Maybe there are no atheists in foxholes, as William T. Cummings famously said. But who wants to live in a foxhole? Most of us would prefer a room with a view. So if religion comforts the oppressed and miserable, it should follow that a lessening of human misery should lead to a drop-off in religious belief. That is, in any event, the conclusion of this paper, published recently in The Journal of Personality and Social Psychology. It analyzed self-reports from hundreds of thousands of people worldwide and found that there is indeed a connection between religious faith and happiness—but only in troubled societies.
That's a major revision to a common social-science line on religion, which is that it's good for individuals and for societies. It's often noted, for example, that in the United States people who report themselves religious also report themselves happier, and turn out to be healthier and more generous to charities, than secular people.
To the authors of the new paper, that oft-repeated fact suggested a question: If religion is so great, why do people leave it behind in countries like the Netherlands (40 percent non-religious) or the Czech Republic (60 percent Godless) or the United States (fastest growing religious affiliation: "None")? The answer, they write, is that religion only has these good effects when a society is in trouble.
Foxhole conditions—fear, hunger, want, sickness—do indeed promote faith, write Ed Diener, Louis Tay and David G. Myers. If you're poor or hungry, if you're sick or if you feel unsafe at night, you're more likely to be religious. More interestingly, though, they found that this relationship between circumstances and religion is truly a social phenomenon: In a poor and dangerous nation, even the privileged rich, who aren't personally menaced, are more religious. (And, similarly, in the United States, it's states with lower life expectancies and higher poverty rates that have the largest proportion of religious people.)
In safer, richer societies religion doesn't play much of a role in happiness, write Diener et al. So in places like Sweden and France, religious and non-religious people reported about the same levels of positive feelings and nearly the same sense of social support, respect from others, and conviction that life is meaningful. In Egypt or Bangladesh, though, people who were religious reported much higher levels of positive feeling than did those who were not.
These results, based on people's self-reports, jibe with this model, developed by physicists, of how religions die out. As I wrote last spring, that model simplified society into two communities—a large one of believers and a small one of non-believers—and assumed that religious commitment is a matter of self-interest. This admittedly very simple scheme predicted that the two groups can co-exist for a long time, but that religious adherence can collapse quite quickly. "The model predicts that for societies in which the perceived utility of not adhering is greater than the utility of adhering, religion will be driven toward extinction," wrote those authors.
I think Diener et al. also have a wider significance. If religion is good for some people and not for others, in accord with patterns that science can discern, then religion is neither a menace to humanity nor an essential part of being human. Which means we secularists needn't side with militant atheists like Richard Dawkins and Daniel Dennett (whose hostility to faith borders on a kind of madness) nor fall into an apologetic crouch, in which we accept without argument the notion that once religion is gone, something important is missing. As religion fades away in many wealthy societies, it's time to consider the possibility, as some of these writers do, that religion is still fading, and that when it does, nothing is missing.
Illustration: from Inquisition Tribunal, by Goya, via Wikimedia
Diener, E., Tay, L., & Myers, D. (2011). The religion paradox: If religion makes people happy, why are so many dropping out? Journal of Personality and Social Psychology DOI: 10.1037/a0024402