constructs is driving this process. In one conceptualization, Trump’s socially conservative positions are attractive to those who value feelings and intuition while rejecting evidence and data (Oliver & Wood, 2018). Viewed this way, affinity for Trump is a manifestation of a feelings-based way of viewing the world and of coming to truth. A second framework involves Trump’s populist style and “shoot from the hip” persona embodying an epistemic value system that appeals to those who see Trump as a kindred spirit – someone who thinks and talks the way they do, trusting their gut over expertise and evidence. But these mechanisms could be reciprocal as well. Longitudinal and experimental work shows that psychological traits not only predict political preferences but are also influenced by political considerations (Bakker, Lelkes, & Malka, 2021). Trump’s threat-oriented rhetoric might prime aspects of political identity that encourage heuristic thinking, embrace of instinct, and rejection of evidence. Meanwhile, Trump’s rhetorical style, which offers a performance of “gut-based” thinking, might reinforce the notion that instinct is the best way to truth, and his criticism of experts might reinforce his supporters’ inclination to reject evidence.
Of the three political constructs explored here, Trump favorability is the only consistently significant correlate of feelings/instinct based epistemic motivations. Trump support is also the only significant mediator between feelings-based motivations and misperceptions. This association is noteworthy given Trump’s populist appeal which prioritizes common sense and instinct over expertise. By tapping into his supporters’ preexisting penchant for intuition and emotion-based thinking, while providing emotional, false stories to satisfy those epistemic needs (Evanega et al., 2020), perhaps Trump was indeed a “rhetorical genius” (Mercieca, 2020).
COVID misinformation and election misinformation both served Republican political goals and as such were shared by conservative media personalities, Republican political leaders, and Trump himself. Because trust in these sources may have contributed to belief in these falsehoods, future research ought to incorporate the role of trust and perceived source credibility as potential mediators and moderators of these relationships (see Traberg & van der Linden, 2022). Future work might also assess belief in non-partisan misinformation to better explicate the mechanisms through which misperceptions are formed.
Given the political valence of both COVID and election misinformation, these relationships might not apply to belief in liberal-serving misinformation. However, research indicates that ideological asymmetry exists in both the production and reception of misinformation (Garrett & Bond, 2021). Since valuing instinct and rejecting evidence are associated with political conservatism and with belief in misinformation (Garrett & Weeks, 2017; Pytlik, Soll, & Mehl, 2020), perhaps we should expect such falsehoods to be peddled more on the right than the left. This is not to say that liberal propaganda does not exist, but that it might employ tactics more compatible with the epistemic motivations of liberals, such as statistics, expertise, or data.
It is striking that discrediting experts and elites is a pillar of populist propaganda (Oliver & Rahn, 2016). As theorized through motivated social cognition (Jost et al., 2003), social and cultural conservatives’ high physical threat salience and resulting desire for efficiency encourage reliance on heuristics. While intuition and emotions can serve as consistent, readily available cues, expertise and the credibility of experts could also serve that purpose. But by strategically eroding trust in experts, Trump and populist conservative leaders might be eroding one of the few heuristics that could guide instinct-driven supporters in the direction of empirical truth.
The relationships identified here suggest how far right populist political leaders might appeal to emotions, intuition, and a rejection of evidence to attract socially conservative voters, and then may reinforce those inclinations through their rhetoric. It is this dynamic that might contribute to social conservatives’ susceptibility to – and even demand for – misinformation. Our findings underscore how the problem of misinformation is unlikely to be solved on the side of supply alone – though efforts like content moderation and fact-checking are still crucial. But as long as falsehoods satisfy some people’s epistemic needs, and as long as their beliefs remain unchanged by evidence, interventions will need to tackle the individual level needs that create fertile ground for misinformation in the first place. Fortunately, it may be possible to tap into instinct and emotions while also highlighting the importance of expertise and evidence (Van Bavel, FeldmanHall, & Mende-Siedlecki, 2015; Van Bavel, Xiao, & Cunningham, 2012). And even brief interventions can trigger analytical reflection and thwart the effects of misinformation (Pennycook et al., 2021). The question then becomes how to scale such interventions to prevent the next public health or democratic infodemic.