If you don’t mine choking down a heaping helping of bloombast, there’s an intriguing suggestion at the heart of this Policy Review essay by Mary Ebserstadt: The well-known correlation between religiosity and family size should be interpreted not (just) as evidence that religious belief or participation induces people to have more children, but rather as revealing that large families tend to make people more religious. She infers from this that the secularization thesis, which many take to be falsified by recent history, faces an additional hurdle, insofar as any number of policy changes that might increase fertility rates would also tend to promote religious resurgence. (Though see this post by Brink Lindsey, which summarizes some reasons to think secularization is indeed proceeding, however slowly.)

A few thoughts. Just as a general point, I find it highly unlikely that demographers and other social scientists generally talk about the effect of religion on family size because they’ve just never given any thought to the direction of causation. There are statistical tools for sorting this stuff out, most of them rather more sophisticated than Eberstadt’s impressionistic eyeball survey. The footnotes here are stunningly thin on serious academic investigations of the interaction and fertility—and they are out there. I’m going to go out on a limb and give them credit enough to assume there has, in fact, been some effort to check this. But since I’m also not intimately acquainted with the literature, I’ll do some impressionistic eyeballing of my own.

Eberstadt frames her analysis as deciding between (or rather, deciding the relative weights of) an A-causes-B account and a B-causes-A account, when a C-causes-A-and-B account seems as plausible as either. That is, there are probably a whole array of other overlapping dispositional and cultural factors that happen to promote both secularism and lower fertility. There is, for instance, some reason to think religiosity varies inversely with education, and that people who seek more education are more prone to delay childbearing, to be more career-focused, and so on. If certain people have (or certain cultures promote) stronger underlying desires for a sense of “rootedness,” as opposed to cosmopolitan autonomy or whatever we want to call it, that too might cash out as a dual impulse to earlier family formation and greater religious participation.

Still, I doubt Eberstadt is wholly off base: Presumably we all know people who were more or less secular but joined a church when they had children. It’s not unheard of that you find even agnostics or avowed unbelievers with the view that it’s good for the kids to be raised with a church, and instrumental participation can segue into genuine commitment. Yet note that the mechanism I’m suggesting here is different from Eberstadt’s idea of childbirth-as-transcendent-experience in a way that matters for projecting future trends. If having kids just directly promotes some kind of visceral yen for God, then we can be fairly confident that an uptick in fertility would tend to promote religion. If, on the other hand, the link turns on this more instrumental sort of consideration, we should expect more ambiguous results, depending on how strongly people continue to believe that kids need religion, whether there are alternative community groups that might serve as credible substitutes, and so on.

Perhaps least persuasive, though, is Eberstadt’s implication that lower fertility (and with it, secularization) can be dramatically reversed if only we reform Social Security or eliminate no-fault divorce or reject gay marriage or institute tuition tax credits or any number of other policy tweaks. Sure, there are variations in fertility from time to time and place to place, there’s the occasional baby boom, and so on. But the clear macro-trend in family size across the planet is not being driven by looser norms about divorce or retirement policy—if anything, these are all epiphenomena of the trends in economic development that are both generally desirable and unlikely to reverse course anytime soon. I’m not seeing a comeback in the cards for the model on which you have eight kids because you need more young farmhands, and you’d better have a few extra because there’s a fair chance a couple will die of whooping cough. Economies that demand ever more education for ever more jobs encourage people to delay reproduction while they finish their schooling—and then to limit what childbearing they do with an eye to the schooling they’ll be paying for down the road. Birth control is only going to get better and more convenient. So Eberstadt should probably be happy that the rest of her case isn’t especially persuasive.