Saturday, May 02, 2009

The paradox of the unaffiliated   posted by Razib @ 5/02/2009 02:28:00 PM

There was a recent Pew survey which came out which showed that around half of Americans "changed religions" in their lifetime. Not too surprising, though there is obviously a qualitative difference in switching from being a Methodist to Presbyterian, as opposed to going from Methodist to Buddhist. But one of the interesting findings, which I suspect many will not be surprised by, is that the "unaffiliated," those with no religion, have low retention rates. About half of those raised with no religion affiliate with a religion as an adulthood. This is being spun as a surprise, but I remember from the 1990s the "surprise" by Gallup researchers that 1/2 of those who claimed no religion were raised with no religion. So it seems that this dynamic hasn't changed that much. The "paradox" is that the population with with no religion has doubled in the last generation because of bleeding from Christian affiliation. Obviously this is due to the reality that the religious form 80-90% of the population, and the nonreligious form 10-20% of the population (depending on how you design the survey), so marginal defections from the former can swamp out substantial defection from the latter. This isn't that insightful, and in any case there are religious groups which exhibit the same high "churn," Mormons and Buddhists for example.

Another issue is the use of self-report survey data to analyze why people changed their religion. You have to be careful about this; the ethnographic data suggest that joining another religion is well predicted by participation in social networks, not the various personal factors people give (from what I know, it is something of a faux paus among many religious groups to admit that their denominational identity has something to do with their parents and social milieu, as opposed to a proactive personal choice). This makes sense, who wants to admit that they joined religion X because all their friends were of religion X, as opposed to some particular mystical or philosophical insight? I think this dynamic explains the high churn rates for small sects and the nonreligious; because they are outnumbered they often interact socially with those who don't share their beliefs and are exposed to more diverse networks.  Cults get around this by imposing behaviors which limit contact with outsiders, but the nonreligious don't have this sort of organizational aspect to their identity so naturally there's going to be a lot of defection. A prediction from this hypothesis would be that defection rates for nonreligious people will be higher the higher the proportion of religious in the particular area. The rise in the proportion of nonreligious in American society wide probably will result in some reduction in the velocity of defection on the margins in the future.