Conspiracies and their believers – Clear Language, Clear Mind

Summary

Conspiracy theory has negative connotations of falseness, implausiblity or irrationality, but we also know many conspiracies are true, ranging from the ordinary criminal case (conspiracy to commit murder) to uncovered elite political conspiracies. From an epistemic perspective, then, the term must be kept neutral.
American conservatives and Trump supporters believe more in conspiracy theories, no matter whether these are kept vague or consist of a broad selection of options.
Cross-national data show that this is not a consistent finding. Rather, there is wide variation between countries, and even within countries in a short time span. This suggests the current political situation causes large changes in which kind of voter currently believe more in conspiracies. This means one should look more for these causes and less for individual causes of such belief with regards to the association with ideology.

Main text

What is a conspiracy theory? Thinking of how the word is used, we might note that it’s a hypothesis or theory about how some person or persons try to accomplish something illegal or morally wrong in a secret manner. Usually, the connotation being that such theories are unlikely to be true or irrational to believe in. The connotation is also why it is a term of abuse, such as when Richard Hanania calls Trump supporters prone to conspiracy thinking. Kieran Zimmer provides the historical background:

The term “conspiracy theory” goes back to the late 19th century. Its use became more common after it was discussed by Karl Popper in his 1945 book The Open Society and its Enemies. In his book Popper describes and disparages the “conspiracy theory of society”, which holds that “whatever happens in society—especially happenings such as war, unemployment, poverty, shortages, which people as a rule dislike—is the result of direct design by some powerful individuals and groups.”⁽¹⁾ He claims this belief is widely held, and it is clear that at that the time of publication “conspiracy theory” did not carry the connotation of being ludicrous and absurd, as it does today. The stigmatization of conspiracy theories began in the 1960s, as did use of the term “conspiracy theorist”. This was in large part a response to criticisms of the Warren Commission report into the assassination of JFK¹, as is demonstrated by a CIA document describing how its propaganda assets could attack the claims of critics. Some suggested strategies include claiming that those promoting alternative theories were politically or financially motivated, or that “parts of the conspiracy talk appear to be deliberately generated by Communist propagandists”. These same techniques still see widespread use today to censure anti-mainstream narratives.

(Chart showing Google Ngrams data on the usage of terms “conspiracy theory”, “conspiracy theory of society”, and “conspiracy theorist” in books from 1870 to 2000. Some percentages are multiplied by the constants shown in order to facilitate visualization. After 2000 usage of “conspiracy theory” increases, and “conspiracy theorist” increases dramatically.)

In light of this background, we might want to steer clear of the psychological connotations and define it in a neutral manner, that is, without the connotation of likely being wrong or irrational. To see why this is necessary, think of how often the police uncovers and prosecutes people for conspiracy to do something illegal, say, murder. These cases are mostly true conspiracies, someone really was planning to murder someone. As such, we know that not all conspiracy theories are false or irrational, rather, they are commonplace and even boring.

But, you say, most things we call conspiracy theories are those relating to political actors in some broad sense that includes politicians like Trump/Biden as well as NASA (moon landing), not only to mere ordinary criminals. Still, even for such theories, there are many, many true historical cases. They may have started out as wild guessing, but were later proven e.g. using unsealed government documents (e.g. from CIA or FBI). For instance, it is true that Richard Nixon authorized spying on his political opponents (Watergate), and it is true that CIA experimented on humans to produce a kind of truth serum for interrogations (MKUltra). The Guardian, not normally a conspiracy theory outlet, has an article from this year listing various proven conspiracies. You can easily find many other such lists. So, epistemically, we cannot just assume that something being a conspiracy theory renders it false or overwhelmingly likely to be false. One must have a closer look at each one. My personal heuristic is just asking how many people are part of the conspiracy and must keep quiet. Humans are not very good at this, so the more people who are in on it, the more likely for it to have already leaked and been conclusively proven.

In the present day political situation, conspiracy theories are associated with right-wing politics, at least in the USA. Is this true? You can find several hundred academic articles saying yes. The problem with them is that it is very easy to produce the result one wants by unfairly sampling conspiracy theories. This is the same thing social psychologists did with ‘prejudice’, where they merely measured how much different kinds of Americans like different groups in society (ratings of various groups from 0-100). In their surveys, they unsurprisingly found that American conservatives like Blacks, homosexuals, foreigners etc. less than American liberals do. Then they equated this with ‘prejudice’ and concluded conservatives were more prejudiced. This won’t do of course. One cannot show that someone is prejudiced by the mere fact that they don’t like some group. Maybe they have good reason to do so, which makes it not pre-judice: “preconceived opinion that is not based on reason or actual experience” as one dictionary puts it. But in any case, if one uses an improved version of this method and have people rate their own and other races in USA, the result is that Europeans are the least prejudiced, indeed, the only race group that isn’t much prejudiced:

But the second reason with these studies (1000s of them) is that they didn’t include a representative group of target groups for people to rate. One researcher finally used a representative set of targets. The findings revealed that there was no particular difference in ‘prejudice’, except for who likes whom. Decades of work wasted. Here’s how results look like (positive values = left-wingers rate higher):

One can argue about which groups should be included in order to make the list truly representative of “all possible social groups”, but I think this is a fairly reasonable list. With this list, there’s not that much difference between the political sides.

How does this look like if one does the same thing for conspiracy theories? One has to begin with a representative list of conspiracy theories. Almost all studies make the same mistake as the social psychologist prejudice studies, and mainly include conspiracies concerning left-wing persons or institutions, thus coloring the results. However, this recent study attempted to include a very broad list of theories:

Uscinski, J., Enders, A., Diekman, A., Funchion, J., Klofstad, C., Kuebler, S., … & Wuchty, S. (2022). The psychological and political correlates of conspiracy theory beliefs. Scientific reports, 12(1), 21672.

Understanding the individual-level characteristics associated with conspiracy theory beliefs is vital to addressing and combatting those beliefs. While researchers have identified numerous psychological and political characteristics associated with conspiracy theory beliefs, the generalizability of those findings is uncertain because they are typically drawn from studies of only a few conspiracy theories. Here, we employ a national survey of 2021 U.S. adults that asks about 15 psychological and political characteristics as well as beliefs in 39 different conspiracy theories. Across 585 relationships examined within both bivariate (correlations) and multivariate (regression) frameworks, we find that psychological traits (e.g., dark triad) and non-partisan/ideological political worldviews (e.g., populism, support for violence) are most strongly related to individual conspiracy theory beliefs, regardless of the belief under consideration, while other previously identified correlates (e.g., partisanship, ideological extremity) are inconsistently related. We also find that the correlates of specific conspiracy theory beliefs mirror those of conspiracy thinking (the predisposition), indicating that this predisposition operates like an ‘average’ of individual conspiracy theory beliefs. Overall, our findings detail the psychological and political traits of the individuals most drawn to conspiracy theories and have important implications for scholars and practitioners seeking to prevent or reduce the impact of conspiracy theories.

Their list is rather long, 39 theories:

The authors fit models for each of the beliefs, as well as a general index of them. Overall, they found that conservatism correlated slightly positively with belief in these:

Each distribution shows the correlations with each of the conspiracies. So for instance we see that how much people like Trump (feeling thermometer rating 0-100) correlates on average 0.20 with belief in these 39 theories. Liking of Biden, on the other hand, correlates on average -0.19. Political ideology correlated 0.09 (conservatives are higher). Clearly, then, Trump likers believe these various theories more. These results also hold up if one controls for the other personality traits they measured:

The effect size is halved in strength for Trump likers and almost entirely removed for ideology (0.02), but the divergence is still there: Trump likers average a standardized beta of 0.10 and Biden likers average a beta of -0.05. We can also look at each belief individually to look for patterns:

Many results are unsurprising. For instance, people who like Biden more are much less inclined to think Obama has a fake birth certificate (birther), the correlation looks to be about -0.50. On the other hand, this correlates about 0.60 with liking Trump. Of course, this is a directly politically motivated theory, so a big partisan split is expected. Similarly, we can look at the theory that Republicans steal elections, which correlates about 0.40 with liking Biden, and -0.40 with liking Trump. More interesting are relatively neutral theories like whether the government is poisoning the citizens with fluoride in the drinking water. Even this one correlates a slightly positively (about 0.10) with liking Trump.

Still, we might be concerned that the balance of theories is not exactly fair. Maybe some more left-wing theories were not included, for instance, systemic racism theory (Whites are conspiring to keep Blacks and others down), which enjoys widespread acceptance among the American left. To avoid this issue, the authors also had a kind of general conspiracy scale, which is what is called ACTS in the figures above. It stands for American Conspiracy Thinking Scale, and just has 4 questions:

Much of our lives are being controlled by plots hatched in secret places.
Even though we live in a democracy, a few people will always run things anyway.
The people who really “run” the country are not known to the voters.
Big events like wars, the current recession, and the outcomes of elections are controlled by small groups of people who are working in secret against the rest of us.

It sounds like elite theory with added secrecy, but is vague on who is doing the conspiring and against whom exactly, aside from the voters. Unsurprisingly, this general elite conspiracy score correlates quite well with the various different beliefs, on average 0.44. So how well does ACTS correlate with their other variables? Or in other words, what kind of person believes in general elite conspiracy theories?

Liking Trump correlates about 0.35, so pretty well, and liking Biden correlates at -0.25. We get the same symmetry even when we leave out the explicit politics and who is doing what to whom aspect. No matter how it is sliced, it seems liking Trump or being more conservative (ideology in figures above) correlates positively with belief in various conspiracy theories.

For good measure, I downloaded the raw data, calculated an overall conspiracy factor score based on the 39 theories, and plotted it against people’s ratings of Trump:

One more idea could be that it is not liking Trump per se, but liking Trump more than Biden, the net Trump score. However, this difference score also correlates 0.40 with the conspiracy theory factor score, and 0.33 with the general elite conspiracy theory scale.

So no matter how we look at the data, American conservatives and Trump likers in particular believe more in conspiracy theories. How does it look around the world? Is this a general conservatism psychological feature, or more of a time and place result? Fortunately, we have a study to answer this too:

Imhoff, R., Zimmer, F., Klein, O., António, J. H., Babinska, M., Bangerter, A., … & Van Prooijen, J. W. (2022). Conspiracy mentality and political orientation across 26 countries. Nature human behaviour, 6(3), 392-403.

People differ in their general tendency to endorse conspiracy theories (that is, conspiracy mentality). Previous research yielded inconsistent findings on the relationship between conspiracy mentality and political orientation, showing a greater conspiracy mentality either among the political right (a linear relation) or amongst both the left and right extremes (a curvilinear relation). We revisited this relationship across two studies spanning 26 countries (combined N = 104,253) and found overall evidence for both linear and quadratic relations, albeit small and heterogeneous across countries. We also observed stronger support for conspiracy mentality among voters of opposition parties (that is, those deprived of political control). Nonetheless, the quadratic effect of political orientation remained significant when adjusting for political control deprivation. We conclude that conspiracy mentality is associated with extreme left- and especially extreme right-wing beliefs, and that this non-linear relation may be strengthened by, but is not reducible to, deprivation of political control.

They used only a short measure of general conspiracy thinking, devoid of particular politics like the ACTS used above, which they call the Conspiracy Mentality Questionnaire (CMQ):

I think that many very important things happen in the world, which the public is never informed about
I think that politicians usually do not tell us the true motives for their decisions
I think that government agencies closely monitor all citizens
I think that events which superficially seem to lack a connection are often the result of secret activities
I think that there are secret organizations that greatly influence political decisions

Their results:

The results are all over the place, literally. Overall, there is a slight correlation with conservatism across countries, either 0.12 or 0.07, of which only the former is beyond chance. So, it doesn’t appear that conservatism per se has much to do with belief in conspiracy theories, rather it depends on time and place.

Some countries show extreme divergence in results despite the two surveys being administered only about 1 year apart, e.g. for Hungary where study 1 showed a strong association with conservatism, and study 2 showed an even stronger association with socialism. The authors offer some plausible speculations about these cases:

As the attentive reader will notice, heterogeneity was present not only between countries but also within countries when comparing the two studies. We can only speculate about the exact reasons for this. One obvious candidate might be different compositions of the samples or slightly different recruiting strategies (for example, study 2 matches demographic population parameters but focuses on self-selected participants in panels on political research, probably with an above-average interest in politics). Another, equally speculative possibility is that these associations are more volatile than is commonly assumed. Although, in general, conspiracy mentality is a relatively stable disposition, political events and rhetoric of political elites might fuel the endorsement of such worldviews and affect the course/direction of conspiratorial beliefs of citizens.

We will illustrate the latter point with two examples. Let us consider the case of Romania. During the collection of data for study 2 (February–May 2018), the governing (leftist) government party (PSD) changed criminal procedure. The alleged aim was to fight a deep state orchestrated by George Soros, while in reality arguably to save their party leader, Liviu Dragnea, from conviction due to corruption. Throughout that period, leftists in Romania endorsed conspiracy mentality to a greater degree than right-wingers. At the (later) time of study 1 (June 2019), Dragnea had been officially convicted for over a year and the (more common) positive linear relationship was also observed in Romania. On the other hand, such a positive linear relationship was present in the Hungarian data in study 1 (July 2017), but it turned to a negative correlation in study 2 (Spring 2018). This might be another indication for the role of rhetoric by the political elites. Namely, 2017 was a pre-electoral year in Hungary, and Viktor Orban and his right-wing party intensified their attacks on the ‘Soros mafia’ and ‘Brussels‘ and gained more and more control over the Hungarian media⁷⁶. At the same time, in the run-up to the elections in April 2018, conspiracy narratives became abundant on the left, with narratives such as the Hungarian Prime Minister being an agent of Vladimir Putin and leading Fidesz politicians secretly taking psychiatric care in Austria. In line with study 2, another public opinion poll from after the elections (autumn 2018) found strong conspiracy narratives on the left and, also, higher conspiracy mentality among left-wing opposition than among governmental voters⁷⁷. Such post hoc explanations of unexpected differences (albeit indicative) remain speculative, but they might serve as welcome inspiration for further explorations on the role of sample characteristics and political elite rhetoric in future studies.

I think this is right. Dedicated truthers aside, most people don’t pay too much attention to such theories unless they are salient in their environment at that time. This can happen because of claims of unfair losses (election fraud or interference claims, as has happened in every last American election I can remember!), or due to other prominent political issues in the national media. Who is doing the believing then can radically change, and does not really reflect information about which kind of voter is more conspiratorial from a psychological perspective in some deeper sense.

If we accept the time and place view, we would say that the reason American conservatives are currently much more conspiratorial is a combination of facts:

In 2024 (but not in e.g. 1980) Republicans have somewhat more uneducated and low intelligence voters, who are more irrational and prone to believing crazy things.
Republicans have a crazy presidential candidate with a personality cult.
Democrats hold the power in almost all areas of American society, so they are the ones that would be doing the political conspiring, and those the theories would tend to concern.
Republicans also lost the last election, resulting in a “conspiracy is for losers” effect.

Here it would be interesting if one could look at survey data for the same general conspiracy questions over time, so one could test these various factors in the American context. I looked at the General Social Survey (GSS), but it doesn’t seem it has suitable questions. Neither does the American National Election Studies (ANES). If you know of other long-running survey collections, please let me know in the comments.

This time I put the summary on top, let me know which you prefer (bottom summary vs. top summary). Generally, summarizing findings at the end makes sense from a reading and memory building perspective, but for those looking to get a summary of what the post is about before deciding whether to read it or not, on top makes more sense.