Readers of this blog will know that I’m a big fan of highlighting commonalities between religious and secular people, so I was very interested to see reporting about a study purporting to show very little difference in moral attitudes between Christians and atheist.
I wasn’t content just to take that reporting at face value, however, and it turns out that the paper in question actually said even more interesting things than the reporting mentioned. (Note: This is a longer post than normal because there’s a lot to tease out of the paper.)
The paper is “How do U.S. Christians and Atheists Stereotype one another’s Moral Values?” by Ain Simpson and Kimberly Rios in the International Journal for the Psychology of Religion (August 2016), and Simpson and Rios attempt in it to test several hypotheses regarding Christians’ and atheists’ moral beliefs and those groups’ perception of the moral attitudes of the outgroup.
To test these hypotheses, Simpson and Rios use Moral Foundations Theory (a theory proposed by psychologists Craig Joseph and Jonathan Haidt and popularized in Haidt’s book The Righteous Mind), which postulates five different moral dimensions: Care, Fairness, Loyalty, Authority, and Sanctity. (A sixth, Liberty, has been added more recently but isn’t addressed in this paper.) Simpson and Rios refer to Care and Fairness as the Individualizing foundations and the latter three as Binding foundations.
Much of the research on Moral Foundations Theory (hereafter MFT) has been on political affiliation, with Haidt and other researchers arguing that political progressives emphasize Care and Fairness, libertarians stress Liberty and Fairness, and conservatives have a more balanced concern for all six dimensions. But Simpson and Rios attempt in this study to look at what Christians and atheists think independently of political leanings.
The hypotheses that Simpson and Rios set out to test were as follows:
- Hypothesis 1: Atheists should endorse the Care and Fairness foundations more strongly than Christians; Christians should endorse the Loyalty, Authority, and Sanctity foundations more strongly than atheists.
- Hypothesis 2A: Atheists should perceive their ingroup to endorse Care and Fairness more strongly than Christians and to endorse Loyalty, Authority, and Sanctity less strongly than Christians.
- Hypothesis 3A: Atheists should overestimate Christians’ endorsements of Loyalty, Authority, and Sanctity and underestimate Christians’ endorsements of Care and Fairness.
- Hypothesis 2C: Christians should perceive their ingroup to endorse all five foundations more strongly than atheists.
- Hypothesis 3C: Christians should underestimate atheists’ endorsements of all five foundations.
(A and C in these hypotheses stand for Atheist and Christian, respectively.)
To test these hypotheses, Simpson and Rios used Amazon’s Mechanical Turk program to recruit participants as part of a larger study, eliminating participants for failing an attention check and those who held beliefs in contradiction with their self-label (e.g. self-described atheists who professed belief in a god or self-described Christians who professed no such belief). Simpson and Rios use the operative definition of atheism as “having no belief in gods,” so “atheists” in this study includes self-described agnostics as well.
Participants were then given questions from the Moral Foundations Questionnaire and asked to answer 1) from their own perspective, 2) from the perspective of a “typical atheist,” and 3) from the perspective of a “typical Christian.” Simpson and Rios also took demographic information on gender, political ideology on an 11-point scale, and religious identification.
Do Atheists and Christians Actually See Morals Differently?
For the first hypothesis, Simpson and Rios compare participants’ own judgments on their moral views. Remember, their hypothesis is that atheists would value Individualizing foundations more than Christians and Christians would value the Binding foundations more than atheists.
The data showed that they were half-right: While Christians did endorse the Binding foundations significantly more than atheists did, Christians actually endorsed the Care foundation slightly more than atheists and there was no real difference in the Fairness foundation.
In their discussion, they note:
To begin with, and consistent with Hypothesis 1, Christians (in Study 1) were much more likely than atheists to endorse the Binding foundations. This is consistent with theorizing by Graham and Haidt (2010), who argue that the Binding foundations are central to distinguishing religious from secular moral systems. Results suggest that Christians are much more likely to value ingroups, authorities, traditions, purity, and self-restraint, and are much more likely to condemn acts that are disgusting or that involve disloyalty, betrayal, disrespect, or insubordination. However, and despite past research (Harper, 2007) indicating that Christians perceive atheists negatively with regard to morally-relevant concerns (e.g., selfish and hardhearted), our findings indicate that atheists have a strong moral sense of compassion and fairness similar in degree to that of Christians. This is consistent with evidence that atheists are no less likely to adhere to concerns pertaining to the Care and Fairness foundations in their beliefs, values, and daily practices. …This is also consistent with Schnell (2015), who found that atheists and believers endorsed humanist ideals (i.e., benevolence, kindness, tolerance, and helping those in need) to similar degrees. (emphasis mine)
Perhaps most importantly, despite noting a general trend toward atheists being liberal and Christians being conservative (which should surprise precisely no one), Simpson and Rios found no change in these results when controlling for political ideology and even said that the difference between atheists and Christians doesn’t particularly resemble the liberal/conservative divide on moral foundations. (This is true of their conclusions generally.)
Do Atheists and Christians Assess the Other Accurately?The answer here according to this data is No, but of course the details are more interesting than that simple answer.
Out of the four hypotheses that dealt with outgroup perception, all but 3C were consistently supported by the responses:
- Atheists perceived no difference in Care but perceived Christians to endorse Fairness much less strongly than they did; they also expected Christians to endorse Sanctity, Authority, and Loyalty more than they do. (Hypothesis 2A)
- Atheists underestimated Christians’ endorsement of Care and Fairness while overestimating their stance on the Binding foundations. (Hypothesis 3A)
- Christians, conversely, perceived themselves to endorse all of the foundations more than atheists, with all but Fairness being perceived as a strong difference. (Hypothesis 2C)
On Hypothesis 3C, Christians underestimated atheists’ endorsement of the Individualizing foundations but, curiously, overestimated atheists’ endorsement of the Binding foundations. Simpson and Rios give a possible explanation in their discussion:
This unexpected result might implicate a role of the False Consensus Effect (Ross, Greene, & House, 1977), which posits that individuals tend to overestimate the extent to which others share one’s attitudes and beliefs. That is, because Christians are an overwhelming majority in the U.S., they may generally overestimate the degree to which their own moral values are shared by outgroups. Hence, they might be prone to slightly over-project their endorsements of these quasi-religious moral concerns even to an outgroup defined by its lack of religious belief.
In Their Own Words
So with this first study, we know what moral foundations atheists and Christians endorse and what they (mis)perceive the other group to value. What we don’t know is whether those perceptions are viewed favorably, unfavorably, or neutrally.
To fill this gap, Simpson and Rios did a second exploratory study, asking respondents to write about their general impression of the outgroup’s moral values and evaluating those responses for clauses that indicated content that reflected the moral foundations. (They did this by having both the first author and a research assistant untrained in MFT code the responses and comparing the two to come to a consensus with both an expert and lay perspective.)
Here’s where it gets interesting: Atheists had far more negative things to say about Christians than Christians did about atheists.
Atheists were about as likely to say positive things about Christians’ endorsement of the Care foundation (e.g. that they want to help the less fortunate) as negative ones (e.g. that they are indifferent to the poor), but they were more negative for all the other foundations. The authors suggest that this is due to atheists’ “pervasive negative stereotypes about and discrimination against their group in the U.S.,” but I would add as well that it is more likely for an atheist to have had bad personal experience being negatively stereotyped by Christians merely because of demographic considerations: More atheists know Christians than Christians know atheists (in part because of atheists being closeted to avoid being mistreated based on negative stereotypes, which is a lovely vicious circle).
For Christians, atheists were described more positively regarding Loyalty (“[atheists] love their families and participate in some community activities”) and more negatively regarding Authority and Sanctity. The authors note a particular hostility regarding the Authority foundation, saying things like “[Atheists] undermine the 10 commandments” and “[atheists are] lawless and destructive.” They hypothesize that this is an explicitly religious perception based on stereotypes of atheists as rebellious, where consideration of non-religious considerations (e.g. respect for secular authorities and traditions) might correct that somewhat.
Many people will not be particularly surprised by these conclusions. I can say that I wasn’t entirely shocked. So how can this be useful?
For one, talking about moral foundations allows us a framework to look at shared moral values. Note here that this is clear data suggesting that Christians and atheists don’t differ all that much in terms of how they value compassion and fairness, despite misperceptions to the contrary. That means that we can look at goals motivated by these values and find some degree of consensus on them, and that gives us something to collaborate on if we’re willing to set aside our differences. (And again, this already happens, so that shouldn’t be a big shock!)
Second, I think this is a reminder to atheists in particular that while we may have good reasons to be angry at religion (and, in the US, at Christianity specifically), there is a degree to which our anger and negative experiences with Christianity may be affecting our ability to perceive the motivations of Christians accurately. I’m not even arguing that people who have been harmed by Christianity should just get over it and work with Christians — that is a space I freely grant to those who have suffered trauma from religion — just that we should try to correct for that in our own individual assessments.
Most of all, I hope that data like this serves as a reminder to all of us that we need to check our ingroup and outgroup perceptions constantly. It’s really easy to let singular events or even apparent trends color our perceptions of both, and that can be an extremely poor way of grounding our own actions. We need data like this as a reality check for the negativity bias that can throw off our judgments.
And all of this sounds like a great reason to further increase our dialogue with people who believe differently than us. Yes, there will be differences in what undergirds our moral beliefs, but there is very likely ample room for agreement, and that space is worth seeking out, especially if we really do care about helping others and making moral progress.