Occasionally blogger abb3w -- who has done a lot of interesting work with the GSS -- has a fascinating blog post where he/she analyzes the rise of the so-called "nones" (those with no particular religious affiliation) according to birthdate. The result is a surprisingly good match with a logistic "S" curve (I'm sure I knew what that was once upon a time, but no longer -- but it sure is shaped like an S!), which if extrapolated shows the "nones" exceeding 90-95% around the year 2100.
This is a highly idealized model of course, and nobody can predict the idiosyncratic nature of future history. But even the historical data is encouraging.
Update: Since some people are obstinately not getting it, let me make some clarifications.
First, as a reasonable person might have guessed from the heavily caveated second paragraph of this post, as well as the presence of a question mark in the title, I am not claiming any sort of certainty about whether this trend is real or not. In fact, it seems just as likely to me that it's not than that it is. I just thought the data abb3w turned up, in particular the close fit with a logistic curve, was intriguing.
Second, as abb3w made clear in the original post, and as I made clear in the comments, even if this trend is real it says nothing about religious belief -- it only speaks to religious affiliation. The so-called "nones" include SBNR types and so forth, for example. I thought that this was obvious from abb3w's post, and I didn't really intend to imply anything other than, "Hey look, abb3w has a cool post," so I felt it unnecessary to state this up front. But since I'm being accused of "intellectual dishonesty" for failing to spell it out for my less astute/more paranoid readers, I have chosen to update the post to reflect this. By the way, I still find the trend encouraging, because while I think that theistic belief in general is not compatible with critical examination, I think that vague personal theistic beliefs are usually not particularly harmful. (They may even be beneficial to some people, but I think the jury is still very much out on this one, and I feel it would be elitist to claim that some people "need" their comforting delusions without possessing very solid evidence to support that position)
Third, although this seems plain as day to me, I now find it necessary to point out that this post was not intended to be polemical. If you want polemics, just browse my blog for a bit, you'll find plenty. But this post is not included there. I'm not trying to "persuade" anybody of anything. I think it ought to be clear that no atheist living in America in the year 2010 is going to find the Argument from Popularity at all persuasive! As such, I would not attempt to employ it to convince others either. It is a rather lousy argument, after all. The percentage of people who are religiously affiliated, or who believe in God, or whatever, are completely independent from the truth value of any given theistic proposition, which itself is completely independent from whether or not religion is a force for good. These are all entirely separate issues. Again, if you want to hear me make an argument in regards to the truth of religion, or in regards to the positive/negative effects of religion, one will not have to look far in this blog -- but this post is not it. This post does not even fringe on those questions.
I hope all is clear now.