![[Adobe Stock]](https://sp-ao.shortpixel.ai/client/to_webp,q_glossy,ret_img,w_750,h_375/https://www.psypost.org/wp-content/uploads/2025/08/conspiracy-theorist-750x375.jpg)
A new study published in the European Journal of Social Psychology provides evidence that believing in one conspiracy theory can slightly increase the likelihood of believing in others over time. While the effects were small, the findings support a long-standing theory that some people may develop interconnected systems of conspiratorial thinking, where one belief helps reinforce others.
Although previous studies had shown that people who believe in one conspiracy theory tend to believe in others, there was little direct evidence about how these beliefs develop over time. One influential theory, proposed by psychologist Ted Goertzel in 1994, suggested that conspiracy beliefs could form a closed system, where each belief supports and amplifies the others.
This model has been widely cited, but its central claim—that believing in one conspiracy theory increases the likelihood of adopting others—had not been tested in a way that could establish causality. To address this gap, a group of researchers from institutions in New Zealand, Australia, the United Kingdom, and Germany conducted two extensive longitudinal studies to examine how beliefs in various conspiracy theories influence each other over time.
“There’s been a lot of research showing that people who believe one conspiracy theory are more likely than average to believe other conspiracy theories. A well-known theory suggests this is because developing a belief in one conspiracy theory causes people to develop beliefs in others,” said study author Matt N. Williams, an associate professor at the School of Psychology at Massey University.
“This is the popular idea of ‘going down the rabbit hole.’ But despite a lot of discussion about this idea in the academic literature, there haven’t previously been any studies on the topic with a credible way of testing causal effects. I realized that by doing longitudinal studies it’d be possible to test whether people who develop a belief in one conspiracy theory are more likely to subsequently develop beliefs in other conspiracy theories.”
In the first study, the team followed 498 participants from Australia and New Zealand across seven monthly survey waves. Participants answered questions about their agreement with 10 well-known conspiracy theories, such as beliefs that the September 11 attacks were an inside job, that fluoride is added to water to control minds, and that COVID-19 vaccines contain microchips. Responses were given on a five-point scale ranging from strong disagreement to strong agreement. Most participants disagreed with most of the theories, but there was still enough variation in responses over time to examine patterns of change.
The researchers used a statistical approach known as a random intercept cross-lagged panel model, or RI-CLPM. This method allowed them to isolate changes within individuals over time, separating those from stable personality traits that might predispose someone to conspiracy beliefs in general. They then analyzed whether an increase in belief in one conspiracy theory in a given month was followed by increases in belief in other theories the next month.
The results showed that on average, small increases in belief in one conspiracy theory were followed by small increases in belief in others. The individual effects were tiny—typically a few hundredths of a point on a five-point scale—but consistent across the data. While none of the individual cross-lagged effects were statistically significant after correcting for multiple comparisons, the average effect across all 90 possible pairings was positive and significant. This pattern supported the idea that, for at least some people, adopting one conspiratorial belief can make it more likely they’ll adopt others in the future.
To confirm the results and strengthen the causal interpretation, the researchers conducted a second, preregistered study. This study included a larger and more diverse sample of 980 participants from the United Kingdom, Australia, and New Zealand. The data were collected over 13 consecutive months, and the survey included 11 conspiracy theories—some overlapping with those in the first study and some newly added. These included beliefs that climate change is a hoax, that genetically modified foods are dangerous and being covered up, and that the COVID-19 pandemic was used as a pretext to restrict civil liberties.
Again, the researchers used RI-CLPM to analyze the data. They preregistered their hypothesis that belief in any one conspiracy theory would positively influence belief in all others over time and laid out strict criteria for what would count as support or refutation of that hypothesis.
According to their preregistered plan, even a single negative and statistically significant cross-lagged effect could have falsified their theory. But in the results, none of the effects were significantly negative. One positive effect remained statistically significant even after a strict correction for multiple comparisons, and the average cross-lagged effect was again small but positive and statistically significant.
“The results suggest that developing a belief in one conspiracy theory probably does make you more likely to subsequently develop beliefs in other conspiracy theories, but only very slightly!” Williams told PsyPost. “The effect sizes we found were statistically significant but very small.”
“This lines up with one of our previous studies showing that people’s levels of belief (or disbelief) in conspiracy theories tend to be quite stable over time. In that paper, we concluded: ‘People developing beliefs in a succession of conspiracy theories are often characterised as falling down a ‘rabbit hole’—a reference to Lewis Carroll’s Alice’s Adventures in Wonderland. It seems that if conspiracy theorists do fall down a rabbit hole, it is typically one with a rather gradual slope.'”
As with all research, there are limitations to note. One concern is that the findings could be influenced by factors that change over time and affect belief in multiple theories, such as exposure to new information or political events. These time-varying influences were not controlled in the analyses, meaning they could potentially explain some of the observed effects.
“Our paper was specifically focused on estimating causal effects,” Williams noted. “We did collect longitudinal data, and applied a form of analysis that can rule out one type of confounding variable—confounding variables that are stable over time. However, our studies definitely aren’t as conclusive as a true experiment would be, and it is possible that our estimates of causal effects could be biased (upwards or downwards!)”
“It’s also worth mentioning that the two studies in this paper are about the effects of beliefs in different conspiracy theories on one another. They weren’t designed to tell us anything about other consequences of belief in conspiracy theories. That’s a different question, and other researchers have published on this topic. But to me, the fact that people are willing to entertain conspiracy theories isn’t a bad thing in and of itself; real conspiracies do happen! That said, when beliefs about conspiracies become unmoored from reality, negative consequences can take place.”
Another limitation is that the surveys may not accurately reflect participants’ true beliefs. Factors such as social desirability bias, inattentiveness, or insincere responding could have distorted how people reported their agreement with the conspiracy theories.
“People sometimes claim to believe a conspiracy theory in a survey when they don’t really believe it (e.g., because they’re making a joke or trying to troll us),” Williams explained. “Such insincere responses could possibly distort some of our findings in this research area. For example, it could inflate estimates of the prevalence of beliefs in conspiracy theories. Some of my collaborators and I are therefore working on a study where we aim to find out more about insincere responding and how it affects research on conspiracy theories.”
“I think it’s important that research on conspiracy theories (or any topic, really) is done in a transparent fashion. So we shared open data and materials along with this paper, so readers can check our analyses and replicate our work. We also preregistered Study 2, so that readers can have more confidence that we didn’t ‘cherry pick’ our findings. The paper itself is also open access.”
The study, “Does Developing a Belief in One Conspiracy Theory Lead a Person to be More Likely to Believe in Others?,” was authored by Matt N. Williams, Mathew D. Marques, John R. Kerr, Stephen R. Hill, Mathew Ling, and Edward J. R. Clarke.