Evolutionary psychologists have been busy proposing explanations for religiosity. Belief in transcendent conscious beings might promote survival, they argue, by instilling hope and optimism. Or it might be a by-product of other naturally selected susceptibilities, such as infant credulity, pattern seeking, or the tendency to attribute strange events to agency.
But last year a certain Gregory Paul announced in Evolutionary Psychology that this project is a false trail. According to Paul, the rapid decline in religious belief in the West shows that it can’t be a hard-wired psychological propensity.
Paul first ignited controversy in 2005 with an article in the Journal of Religion and Society arguing that religiosity was highly correlated with high rates of homicide, drug addiction, teenage pregnancy and other social disorders. Despite his explicit protestations to the contrary, critics initially inferred that Paul was blaming religion for social problems. In fact his aim was merely to refute the conventional wisdom that religion is socially beneficial, and the best way to achieve this was by showing that the relationship is if anything negative.
In the 2009 article, Paul went on to propose a causal relationship, but with the direction of causation running the other way: that is, religion is a response to social dysfunction and anxiety. This is evident, he argues, from the observation that religiosity has retreated in response to three pressures: (1) the march of science and rational thought; (2) increased economic security including employment protection and free medical care; and (3) the ascendancy of materialism as a legitimate outlook. These factors neatly explain why religiosity has declined so much in Europe, Japan and Australia, and why the USA is such an outlier.
In the US, says Paul, ‘half a million people go bankrupt because of medical bills.’ ‘You can lose your middle lifestyle in a year or less’. Nonetheless, materialism will be an irresistible force when that obstacle has been removed. As he puts it, ‘religious conservatives have lost the culture war’.
Paul’s main finding, captured in the second graph, is the correlation between two synthetic variables, a ‘Successful Societies Scale’ and a ‘Popular Religiosity Versus Secularism Scale’. I don’t have the time or expertise to evaluate these variables or the statistical properties of the correlation. Would-be debunkers point out that Paul is neither a sociologist nor a statistician — he builds model dinosaurs for films like Jurassic Park — but the hypothesis is not self-evidently ridiculous, and his correlations look like a good starting point for a deeper empirical analysis. The broad thrust is consistent with at least one reputable looking study (which Paul himself surprisingly doesn’t seem to have discovered), by Gill and Lundsgaard, who conclude:
It is quite apparent that there is a strong statistical relationship between state social welfare spending and religious participation and religiosity. Countries with higher levels of per capita welfare have a proclivity for less religious participation and tend to have higher percentages of non-religious individuals. People living in countries with high social welfare spending per capita even have less of a tendency to take comfort in religion, perhaps knowing that the state is there to help them in times of crisis.
I have a couple of reservations. First, I’m troubled by the either/or framing of the theory. The fact that religion can be ‘shrugged off’ doesn’t discount the role of evolved psychological factors. Societies can shrug off diabetes, but the disease nonetheless has a heritable, biological basis. According to Michael Shermer, ‘a study of 53 pairs of identical twins and 31 pairs of fraternal twins, each reared apart, researchers at the University of Minnesota, Minneapolis looked at five different measures of religiosity and found that the correlations between identical twins were typically double those for fraternal twins’.
Second, given that there is considerable variation in economic insecurity within countries, I’d want to see evidence that this variation explains variations in the intensity of religious belief in a given country. When you adjust for education, for example, are wealth and good health insurance reliable predictors of an individual’s religiosity?
In any case, given the weight Paul puts on health insurance, it will be interesting to see whether the US experiences a sharp decline in religiosity as Obama’s health reforms take effect, given that, according to the Congressional Budget Office, ‘the combined effect of enacting H.R. 3590 and the reconciliation proposal would be to reduce the number of nonelderly people who are uninsured by about 32 million’. With ten percent of the population experiencing a considerable increase in economic security, one might well expect to see a sharp decline in religiosity in the US in the next, say, two decades.
(Which of course isn’t the only reason for supporting health care reform.)