Atheists: The Puppy Kicking, Chicken ‘Loving,’ Serial Killing Cannibals Next Door?

Sarah R. Schiavone

Atheists don’t win many popularity contests. Rather, nonbelievers are generally disliked. But just how severe are popular negative attitudes toward atheists?

First, atheists are one of the least electable groups, with only 58 percent of Americans saying they would vote for an atheist, even if he or she were a well-qualified candidate and nominated by his or her own preferred political party. Of course there are plenty of candidates many of us wouldn’t vote for … but to entirely cross off every single member of a given group? A little harsh. Candidates may have policies we disagree with, but personal preference becomes prejudice when individuals are judged not on one’s merits but according to group memberships. Thus, refusing to vote for an atheist candidate—that is, crossing off a group that might make up a quarter of U.S. adults, according to some recent research—seems problematic. Beyond not voting for atheists, roughly half of Americans would be unhappy if someone in their family married an atheist. But then, people are probably fairly selective when it comes to who they want their kids to marry.

Self-reports of voting or marriage eligibility provide valuable information about the prejudices people are willing to admit to holding. Yet decades of research in psychology indicate that people rarely want to be viewed as prejudiced by their peers. Although everyone likely holds some forms of prejudice, most folks strive to appear egalitarian and unprejudiced. This presents a difficult problem for researchers studying stereotypes and prejudice: How do we measure attitudes that people are reluctant to admit to others, perhaps even to themselves? To investigate the attitudes people are less willing to admit, indirect measures of prejudice (rather than self-reports) might be more informative regarding attitudes toward various stigmatized groups and identities.

In other words, to gauge negative attitudes toward various people in society, it is often necessary to delve deeper than people’s openly admitted attitudes and consider more subtle ways in which prejudicial attitudes might reveal themselves. People may overtly and explicitly appear nonprejudiced but nonetheless hold prejudices at an intuitive level. People rely on various heuristics, or mental shortcuts, when making quick judgements. These intuitions enable people to respond quickly and automatically while generally saving the time and energy that would be required for more deliberative rational thinking. However, relying on heuristics can also have more sinister consequences, such as when the color of a suspect’s skin subtly biases a police officer’s split-second decision to unload a firearm when an innocent suspect reaches into the glovebox of his or her car. Overall, heuristics may help us make split-second judgements that are often useful, but they are also often inaccurate and imprecise.

One commonly used shortcut is the representativeness heuristic, or the tendency to categorize events or people based on how similar they seem to our assumptions about certain events or groups (for instance, a stranger looks like what we imagine criminals to look like, therefore we assume he or she is probably a criminal). The use of the representativeness heuristic can be observed through an odd psychological trick called the “conjunction fallacy,” famously used in the Linda Problem developed by Daniel Kahneman and Amos Tversky. The Linda Problem presents folks with the following scenario and question:

Linda is thirty-one years old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice and also participated in anti-nuclear demonstrations.

Which is more probable?

a) Linda is a bank teller.

b) Linda is a bank teller and is active in the feminist movement.

Considering basic probability, the second choice—specifying the conjunction of two events occurring (Linda being a bank teller and Linda being active in the feminist movement) is always less probable than the first option that only specifies one of the events occurring (Linda being a bank teller). Despite the laws of probability, most people choose the second option. What gives?

Choosing option b reflects the operation of a heuristic whereby people intuitively match the group implied in the second option (active in the feminist movement) to the provided description of Linda. However, people only tend to select the second option when there is an intuitive match between the description and the implied group. Imagine you are given the same description of Linda and then asked whether it is more probable that Linda is a) a bank teller, or b) a bank teller who is an avid big game hunter. Still tempted to pick option b? Almost nobody is, because there is no intuitive match between Linda’s description and the stereotype of big game hunters. This makes the representative heuristic a useful tool for researchers to explore intuitive stereotyping and prejudice, including anti-atheist prejudice. More specifically, researchers can examine the types of behaviors that evoke conjunction errors (selecting option b), reflecting an intuitive association between atheists and a given description. Which leads to a pretty straightforward question: What behaviors do people intuitively associate with atheists?

Selfish and Criminal Behavior

Imagine Richard, who backs his car into a van. He pretends to leave a note with his information. Later that day, Richard finds a wallet on the sidewalk, takes the money, and drops the wallet in a trash can. In 2011, Will Gervais, Azim Shariff, and Ara Norenzayan gave participants this description and then varied the group implied by option b in a representative heuristic task. People were quite likely to indicate that it is less probable that Richard is a teacher than that he is a teacher and an atheist … far more likely than they were to display the conjunction error when the latter option implied that Richard might be either a Christian or a Muslim. Indeed, only one group evoked a comparable rate of conjunction errors. Participants intuited that the perpetrator of insurance fraud and theft in the provided description was either an atheist … or perhaps a rapist.

Kicking Puppies

Using the same method, a scenario described someone who cruelly made fun of an overweight woman on a bus ride and then kicked a begging dog in the head. Participants rarely committed the conjunction error if it implied that the puppy kicker was gay … yet roughly half of participants associated cruel jokes and animal cruelty with atheists.

Consensual Incest

Yet another iteration of this experimental setup described a brother and sister spending a night having passionate yet safe sex together while traveling. Although they enjoyed it, they decided not to do it again. Few participants assumed the amorous siblings were Buddhists (21 percent conjunction errors), Christians (5 percent), Hindus (10 percent), Jews (3 percent), or Muslims (15 percent). Yet again, half of the participants intuitively associated consensual incest with atheists.


Instead of having sex with his sister, one experiment described a man having (sterile) sex with a dead chicken before cooking and eating it alongside a nice glass of Chardonnay. This behavior was believed to reflect the behavior of atheists (58 percent errors) but not any of the five ethnic groups used for comparison. Gross, but what about other dietary proclivities?


Imagine Catherine, a research assistant in a pathology lab. Bemoaning the waste of simply cremating cadavers after examination, one night she decided to cut off a piece of flesh from a human cadaver to take home and eat. Again, people intuitively assumed that immoral behavior as extreme and bizarre as cannibalism was representative of atheists (43 percent errors).

Serial Murder

Incest and necrobestiality are disgusting. Cannibalism is just wrong. Kicking puppies isn’t nice. But what about behaviors that are clearly and unambiguously harmful and morally heinous? Another description included someone who tortured animals as a child and grew up to murder, dismember, and bury five homeless people in his basement. When given a description of this textbook psychopathic serial killer, people strongly considered this behavior to be representative of atheists but not of any other religious groups.

Global Moral Distrust

Surely such rampant anti-atheist sentiment is just a quirk of contemporary American society, right? Well unfortunately, no. In a recent comprehensive study spanning thirteen diverse countries, it turns out there is evidence of extreme moral anti-atheist prejudice globally—even in more secular countries such as Great Britain, the Netherlands, China, and the Czech Republic. Using the same description of a person who grew up killing animals as a child and became a serial murderer as an adult, people were roughly twice as likely to assume this murderer was an atheist than a religious believer across the thirteen countries. Intuitive moral distrust of atheists was predictably strongest in the more religious societies (such as India, the United Arab Emirates, and the United States); yet, intuitive anti-atheist prejudice was far from confined to the religious societies, as it was evident in every sampled country except Finland! To make matters worse, distrust of atheists was not just evident among religious participants. Surprisingly, across the thirteen countries, even atheists shared the intuition that gross immorality was likely perpetrated by atheists. In other words, even atheists living in many of the most secular societies that have ever existed (such as the Netherlands, the United Kingdom, or the Czech Republic) show intuitive moral distrust of their fellow atheists.

Congratulations, atheist readers! Global majorities intuitively view you as a potentially puppy kicking, chicken “loving,” serial killing cannibal.

In terms of global opinion, it may sound like atheists are well and truly screwed, and perhaps they are. Nevertheless, a sliver of hope can be found in reports that not only are the number of people who identify as nonreligious increasing, but there are even more who may not identify as atheists but nevertheless do not believe in a god (perhaps one in four Americans, according to recent research). Further, self-reported tolerance may be improving, even if only slightly. In a 2017 Pew survey, Americans reported their feelings toward various groups using a “feelings thermometer” by indicating how warm or cold they felt toward each group on a scale from 0° to 100°. Atheists received an average score of 50° out of 100°. The only group given a colder rating was Muslims (48°). While 50° may sound low, it is still somewhat of an improvement for atheists over similar polls taken in previous years. In 2014, atheists were given only an average rating of 41°, suggesting there has been a slight uptick in warmth in recent years.

Now that we have savored a brief moment of potential positivity for atheists, it’s time for another gut punch. People clearly intuitively associate immorality with atheism, but religious people can also behave in strikingly immoral ways. Consider stereotypically immoral acts associated with religious believers (prayer-healers letting children die of preventable diseases, televangelists swindling money, priests molesting children). If a religious figure does something immoral, surely they, too, wouldn’t be perceived as atheists?

To explore perceptions of religious people acting badly, we described a sixty-year-old man who was widely respected and generally viewed as very caring and friendly. However, he actually had spent the last ten years molesting more than thirty boys in his office. Even when given information that the man is a priest, far more people made the error of choosing option b that He is a priest and does not believe in God (57 percent) than those given the option of He is a priest and believes in God (40 percent). These findings suggest that even when there is direct evidence of an individual’s religious commitment (being a priest), his immoral actions were enough to sway intuitions to the point that people view him as a priest who also happens to be a covert atheist.

Ouch …


Further Reading

Sarah R. Schiavone

Sarah R. Schiavone is a graduate student in social psychology at the University of Kentucky and a fellow of the National Science Foundation’s Graduate Research Fellowship Program. Her research focuses on anti-atheist prejudice and attitudes toward science and religion.

Prejudice against atheists is real. It’s global. And it’s worse than you think.

This article is available to subscribers only.
Subscribe now or log in to read this article.