Tom Rees reports on a smart new study which tests the effect of religiosity on attitudes toward torture in the US. Using two different large surveys, the researchers first simply examined the correlation between religiosity and support for permitting torture. Realizing that conservative political ideology can also induce greater support for torture and can itself be driven by religiosity, they then compared the direct impact of religion on torture support with the indirect effect as mediated through political ideology.
What they found is that religiosity (measured in one case simply by how often people attend church, and by also looking at how often people pray, and how important they think religion is in their lives in the other survey) tends to correlate with a decline in support for torture. The correlation isn’t huge, but it is statistically significant, and being religious corresponds to a 9–14% decline in support for torture.
When considering the indirect effect, they found that increased religiosity is associated with about 15% increase in conservatism, and that being conservative makes someone between a quarter and a third more likely to support torture. But that effect varies by education and political engagement more generally. In populations with less education, the correlation of religiosity to conservatism, and of conservatism to torture support, both decline. In the least engaged group, there’s no significant correlation of religion to conservatism at all, and the least engaged group has less than half the correlation between conservatism and torture as the most engaged.
This makes a certain sense, since religiosity tends to decline with education (education being a good proxy for political engagement), and it’s long-established that the folks with high educational attainment who go to church tend to be more politically conservative. That means that the low-education population will have more religious liberals, which ought to cut the correlation of religiosity to conservatism. And it may be that, for cultural reasons, the boundaries of conservatism shift as you move from low-education to high-education populations, such that folks with more liberal attitudes will call themselves conservative and therefore reduce the correlation between torture support and conservatism.
What’s interesting is that, despite these shifts in the nature of religiosity across educational categories, you see almost no change in the correlation between religiosity and support for torture. The shifts in attitude that become clear in the indirect pathway remain almost perfectly constant on the direct pathway. Even though the more educated religious folks are more conservative, the basic statistical effect of religiosity holds steady, in opposition to torture.
Whether this explains the disturbing tendency of certain prominent (and putatively liberal) atheists to defend torture is an open question.
Ariel Malka and Christopher J. Soto (2011) “The Conflicting Influences of Religiosity on Attitude Toward Torture” Personality and Social Psychology Bulletin DOI: 10.1177/0146167211406508