Abstract
Objective: This study disentangles the known correlates of conspiracy beliefs—such as the general predisposition toward conspiratorial thinking, authoritarianism, and partisan and ideological predispositions—in order to better understand the psychological antecedents of such beliefs and answer the question: Who are conspiracy theorists?
Methods: We use classification and regression tree models to explain individual beliefs in specific conspiracy theories, employing a large set of known correlates of conspiratorial thinking.
Results: Depending on the characteristics of the conspiracy theory employed on the survey, we find that political orientations and conspiratorial thinking provide the most analytical leverage in predicting individual conspiracy beliefs. Furthermore, paranormal beliefs were more predictive than previous literature suggests, while psychological biases demonstrated very limited predictive utility.
Conclusions: The psychological antecedents of conspiracy beliefs used to explain those beliefs vary considerably by the stimuli or events at the center of a given conspiracy theory. Therefore, disproportionately favoring one type of conspiracy theory on one's survey may result in inferences about conspiracy theorists that do not translate across studies. Furthermore, though we are not yet capable of fully determining who conspiracy theorists are, conspiratorial thinking, paranormal beliefs, and political orientations are more predictive of particular conspiracy beliefs than other attitudes, predispositions, and orientations.
---
A great deal of recent research has demonstrated that conspiracy beliefs are a central component of modern culture and politics (e.g., Brotherton, 2015; Oliver and Wood, 2014; Uscinski and Parent, 2014), rather than indicators of radical alienation or psychopathology. More commonplace than once thought, belief in conspiracy theories can have negative effects on individuals and democratic society. For example, mere exposure to conspiracy theories has been found to reduce individual intention to vote and make political donations (Butler, Koopman, and Zimbardo, 1995), participate in politics more generally (Jolley and Douglas, 2014), and place trust in government services and institutions (Einstein and Glick, 2015). Thus, understanding conspiracy theories and those who believe them has positive implications not only for “conspiracy theorists,” but also for those around them, as well as normative implications for theories of culture and government. The rapidly expanding, albeit nascent, literature on conspiracy theories has identified a substantial set of attitudes, predispositions, orientations, and psychological mechanisms that are related to conspiracy beliefs. One characteristic shared by many of these studies is a focus on one, or a small subset of, correlate(s) of conspiracy beliefs. It is perfectly reasonable and expected that a young literature may move forward in such a piecemeal fashion as researchers strive to build a theoretical foundation for the subfield. However, such an erratic procession is not without consequential limitations. While we have learned a great deal about the correlates of conspiracy beliefs in a vacuum, we have little empirical foundation for understanding how those correlate work together. Indeed, we have only a tenuous grasp of who conspiracy theorists are.
In this article, we offer information about which factors related to conspiracy beliefs provide the most analytical leverage in correctly predicting individual conspiracy beliefs. Such information is useful for both identifying the relevant characteristics of conspiracy theorists more precisely, and in further developing our theories about the psychological antecedents of specific conspiracy beliefs (Wood, 2017). Ultimately, a firmer understanding of the characteristics of conspiracy theorists can help us understand what about the conspiratorial frame of mind gives rise to the negative effects of conspiracism, and possibly help combat such effects.
To be more specific, this study offers three important developments of recent work on the correlates of conspiracy beliefs. First, we decipher which correlates seem to analytically matter most when a substantial set of previously identified factors are accounted for. In doing this, we find that the most predictive correlates of nine specific and commonly employed conspiracy beliefs are conspiratorial thinking, political orientations, and paranormal beliefs, while measures of several psychological biases and sociodemographic characteristics offer little utility in correctly classifying conspiracy beliefs. Second, we identify the characteristics of specific conspiracy theories that seem to affect which correlates are most predictive of related beliefs. More specifically, we find that the previously identified correlates differentially affect different “types” of conspiracy theories—namely, those with and without partisan political content. Finally, we discuss our general ability to answer the question “who are conspiracy theorists?” and provide practical recommendations for scholars interested in explaining conspiracy theories and those who believe them.
Background
A conspiracy theory is a “proposed explanation of events that cites as a main causal factor a small group of persons (the conspirators) acting in secret for their own benefit, against the common good” (Uscinski, Klofstad, and Atkinson, 2016:2). A conspiracy belief, then, is a belief in a specific conspiracy theory. Perhaps the most obvious theoretical cause of specific conspiracy beliefs is “conspiratorial thinking,” a psychological state that is the amalgamation of “stable individual differences in the general tendency to engage with conspiracist explanations for events” (Brotherton, French, and Pickering, 2013:1). Conspiratorial thinking is a general tendency to view the world a certain way (Brotherton, 2015) rather than any specific trait, predisposition, or set of attitudes. A great deal of recent work on conspiratorial thinking has turned its attention away from specific conspiracy beliefs and toward the goal of identifying this mindset in individuals (e.g., Imhoff and Bruder, 2014; Lantian et al., 2016; Uscinski and Parent, 2014; Wood, 2017). Most of the unique measurement strategies utilized to identify conspiracy theorists ask survey respondents to what extent they agree with claims such as “I think that events which superficially seem to lack a connection are often the result of secret activities” (Bruder et al., 2013), or “Much of our lives are being controlled by plots hatched in secret places” (Uscinski and Parent, 2014).
Conceptually related to conspiratorial thinking are specific psychological biases that may cause—individually or cumulatively—one to view the world through the lens of conspiracism. Patternicity, the predisposition to delineate patterns from what are truly random events, is an obvious hallmark of conspiratorial thinking (Brotherton, 2015). Those individuals who are particularly susceptible to patternicity are more likely to “see” the connections between disparate events. Brotherton and French (2014, 2015) have also identified susceptibility to the conjunction fallacy and intentionality bias as characteristics of individuals with conspiracy beliefs. Indeed, conspiracy theorists are likely to believe in several conspiracy theories that are logically at odds with each other and ascribe intention to perceived actions that are truly accidental or coincidental in nature.
Political scientists focusing on the role of political identities and orientations have found that partisanship and ideological self‐identifications are significantly related to many American conspiracy theories. For instance, Hartman and Newmark (2012) and Pasek et al. (2015) identify partisanship as a key predictor of conspiracy beliefs about the birthplace of Barack Obama. Miller, Saunders, and Farhart (2016) find that left–right political orientations serve as the latent structure that underlies a host of American beliefs in conspiracy theories about climate change, the birthplace of Barack Obama, the 9/11 terror attacks, and electoral fraud. The effects of partisan and ideological orientations also appear to translate in nature, though not exact character, across political context. van Prooijen, Andre, and Pollet (2015) find politically extreme individuals—those who reside on the edges of scales of ideological orientations, regardless of political context—tend to be more conspiratorial than their moderate counterparts.
A host of other individual differences have also been identified as correlates of conspiracy beliefs. Authoritarianism, for example, has long been a fixture in the literature on conspiracy beliefs (Abalakina‐Paap et al., 1999; McHoskey, 1995). Since authoritarian individuals are more likely to blame others for their problems, they are also more likely to believe in conspiracy theories that tend to posit that some unknown force is guiding events. More recent work has also identified paranormal thinking—the tendency to believe in the role of supernatural forces in everyday life, and individuals' ability to engage such forces—as an important correlate of conspiracy beliefs (Darwin, Neave, and Homes, 2011; Oliver and Wood, 2014). Lastly, personality characteristics (Swami, Chamorro‐Premuzic, and Furnham, 2010), anomia (Goertzel, 1994), and self‐uncertainty and belongingness (van Prooijen, 2016) have been found to correlate with conspiratorial thinking.
Finally, some have asserted the role of nonpsychological characteristics, such as age, race, gender, socioeconomic status, and educational attainment, in promoting conspiratorial thinking. Much of this work comes from a more historical, sociological tradition that emphasized “fringe” status in society as a major cause of conspiratorial worldview (e.g., Hofstadter, 1964). Although we are more interested in the psychological correlates of conspiracy beliefs, we nevertheless consider whether sociodemographic characteristics affect conspiratorial thinking in addition to other psychological variables.
We do not purport to have included every known correlate of conspiracy beliefs in the above lists, but we do believe that our review of the literature conveys the simple point that there are a great many of such correlates. And, even though none of the correlates of conspiracy beliefs we have outlined above strikes us as unreasonable or improbable, we are left wondering which of these correlates actually matters for classifying conspiracy beliefs. Very little of the research outlined above considers simultaneously the effects of several of the identified constructs on conspiracy beliefs. Gaining some grasp on this question is important for two reasons. First, it will help us better understand which predispositions, orientations, and attitudes best characterize conspiracy theorists. An understanding of the characteristics of conspiracy theorists will aid us in deciphering the best candidates for causal first‐movers of conspiracy beliefs. This has implications for our theories of the psychological antecedents of conspiracy beliefs, and the measurement strategies we employ to capture and explain such beliefs. Second, this work will help us more fully answer perhaps the most illusive question faced by the literature: Who are conspiracy theorists?