Polls show that many Americans, occasionally majorities, believe in various conspiracy theories (Enders et al., 2021). These beliefs are associated with numerous counterproductive behavioral tendencies (Jolley, Mari, et al. 2020a), including poor health practices (Romer & Jamieson, 2020), nonnormative political engagement (Sternisko et al., 2020), and discrimination against racial and ethnic minority groups (Jolley, Meleady, et al. 2020b). Conspiracy theory beliefs are also associated with political violence (Greenhill & Oppenheim, 2017), such as that which occurred on January 6, 2021 at the U.S. Capitol Building.

While an emerging research agenda has identified many psychological factors associated with beliefs in conspiracy theories (Douglas et al., 2019), the literature remains divided about the political characteristics associated with those beliefs. Are, for example, partisanship or political ideology related to conspiracy theorizing? One argument, dating back at least to Richard Hofstadter’s The Paranoid Style (1964), holds that Republicans and conservatives are more likely to believe conspiracy theories than Democrats and liberals. On the one hand, several studies find supportive evidence for this asymmetry thesis in the form of correlations between Republican/conservative self-identification and various operationalizations of conspiricism (e.g., van der Linden et al., 2021). On the other hand, some studies fail to identify such asymmetries, instead concluding that conspiracy theorizing is a “widespread tendency across the entire ideological spectrum” (Oliver & Wood, 2014; see also Enders et al., 2021). As this impasse persists, much hangs in the balance, including the development of strategies for combating the spread of conspiracy theories (Wittenberg & Berinsky, 2020).

We argue that contradictory findings are the result of researchers’ choices and assumptions. For example, most studies center their test of the asymmetry thesis on only a small number of conspiracy beliefs (e.g., van der Linden et al., 2021). This is potentially problematic because conspiracy theories are not created equally––they exhibit qualities (e.g., villains, schemes, victims) that make them differentially attractive depending on one’s psychological, political, and social characteristics and motivations (Miller, 2020). Indeed, conspiracy theory beliefs are frequently the product of motivated reasoning and elite influence (Miller et al., 2016). While both forces influence people regardless of political orientations (Guay & Johnston, 2021), they can also impact the relationship between conspiracy beliefs and political orientations depending on the theories queried.

Previous work also shows that the political context (both temporal and geographic) in which conspiratorial views are polled can affect who is most likely to believe specific conspiracy theories (Uscinski & Parent, 2014). Thus, findings of left–right asymmetries (or lack thereof) may be an artifact of which conspiracy theories researchers investigate, as well as when, where, and how they are investigated (Enders & Uscinski, 2021). One way to address these concerns is to conduct additional tests of the asymmetry thesis using multiple datasets, varying assumptions, conditions, and operationalizations. This is our goal.

First, we examine correlations between partisanship/ideology and beliefs in 52 conspiracy theories in the U.S. We buttress this analysis with an examination of the relationship between left–right ideology and belief in 11 conspiracy theories across 20 additional countries. In our second analysis, we provide an alternative test of the asymmetry thesis by examining beliefs in five “content-controlled” conspiracy theories in the U.S. for which we varied only the partisanship of the accused conspirators. In our final analysis, we examine the relationship between political orientations and conspiracy thinking––the general predisposition to interpret events and circumstances as the product of conspiracies (Uscinski & Parent, 2014)––in 18 datasets collected over the span of a decade (2012–2021). Across all studies, we fail to observe consistent evidence that the right exhibits higher levels of conspiricism––however operationalized––than the left. While some specific conspiracy theories find more support among Republicans/conservatives, others find greater support among Democrats/liberals. Moreover, many conspiracy theories appeal equally to the left and the right. Overall, we find that the strength and nature of the relationships between various operationalizations of conspiricism and partisanship/ideology depend on the conspiracy theories under examination and the political context in which those theories are examined.

Conspiracy Theories and Asymmetry

Since the development of Hofstadter’s “paranoid style” (1964), many researchers and journalists have argued that Republicans and conservatives are more prone to believing in conspiracy theories than Democrats and liberals (e.g., Bump, 2021). This perspective generally assumes, or surmises, that there are “bottom-up” psychological processes differentially affecting individuals on the right (Carney et al., 2008; Jost et al., 2009) that predispose them to believe conspiracy theories at higher rates and with greater intensity (e.g., van der Linden et al., 2021; for outside the U.S., see Walter & Drochon, 2020).

A considerable body of evidence has been harnessed to support this conclusion. First, researchers have identified numerous conspiracy theories that garner more support among the right than left, including those calling into question the scientific consensus on climate change (Miller et al., 2016), the authenticity of Barack Obama’s birth certificate (Pasek et al., 2014), COVID-19 (Miller, 2020), and the election of Joe Biden (Pennycook & Rand, 2021). Second, some researchers show that conservatives display higher levels of generalized conspiracy thinking than liberals (van der Linden et al., 2021); similarly, others show that “extreme” conservatives are more likely to believe conspiracy theories than “extreme” liberals (Imhoff et al., 2022; van Prooijen et al., 2015). Third, conspiracy theory beliefs are sometimes found to be associated with psychological tendencies more prevalent among conservatives, such as authoritarianism, need for cognitive closure, and threat perception (e.g., Dyrendal et al., 2021). Finally, several studies find evidence that those on the right are more likely than the left to engage with fake news and misinformation online (Garrett & Bond, 2021; Grinberg et al., 2019; Guess et al., 2019).

That said, not all investigations produce evidence for this form of partisan or ideological asymmetry and some studies even suggest that the left is more prone to conspiracy theorizing than the right. For example, Democrats/liberals are more likely than Republicans/conservatives to believe in conspiracy theories that identify Republicans, conservatives, corporations, and the rich as conspirators (Enders et al., 2021; Miller et al., 2016; Oliver & Wood, 2014; Uscinski & Parent, 2014). There are also many conspiracy theories finding equal support among the left and right, including theories involving “chem-trails”, the moon landing, fluoridated water, Freemasons, lizard people, and television mind control, to name a few (Jenson, 2013; Smallpage et al., 2017). Moreover, the asymmetries identified in studies of online behavior are often due to small numbers of Republicans/conservatives who are not representative of the right as a whole (Lawson & Kakkar, 2021) or idiosyncratic socio-political circumstances (Garrett & Bond, 2021). Finally, several studies find that the general predisposition toward conspiracy theorizing is balanced between the right and left (e.g., Enders et al., 2021; Uscinski et al., 2016, 2021). For every investigation demonstrating asymmetry there is a counterweight showing the opposite.

We posit that inconsistencies across findings are partially due to the concept of conspiracy theory itself. A conspiracy theory is an explanation of an event or circumstance that accuses powerful actors of working in secret for their own benefit, against the common good, and in a way that undermines bedrock societal norms, rules, or laws (Uscinski, 2020). Conspiracy theories are not likely to be “true”––i.e., no even-handed burden of empirical proof has been satisfied––according to appropriate epistemological authorities (Levy, 2007). The concept of “conspiracy theory” can refer to an infinite number of ideas that vary in myriad ways, including who they accuse, what the supposed scheme entails, and what the theory seeks to explain (Sternisko et al., 2020). In this light, scholars should not expect “a single ‘style’ of conspiricism, a uniform embrace of all conspiracy theories, or for conspiricism to be limited to one side of the ideological spectrum” (Oliver & Wood, 2014).

When researchers focus on conspiracy theories that are explicitly partisan (i.e., those accusing the out-party or its coalition, or otherwise bolstering the image of the in-party) or are championed by elites in government or media, they are likely to observe relationships between beliefs in such theories and political orientations. But this is so for reasons unrelated to political orientations themselves. First, since the mass public takes cues from co-partisan political and media elites (Zaller, 1992), mass opinions—even those involving conspiracy theories—are often a reaction to elite political discourse (Berinsky, 2015). Second, motivated reasoning, a ubiquitous psychological process by which individuals accept (reject) information congruent (incongruent) with their previously held beliefs and dispositions (Lodge & Taber, 2013), oftentimes leads individuals to adopt accusatory perceptions (Zell et al., 2021), including conspiracy theories about political out-groups (Miller et al., 2016). Given that elite cues and motivated reasoning appear to similarly affect both Republicans/conservatives and Democrats/liberals (Bolsen et al., 2014; Clark et al., 2019; Ditto et al., 2019; Guay & Johnston, 2021), beliefs in conspiracy theories born of these processes signal little about an innate connection between such beliefs and partisanship/ideology.Footnote 1

Conspiracy theories, like other ideas, can also attract different adherents as political and cultural circumstances shift over time. For instance, belief in conspiracy theories about election fraud are tightly tethered to the electoral fortunes of one’s preferred party and that party’s messaging (Edelson et al., 2017; Pennycook & Rand, 2021). Whereas 9/11 “truther” theories found more support among Democrats/liberals in the years immediately following the attacks, likely due to both partisan motivated reasoning and elite cueing, support for these theories has become symmetrical as the 9/11 attack faded from partisan political discussion (Enders et al., 2020). In cases such as these, researchers’ choices of when to poll the public leads to different observed relationships between conspiracy theory beliefs and political orientations. Similarly, the choice of where to poll can affect observed relationships. For example, climate change conspiracy theories find more support in socio-political contexts (e.g., the United States, Australia) where elites have politicized climate change (Dunlap et al., 2016) than other contexts where the subject is less politicized.

Simply put, making generalizable claims about the nature, scope, and correlates of conspiracy theory beliefs is no easy task. We surmise that many of the observed disagreements in the literature are due to a combination of limitations regarding the operationalizations of conspiracy theorizing employed and the context––both temporal and socio-political––in which beliefs are assessed. Our contribution is to provide a comprehensive set of tests of the asymmetry thesis using dozens of specific conspiracy theories and different measurement strategies across time in the U.S. and other countries. In the following three sections, we assemble an expansive body of survey evidence to answer the question: Are those on the political right (e.g., Republicans and conservatives) more likely than those on the left (e.g., Democrats and liberals) to believe in conspiracy theories?

Finding 1: The Relationship between Conspiracy Theory Beliefs and Political Orientations Depends on the Conspiracy Theory

Much of the evidence provided for the asymmetry thesis comes in the form of correlations between partisanship/ideology and beliefs in one or a few conspiracy theories. However, the relationships identified may be due to the specific conspiracy theories under investigation rather than “bottom-up” psychological asymmetries between left and right. To interrogate the asymmetry thesis, we examine the relationship between political orientations and a wides range of conspiracy theory beliefs. Figure 1 displays Pearson correlations between beliefs in 52 conspiracy theories and both partisanship and ideological self-identifications.

Fig. 1
figure 1

Pearson correlations between beliefs in conspiracy theories and partisan and ideological self-identification. Horizontal bands represent 95% confidence intervals, two-tailed tests

These data were collected across eight national surveys fielded between October 2016 and May 2021.Footnote 2 All surveys were fielded by either YouGov, as part of the annual Cooperative Congressional Election Study, or Qualtrics, and all samples, which vary in size from 1,000 to 2,023 respondents, were designed to be representative of the U.S. population based on age, sex, race, and educational attainment; see the online appendix for additional information about the sociodemographic composition of each sample, sampling procedures, and other details.

We compiled a set of conspiracy theories that vary in their topical domains (e.g., health, science, politics), accused conspirators (e.g., partisan and non-partisan groups and figures, “the government”), alleged activities (e.g., causing direct harm, undermining democracy, covering-up vital information), salience (e.g., the activities of the Rothschild family versus the current COVID-19 pandemic), and level of specificity (e.g., explicitly identifying the conspirator versus positing a more ambiguous conspirator or group). All items meet the standard definition of “conspiracy theory” described above. Both partisanship and ideological self-identification are operationalized using the familiar seven-point measures coded such that greater values correspond to stronger Republican/conservative identification.Footnote 3 If individuals on the right are asymmetrically predisposed to adopting conspiracy theory beliefs, then we should consistently observe positive, statistically significant correlations between conspiracy theory beliefs and partisan and ideological identities across Fig. 1.

Figure 1 reveals that both the direction and magnitude of the correlations between conspiracy theory beliefs and partisanship/ideology vary considerably across conspiracy theories. When the conspiracy theory implicates actors associated with the political left in wrongdoing (e.g., Birther), or are endorsed by Republican and conservative elites (e.g., Global Warming Hoax), Republicans/conservatives exhibit greater levels of belief than Democrats/liberals (see upper third of Fig. 1). Likewise, when the conspiracy theory implicates actors associated with the political right (e.g., Koch Brothers World Control), or are endorsed by Democratic and liberal elites (e.g., Trump is a Russian Asset), Democrats/liberals exhibit greater levels of belief than Republicans/conservatives (lower third of Fig. 1). We also observe that many conspiracy theories find equal support among the left and right. Online appendix Figure A1, which shows the percentages of respondents on the right and left believing in each conspiracy theory shown in Fig. 1, also reveals parity in the proportion of believers across the political spectrum.

Our results regarding two specific conspiracy theories deserve additional emphasis. First, those on the left and right equally express belief in the general theory that “Regardless of who is officially in charge of governments and other organizations, there is a single group of people who secretly control events and rule the world together.” This question captures a sentiment that is presumably foundational to many specific conspiracy theory beliefs; that we observe no difference between left and right may suggest that the psychological bedrock for conspiricism traverses mainstream political orientations. Second, we find political balance in the belief that “The U.S. government is mandating the switch to compact fluorescent light bulbs because such lights make people more obedient and easier to control,” which was fabricated by researchers (Oliver & Wood, 2014). That the right does not exhibit greater levels of belief in this conspiracy theory prompts us to further question whether there is an innate connection between right-wing identification and conspiracy theory beliefs.

While our intention is to test the asymmetry thesis, specifically, we note that other research has found that extremists––those who identify with the most extreme left/right categories of political identity measures––are more likely than moderates to exhibit conspiracy beliefs (Imhoff et al., 2022). Such a pattern would manifest as a nonlinear, parabolic relationship between political orientations and conspiracy beliefs. Of course, correlations (like the ones presented in Fig. 1) are only capable of deciphering linear relationships. Whereas some nonlinear functional form may be the “correct” model—although this, too, is in contention, (Enders & Uscinski, 2021; van der Linden et al., 2021)—our goal is not to provide the “best” model of each conspiracy belief, but to detect asymmetries. Even if some nonlinearity was present, correlation coefficients would still capture left–right asymmetries should they exist (this applies to all analyses presented below). Regardless, an investigation of potential nonlinear relationships suggests that our empirical strategy is appropriate; in the appendix, we provide evidence that the relationships we are interested in tend to be linear.

Moving Beyond the US

Thus far, we found that the observed relationship between beliefs in various conspiracy theories and political orientations in the U.S. is dependent on the specific conspiracy theories in question. Still, an examination of the relationship between political orientations and conspiracy theory beliefs across a wide range of socio-political contexts can provide further clarity and demonstrate the role of political context. If there is an innate connection between right-wing ideology and conspiracy theory beliefs, we should observe greater support for conspiracy theories among those on the right across countries, regardless of variation in political and economic systems, social (in)equality, racial and ethnic composition of the populace, and many other factors.

To extend our test of the asymmetry thesis in this way, we examine the correlations between a seven-point measure of left–right ideologyFootnote 4 and 11 conspiracy theory beliefs across 20 countries that span six continents (total n = 26,416), which are presented in Fig. 2.Footnote 5 All surveys were conducted by YouGov between July 30–August 24, 2020, who constructed the samples to be representative of each country’s population based on available census records. Questions were approved of and translated by YouGov and their partners in each country. Additional details about each survey, including sociodemographic information, appear in the appendix. Asking about the same 11 conspiracy theories that address topics traversing socio-political contexts (e.g., AIDS, COVID-19) allows us to examine the impact of context on the relationship between ideology and conspiracy theory beliefs, providing additional tests of the asymmetry thesis under different conditions.

Fig. 2
figure 2

Pearson’s correlations between beliefs in conspiracy theories and left–right ideological self-identification across 20 countries. Horizontal bands represent 95% confidence intervals, two-tailed tests

Remarkably, not a single conspiracy theory investigated in Fig. 2 exhibits correlations with consistently positive or negative signs (regardless of statistical significance) across all 20 countries. For example, the 9/11 conspiracy theory finds significantly more support among those on the left in 8 of the 20 countries and is not significantly related to ideology in the remaining 12. Even the Global Warming Hoax belief, a feature of the political right in the U.S., finds significantly more support among the left in 5 countries (Egypt, Mexico, Nigeria, Saudi Arabia, and Turkey) and exhibits no correlation with ideology in 2 others (Hungary and South Africa). Despite some ideological “balance” across Fig. 2, most correlations are quite weak, suggesting a lack of ideological discrepancy altogether. Indeed, the average correlation––across all conspiracy theory beliefs and countries––is a meager 0.03 (ranging from − 0.05 for 9/11 Truther to 0.10 for Global Warming Hoax).

To be sure, there are many possible explanations for the variability we observe across conspiracy theories and countries. Our goal, however, is not to provide post-hoc explanations for this variation, but rather to leverage it towards additional tests of the asymmetry thesis. While some countries (e.g., France, Germany, Italy) and some conspiracy theories (e.g., Global Warming Hoax, COVID is a Myth) provide more support for the asymmetry thesis, others exhibit the opposite pattern. The relationship between political ideology and conspiracy theory beliefs appears to be dependent on both the details of the conspiracy theories and socio-political context.

Altogether, we find that symmetry comes in two forms. First, there are conspiracy theories that prove to be systematically attractive to those on either the left or right. Second, many conspiracy theories find similar levels of support (or lack thereof) across the political aisle. We do not argue, nor do our data demonstrate, that beliefs in any given conspiracy theory are or should be symmetric across political predispositions––this is entirely dependent on the details of the conspiracy theory, and perhaps the context in which beliefs were assessed.

Of course, one might protest that the patterns depicted in Figs. 1 and 2 are artifacts of the conspiracy theories we chose to examine and, indeed, they would be correct! However, as much would be the case for any study of specific conspiracy theories. This critically important point explains the discrepancies among previous studies: substantive inferences are heavily dependent on which conspiracy theories are considered. Inferences about the fundamental nature of conspiracism should not be made from patterns in a single or small number of conspiracy beliefs, even though precisely such generalizations are commonplace in the conspiracy belief literature.

Decisions about which conspiracy theories to poll and analyze are always fraught with challenges. For example, one could criticize Finding 1 by claiming that some of the conspiracy theories we employed are more plausible (e.g., believable, evidenced, or rational) than others, and that this variability in plausibility is correlated with ideology or partisanship. However, judgements of this nature (Douglas et al., 2022), even among otherwise discerning researchers, are colored by motivated reasoning. It should not be a surprise that a Democrat, for example, would believe that the conspiracy theories that other Democrats believe in are more plausible than the conspiracy theories that Republicans believe in. Nevertheless, the conspiracy theories at the top and bottom of Fig. 1 are different from each other and not necessarily comparable (though, this is again a matter of subjective judgment). We, therefore, point readers to the conspiracy theories in the middle of Fig. 1. These are particularly important because the correlations with partisanship and ideology––or the lack thereof––are inconsistent with the asymmetry thesis (i.e., the right should be more likely than the left to believe these conspiracy theories should the asymmetry thesis be correct). For example, partisanship and ideology are not correlated with beliefs in conspiracy theories about the JFK assassination, the MMR vaccine, the Holocaust, GMO’s, Fluoride, cellphones, AIDS, pharmaceutical companies, government mind control, and lightbulbs.

Furthermore, even though Finding 1 involves more conspiracy theory beliefs than previous studies, our findings remain an artifact of which conspiracy theories researchers investigate because there is no “correct” or “representative” set of conspiracy theories that researchers should employ. This problem motivates the analyses that follow: an experiment which holds constant all but the partisanship of the accused conspirators in a set of conspiracy theories, as well as an analysis of the general predisposition toward conspiracy thinking. These analyses are attempts to avoid the trappings of individual conspiracy theories.

Finding 2: The Political Left and Right Both Accuse the Out-Party of Conspiring

Even though we considered a larger group of conspiracy theories above than any study we are aware of, one might still wonder how the idiosyncrasies of the conspiracy theories we examined may affect our inferences about partisan and ideological (a)symmetry. To address this concern, we present an alternative strategy for exploring the relationship between conspiracy theory beliefs and political orientations that is not subject to alternative explanations owing to the idiosyncrasies of the conspiracy theories themselves. Specifically, we hold the details of the conspiracy theory constant, varying only the political orientations of the presumed conspirators. This approach enables a test of a more nuanced version of the asymmetry thesis—that people on the right engage in partisan/ideologically motivated conspiracy theory endorsement to a greater extent than people on the left. A finding of asymmetry in this circumstance would show that Republicans/conservatives are more likely than Democrats/liberals to endorse conspiracy theories that impugn political out-groups.

To this end, on two surveys––one fielded by MTurk in 2017 (n = 2,041) and the other by Lucid in 2020 (n = 3,994)––we randomly assigned half of our respondents to receive the “Republicans are conspirators” versions of five conspiracy theory questions, and the other half to receive the “Democrats are conspirators” versions. To hold the content of the conspiracy theory constant, we created five general conspiracy theories on the topics of election fraud, political extremism, the economy, health policy, and crime. For example, the “economy” conspiracy theory question asked:

Do you think that Democratic [Republican] political elites are…

definitely secretly plotting with large banks to lie about the health of the economy to gain support for their economic policy proposals (coded 4)

probably secretly plotting with large banks to lie about the health of the economy to gain support for their economic policy proposals (3)

probably not secretly plotting with large banks to lie about the health of the economy to gain support for their economic policy proposals (2)

definitely not secretly plotting with large banks to lie about the health of the economy to gain support for their economic policy proposals (1)

Full question wordings for the election fraud, political extremism, economy, health policy, and crime questions appear in the online appendix. Partisanship was measured via the standard seven-point branched question, and ranges from “strong Democrat” to “strong Republican.” Symbolic ideology was measured via a seven-point scale ranging from “extremely liberal” to “extremely conservative.”

Because of motivated reasoning (and consistent with the general pattern depicted in Fig. 1), we expect that Republican/conservative respondents would be more likely to endorse the “Democrats are conspirators” version of the questions (as indicated by positive correlations with partisanship and symbolic ideology), and Democratic/liberal respondents would be more likely to endorse the “Republicans are conspirators” version (as indicated by negative correlations with partisanship and symbolic ideology). For our purposes, the more pertinent question is whether partisan/ideologically motivated conspiracy theory endorsement is asymmetrical. Does the relative size of the relationship between motivated conspiracy theory endorsement and political orientations vary depending on whether the purported conspirators are Democrats or Republicans?

Figure 3 displays the absolute value of the correlations between political orientations and both partisan versions of the conspiracy theory questions described above. First, we should always observe correlations that are significantly distinguishable from 0 (i.e., p < 0.05, two-tailed test); this would indicate that some level of partisan motivated reasoning is behind the endorsement of conspiracy theories that involve partisan groups/figures, consistent with findings presented above. Second, we should observe significantly larger correlations between political orientations and beliefs in conspiracy theories that malign Democrats (the red points) compared to those that malign Republicans (blue points) if Republicans and conservatives are asymmetrically motivated to believe in conspiracy theories about the left.

Fig. 3
figure 3

Pearson’s correlations between beliefs in specific content-controlled conspiracy theories and partisan and ideological self-identification. Horizontal bands represent 95% confidence intervals, two-tailed tests. Data from the following surveys: Lucid February 2020 (nRepublican Conspirator = 2,009; nDemocratic Conspirator = 1,985), MTurk October 2017 (nRepublican Conspirator = 992; nDemocratic Conspirator = 1,023). See online appendix for sample characteristics

While we observe evidence of partisan motivated conspiracy theory endorsement in all instances, we do not observe a tendency for Republicans or conservatives to engage in such reasoning to a greater extent than Democrats and liberals. Regarding partisanship, in five of the 10 tests, the correlations are not significantly different from one another, indicating symmetry (i.e., p > 0.05). In the other five, Democrats evidence a greater propensity for partisan motivated conspiracy endorsement than Republicans. A similar pattern emerges for ideology. In four of the 10 tests, the correlations are not statistically significantly different from one another. In the remaining six, liberals engage in greater motivated conspiracy theory endorsement than conservatives.

In sum, we find no support for the hypothesis that those on the right are more likely to endorse conspiracy theories that impugn liberals than liberals are to endorse the exact same conspiracy theories when they impugn conservatives. This finding has an important implication for the asymmetry thesis: many individuals accept the opportunity to derogate the opposing political side, regardless of the specific factual details of the charge and even if the charge alludes to a conspiracy. These results explain the patterns in Fig. 1 and lead to our final analysis: an examination of potential partisan/ideological asymmetries in conspiracy thinking, the general predisposition. Despite our findings thus far, this predisposition may be asymmetric across partisan/ideological lines.

Finding 3: Conspiracy Thinking is Unsystematically Related to Political Orientations

A final analytical strategy for assessing the relationship between political orientations and conspiracy theory beliefs entails shifting focus from belief in specific conspiracy theories to the general predisposition to engage in conspiracy theorizing (e.g., Bruder et al., 2013). This predisposition, conspiracy thinking (sometimes referred to as conspiracy ideation, conspiratorial predispositions, or conspiracy mentality), can be thought of as a bias toward interpreting events and circumstances as the products of conspiracies carried out by powerful, malign actors (Cassese et al., 2020; Uscinski et al., 2016). Empirical measures of conspiracy thinking are specifically designed to transcend the trappings of individual conspiracy theories; in other words, an examination of the relationship between political orientations and conspiracy thinking has the advantage of avoiding the complications of substantive interpretation inherently posed by the idiosyncrasies of beliefs in particular conspiracy theories.

To measure this predisposition, we seek to measure general conspiratorial sentiments in which abstract nefarious groups are subverting norms, rules, and laws by secretly engaging in harmful actions or by covering-up such actions. Here, we employ the 4-item American Conspiracy Thinking Scale (ACTS), first developed by Uscinski and Parent (2014) and based on items from McClosky and Chong (1985). This scale is very similar in content and construction to other scales intended to capture the disposition to believe conspiracy theories, such as the five-item Conspiracy Mentality Questionnaire (Bruder et al., 2013) and the 15-item Generic Conspiracist Beliefs Scale (Brotherton et al., 2013). Subjects responded, using five-point response options ranging from “strongly disagree” (1) to “strongly agree” (5), to each of the following items:

  1. 1.

    Even though we live in a democracy, a few people will always run things anyway.

  2. 2.

    The people who really “run” the country are not known to the voters.

  3. 3.

    Big events like wars, the recent recession, and the outcomes of elections are controlled by small groups of people who are working in secret against the rest of us.

  4. 4.

    Much of our lives are being controlled by plots hatched in secret places.

The resultant additive scale has been used in numerous studies, consistently accounting for variation in beliefs in a wide range of conspiracy theories (Cassese et al., 2020; Miller, 2020; Uscinski et al., 2016). Cronbach’s alpha reliability estimates range from 0.76 to 0.86 across the 18 datasets employed in this analysis, suggesting high reliability; the proportion of variance explained by the first factor of an exploratory factor analysis ranges from 0.86 to 0.94, suggesting unidimensionality. See the online appendix for detailed information about each individual dataset, including the measurement properties of the ACTS (Figure A2).

To validate this measure in the context of the current investigation, Fig. 4 presents the correlations between the ACTS and beliefs in each of the specific conspiracy theories examined in Fig. 1. Our intention here is to show that the ACTS measure is a predictively valid indicator of the general tendency to believe in conspiracy theories. We should expect all correlations to be positive and statistically significant. Figure 4 shows universal support for this expectation. Moreover, we observe a pattern consistent with our earlier arguments about the impact of partisan/ideological motivated reasoning and opinion leadership on beliefs in some conspiracy theories: there is a weaker correlation between the ACTS and beliefs in the specific conspiracy theories exhibiting the strongest political asymmetries in Fig. 1 (i.e., those listed at the very top and very bottom of the vertical axis).

Fig. 4
figure 4

Pearson correlations between beliefs in specific conspiracy theories from Fig. 1 and conspiracy thinking scale. Horizontal bands represent 95% confidence intervals, two-tailed tests. Qualtrics May 2021 (n = 2,021), Qualtrics October 2020 (n = 2,015), Qualtrics June 2020 (n = 1,040), Qualtrics March 2020 (n = 2,023), Qualtrics July 2019 (n = 2,000), CCES October 2018 (1, n = 1,000), CCES October 2018 (3, n = 1,000), CCES October 2016 (2, n = 1,000); see online appendix for sample characteristics

Having established the criterion validity of the ACTS, we now turn to the empirical test of central concern: the relationship between political orientations and conspiracy thinking. Figure 5 depicts the correlations between conspiracy thinking and ideological self-identification, partisanship, and operational ideology. This analysis includes a total of 18 datasets (n = 31,741) spanning 10 years (2012–2021). Details about all datasets appear in the appendix.

Fig. 5
figure 5

Pearson’s correlations between conspiracy thinking scale and partisanship, symbolic ideology, and operational ideology across studies and time. Horizontal bands represent 95% confidence intervals, two-tailed tests. Data sources listed along vertical axis. Individual sample sizes and sample characteristics appear in the appendix

Beginning with symbolic ideology (measured via a seven-point scale ranging from “extremely liberal” to “extremely conservative”), no consistent pattern emerges (see the first panel of Fig. 5). In some surveys the correlation is positive (indicating higher levels of conspiracy thinking on the right than left), in others it is negative (indicating the opposite). In others still, correlations are statistically indistinguishable from 0. Moreover, those correlations that are statistically significant tend to be small, rarely exceeding an absolute value of 0.10 and never exceeding an absolute value of 0.25. The average correlation between conspiratorial thinking and ideological self-identification across all 18 national surveys is 0.04. The relationship between conspiratorial thinking and partisanship (a seven-point scale ranging from “strong Democrat” to “strong Republican) is similarly inconsistent and weak across surveys/time (middle panel of Fig. 5); the average correlation is 0.03.

Eight of our surveys contained multiple questions measuring issue attitudes, such as those regarding healthcare reform. For each survey we generated summary scales of these attitudes, a common method for measuring operational ideology (Ansolabehere et al., 2008; see online appendix for question wordings). Thus, we can investigate whether measures of ideology based on policy preferences rather than on identity reveal relationships between ideology and conspiracy thinking. The third panel of Fig. 5 demonstrates that the relationship between conspiratorial thinking and operational ideology is also inconsistent (ranging from 0.00 to 0.23), with an average correlation of 0.09. Although these correlations are slightly stronger than the averages we observe for symbolic ideology or partisanship, we note that we have fewer data points; moreover, the positive relationship is substantially driven by the same three 2016 data points for which we observed the largest correlations with respect to symbolic ideology and partisanship in the first two panels of Fig. 5.

We suspect that this spike––observed across all three panels of Fig. 5 in 2016––owes to the political context in which the surveys were fielded: 2016 was a presidential election year during which Donald Trump openly and frequently trafficked in conspiracy theories, sharply deviating from past norms. While conspiracy thinking is conceived of as a largely stable predisposition (Uscinski & Parent, 2014), emerging scholarship suggests that conspiracy thinking can be temporarily heightened or diminished in some groups by situational political factors (Einstein & Glick, 2013; Farhart et al., 2020). This is not unlike how other predispositions like partisanship, for example, operate: salient stimuli cause short-term fluctuations around an otherwise stable mean (Smidt, 2018; West & Iyengar, 2020). The data offer suggestive evidence for our proposition. In October 2012, the mean score on the ACTS (range: 1–5) was 3.19 for Democrats and 3.13 for Republicans. In October 2016 (average of both CCES datapoints), this average increased to 3.51 for Republicans, but remained stable for Democrats at 3.19. By October 2018 (average of all 3 CCES datapoints), it had returned to 2012 levels––3.17––for both parties.

Regardless of the changing political circumstances and a potential Trump effect, the totality of the evidence presented above is simply not indicative of a systematic left–right asymmetry in conspiracy thinking. As an additional check, we also examined each bivariate relationship between ideology/partisanship and conspiracy thinking separately, looking for evidence that political extremity, rather than valence (i.e., left/right), is related to conspiracy thinking. We found only mixed evidence for the proposition that strong partisans or extreme ideological identifiers exhibit higher levels of conspiracy theorizing than weak identifiers or moderates/independents. Details about these analyses appear in the appendix.

Discussion and Conclusion

Are those on the political right (Republicans/conservatives) more prone to conspiracy theorizing than those on the left (Democrats/liberals)? The smattering of evidence across the literature provides conflicting answers to this question. We surmise that disagreement in the literature is substantially the product of limitations regarding both the operationalizations of conspiracy theorizing and the context––both temporal and socio-political––in which beliefs are assessed in previous work. Given the imperative of better understanding conspiracy theories and the people who believe them, we compiled a robust body of evidence for testing the asymmetry thesis. Across multiple surveys and measurement strategies, we found more evidence for partisan and ideological symmetry in conspiricism, however operationalized, than for asymmetry.

First, we found that the relationship between political orientations and beliefs in specific conspiracy theories varied considerably across 52 specific conspiracy theories. Conspiracy theories containing partisan/ideological content or that have been endorsed by prominent partisan/ideological elites will find more support among those in one political camp or the other, while theories without such content or endorsements tend to be unrelated to partisanship and ideology in the U.S. We also observed considerable variability in the relationship between left–right ideology and 11 conspiracy theory beliefs across 20 additional countries spanning six continents; this variability suggests that the relationship between left–right ideology and conspiracy theory belief is also affected by the political context in which conspiracy theories are polled. To account for the potential impact of idiosyncratic factors associated with specific conspiracy theories, we next examined the relationship between beliefs in “content-controlled” conspiracy theories and political orientations. We found that both Democrats/liberals and Republicans/conservatives engage in motivated conspiracy endorsement at similar rates, with Democrats/liberals occasionally exhibiting stronger motivations than Republicans/conservatives. Finally, we observed only inconsistent evidence for an asymmetric relationship between conspiracy thinking and either partisanship, symbolic ideology, or operational ideology across 18 polls administered between 2012 and 2021. Even though the average correlations across studies were positive, indicating a relationship with conservatism/Republicanism (owing mostly to data collected in 2016), they were negligible in magnitude and individual correlations varied in sign and statistical significance over time.

Equally important as our substantive conclusions is an exploration of why we reached them, which can shed light on existing inconsistencies in the literature. While the core inferences we make from our investigation may deviate from the conclusions of others, empirical patterns are not irreconcilable. Take, for example, the study conducted by van der Linden and colleagues (2021). They infer from a strong, positive correlation between beliefs that “climate change is a hoax” and conservatism that conservatives are inherently more conspiratorial than liberals. However, we demonstrate that such conclusions cannot be made using beliefs in a single conspiracy theory. As can be seen in Fig. 1, climate change conspiracy theories show one of the highest levels of asymmetry; therefore, exclusive examination of almost any other conspiracy theory would lead to a result less supportive of the asymmetry argument.

Van der Linden et al. (2021) also find a positive, albeit weak, correlation between conservatism and generalized conspiracy thinking. While this relationship is statistically significant, liberals still exhibit high levels of conspiracism. Indeed, even strong liberals score above the 50-point midpoint on their 101-point measure (between 60 and 65, on average), whereas strong conservatives typically score about 10 points higher (see Figs. 1b and 3b). In other words, liberals, like conservatives, are more conspiratorial than not. Moreover, van der Linden et al.’s data hail from 2016 and 2018––years in which we also observed relatively elevated levels of conspiracy thinking among conservatives. However, this was not the case in other years and samples we examined. This is exactly what we might expect of a disposition that is not inherently connected to partisanship and ideology, but which may be sporadically activated by political circumstances. We do not question the veracity of van der Linden et al.’s empirical findings or those of any other study with conclusions that disagree with ours; rather, we argue that differences largely stem from the inferences made from empirical relationships, which are frequently more general than the data allows.

Despite the magnitude of data we employ, our study is not without limitations, and we wish to emphasize that ours should not be the final word on this topic. Although our data spans a decade, it was collected over the course of only three U.S. presidential administrations. As political culture changes so, too, might the relationship between political orientations and conspiracy theories. Unfortunately, measures of general conspiracy thinking (to our knowledge) were not deployed on national surveys until 2012 and specific conspiracy beliefs were only intermittently polled in the past 70 years, severely limiting how much we can know about conspiracy theorizing in the past. We encourage researchers to track multiple operationalizations of conspiracy theorizing into the future so that we may better understand their political dynamics and consequences.

We also recognize that, while an investigation of the asymmetry thesis across 21 countries constitutes a robust test, the more tests the asymmetry thesis undergoes the more confident we can be about its (lack of) veracity. We encourage an examination of more conspiracy theory beliefs across socio-political contexts, especially those that are closely tethered to each country’s political culture. We also recommend more robust examinations of the asymmetry thesis in regions that have been understudied, such as South America, Africa, and Asia. Even though we included countries from each of these continents, very little is known about the basic nature and scope of conspiracy theorizing outside of North America and Europe.

In a similar vein, conspiracy theories differ not only in who believes them, where, and when, but in their consequences and dangers. As such, it may be useful for researchers to consider categorizing conspiracy beliefs by various attributes, such as their consequences, just as they do for political attitudes (e.g., issue attitudes, affective versus ideological attitudes, etc.)––perhaps the asymmetry thesis finds stronger evidence among certain “classes” of conspiracy theories. Recent events in American politics are suggestive of this possibility. Donald Trump and his allies in government and media fostered election fraud conspiracy theory beliefs to the point of the violent intimidation of elected representatives attempting to certify the 2020 election. In this way, election fraud conspiracy theories––at least under the particular circumstances that Trump and colleagues nurtured––are of more consequence than, for example, conspiracy theories regarding the moon landing or lizard people. While forecasting which conspiracy theories will result in tangible consequences and when is surely difficult, we nevertheless note that symmetry of tendency to believe in conspiracy theories need not equal symmetry in consequence of conspiracy theories, along political lines or otherwise.

Finally, we believe it is critical that work on beliefs, like that presented here, be reconciled with related research examining political asymmetries in the tendency to interact with or “spread” conspiracy theories on social media. Related work by Guess, Nagler and Tucker (2019), Garrett and Bond (2021), and Grinberg et al. (2019), for example, finds evidence for minor asymmetries in the extent to which Democrats/liberals and Republicans/conservatives share misinformation or distinguish between fake and true news stories online. By fusing social media data with survey data researchers can gain greater leverage over questions about the conditions under which online behaviors are reflective of, or even impact, beliefs and offline behaviors. For now, we simply note that findings of asymmetries online may not generalize to the broader population, as politically active social media users are not representative of average Americans when it comes to various political and psychological characteristics (Lawson & Kakkar, 2021). Just as social media data can be fused with survey data, so, too, can top-down data on conspiratorial rhetorical strategies employed by political elites. Few studies of this sort have been undertaken, particularly in the U.S. (see Oliver & Rahn, 2016 for an example), but they are sorely needed––especially to test earlier studies on the rhetoric of conspiratorial elites (Adorno, 2000; Lowenthal & Guterman, 1948).

The last five years have witnessed Republican elites in government and media (most notably Donald Trump) utilizing conspiracy theories in a way unprecedented in the last half century of American politics, and with severe, deleterious consequences for democratic institutions. This alone has encouraged renewed conjecture about an asymmetry in conspiracy theory beliefs. However, elites are an imperfect reflection of the public––they have different goals, incentives, and knowledge about politics. Moreover, elite rhetoric rarely changes predispositions, such as conspiracy thinking, so much as it activates predispositions and connects them to salient political choices (Leeper & Slothuus, 2014). In other words, while Republican elites may have recently activated conspiratorial predispositions among supporters in the mass public––where they exist––in a way that Democratic elites did not, they are unlikely to be able to cause once non-conspiratorial supporters to become highly conspiratorial.

That we find little difference in conspiracy theorizing between the right and left among the mass public does not indicate that there are no differences between partisan elites on this score, nor does it imply that there will not be asymmetries in beliefs in specific conspiracy theories at any given point in time. Specific conspiracy theories can find more support among one partisan/ideological side than the other even though partisan/ideological motivated reasoning and conspiratorial predispositions operate, on balance, in a symmetric fashion. Likewise, the content of those theories and the way they are deployed, particularly by elites, can result in asymmetrical consequences, such as political violence and the undermining of democratic institutions. We encourage future work to integrate the conspiratorial rhetoric of elites with studies of mass beliefs and investigate elite conspiratorial rhetoric from actors including and beyond Donald Trump.