Short version: For all of the current sound and fury about the secularization of Europe and the angry middle-aged men of the New Atheism, the tendency to see the supernatural as real is not going anywhere, and is probably a fairly basic artifact of the way the human mind works. Thus, when "religiosity" declines, it usually does so only to be replaced by more vague forms of mysticism.
The real question, to my mind, is of course whether our inclination towards belief in the supernatural points to the very real existence of the supernatural, or is simply a side-effect of our ability to look for patterns and intentionality in nature. (The answer to that question being where I would part company with Razib and Reihan.)
Since I'm wandering... The question of why people tend to remain open to supernaturalism, regardless of their affiliation with an organized religion, reminds me of a thread of comment I read the other day in which John Wilkins of Evolving Thoughts laid out his explanation of why one should be agnostic (rather than atheistic) on supernatural questions.
He sets forth five steps:
1. Each claim is independently assessed. You make a Thor or a YHWH claim, and I will take each one separately.Maybe I'm unduly narrow in what I tend to consider "empirical" (indeed, it seems to me essential to science to be pretty restrictive with the term) but I can certainly see why (for all of Wilkins' rigor as a thinking) this is not an appealing worldview to most people. Indeed, I can't help wondering if it's even a livable worldview for life as a whole (as opposed for one's specifically scientific activities).
2. For each claim my first question is: is there a factual claim here that is empirically decideable?
3. If so, has that claim been empirically disconfirmed? [Does that god require belief in something false?]
4. If so, be atheist about that claim.
5. If not, be agnostic about that claim.
I mean, taken strictly, if someone comes to you and says "It's wrong to squash your aged and ailing mother with a steam roller," I'm not clear that this would actually be an empirically falsifiable claim. Indeed, the whole idea of "wrong" is not an empirically falsifiable one so far as I can tell. One could reason to the conclusion, but one could not discover it empirically.
And it's because of that fairly restricted realm of empirical information that those who abandon religion, for whatever reason, generally do not become strict empiricists. Invariably they end up believing in something. And (to allow Chesterton to get the last word in -- as he is so wont to do) on many occasions they simply end up believing in everything.