Over the next several weeks I will be reviewing the literature regarding prejudice towards atheists in what will ultimately become a comprehensive literature review. This review, once complete, should suffice to say, “Prejudice against atheists is rampant.” Most, if not all, of these studies have been conducted in the United States, but I will be actively seeking all studies — even any that contradict me — which will, hopefully, examine this issue across the globe.
This week I will be examining “A sociofunctional approach to prejudice at the polls: are atheists more politically disadvantaged than gays and blacks?” by Andrew S. Franks and Kyle C. Scherr from the Department of Psychology, Central Michigan University. This article appears in the Journal of Applied Psychology 2014, 44, pp. 681-691. The authors use four scales to determine which group of political candidates — atheists, gays, or blacks — suffer the greatest disadvantage.
Franks and Scherr begin their study by referencing previous research that has exposed perceptibly systemic prejudice against atheists in all facets of everyday life and against potential atheist candidates for public office. It’s here where they identify the gap in our knowledge and the problem:
[D]espite naturalistic polling indicating a deficit in support for atheist candidates compared with other types of candidates, no research has experimentally examined the magnitude of these deficits and, perhaps more importantly, the content associated with voters’ unwillingness to support atheist candidates (681).
They then point to previous research indicating core prejudice factors of minority groups, which indicate distrust is a recurring theme. They therefore examine the content of political public opinion polls and pit atheist candidates against gay, black, and other minority groups to determine to what extent atheist candidates suffer a polling deficit.
The researchers first examine the wealth of literature in this field and discover a persistent theme among groups who share set religious identities. That is:
The feeling that “God is watching” seems to promote adherence to binding forms of morality (e.g., Haidt, 2007) and shared social norms (Henrich et al., 2010) that serve to both regulate a person’s behavior and give others reason to believe their behavior will conform to social norms even when nobody else is around (682).
In other words, the perception among certain religious groups of an all-watching god tends to paint non-believers as free to break from social norms when nobody is watching them. For example, they cite Christian fears that an atheistic position on the existence of god would lead to “uncontrolled violent and sexual impulses and selfish behaviors that undermine the social order” (682). These findings, they conclude, might help explain why atheists are more likely to be distrusted than other minority groups.
From an evolutionary standpoint, they argue, this makes sense when applied to political candidates. Having a wildcard leader is not an effective way to promote the survival of the group.
Distrust is not necessarily a more important variable to voters than other variables. Gays, they find, suffer prejudice because of the perception of being “potential sources of pathogen contamination due to associations with sexually transmitted infections” and blacks, who are “perceived as threats to physical safety due to associations with violent street crime” (682). Therefore, the researchers develop four hypotheses to determine which group is more disadvantaged at the polls than the others.
Hypothesis 1: Christians will most likely vote for straight white Christians. After this, in descending order they are most likely to vote for a black candidate, a gay candidate, and then an atheist candidate in last place.
Hypothesis 2: Atheists will be deemed the least trustworthy.
Hypothesis 3: Gays will be deemed the most disgusting.
Hypothesis 4: Blacks will be deemed the most threatening.
Methods and Procedures
The researchers then used a sample (n=200) of US citizens gathered from Amazon’s Mechanical Turk online workforce. This sample size is smaller than I would have liked. I would have preferred a sample size about twice as large, but it is sufficient for the purposes of this research.
Participants then completed a survey about a political candidate, only using information about the candidates supplied by the research team:
The candidate was always described as a 39-year-old, male named Greg whose agenda prioritized the economy, health care, and education. Participants were randomly assigned to read about a candidate who was characterized by one of four descriptions: White heterosexual Christian (WHC), Black heterosexual Christian (BLK), White gay Christian (GAY), or White heterosexual atheist (ATH) (684).
The participants rated these candidates along four scales. First, how likely they are to vote for the candidate. Second, third, and fourth, how trustworthy, threatening, and then disgusting the candidate is.
The findings were rather surprising. Not only did atheists score lowest on the distrust-trust scale, they also scored lowest on the threatening-comforting scale and low on the disgusting-appealing scale. Atheists are perceived as untrustworthy, threatening, and disgusting. In contrast, gays and blacks overcame their respective hypotheses either with insignificant correlations or by scoring higher on other scales:
Our results suggest that atheists, compared with gays and Blacks, suffer the greatest disadvantage as political candidates because they received the lowest levels of voting intentions and elicited fear, disgust, and strong levels of distrust from Christian voters. Although distrust may be the core affective component in anti-atheist prejudice (e.g., Gervais et al., 2011), the broad range of negative emotions elicited by the atheist candidate in this study is consistent with other research demonstrating that exposure to atheist ideas causes Christians to feel disgust (Ritter & Preston, 2011) (687)
The study concludes:
Atheism and secularism are particularly under-represented in the political sphere. The current study demonstrates not only the magnitude of anti-atheist prejudice in politics but also the depth and breadth of negative feelings underlying that prejudice (689).
The findings of this research are very interesting. There are some very strong p-values in there, as well as some weaker p-values — the weaker p-values actually strengthen the first hypothesis. Personally, I would like to see this study replicated using a larger sample to see whether or not we get the same results (n=500, for example). Of particular note, their sample was relatively split between Christians and non-religious people. This leaves room for future research: 1) we can replicate the study using only religious people, 2) we can replicate the study using only non-religious people, and 3) we can replicate the study using a sample that accurately reflects the American demographic. These studies might help us better understand to what extent prejudice against atheists is systemic in American religious populations.
Of course, this is only in America. For whatever reason, electing atheists to public office is rather normal in other parts of the world.