There is no decline. There are only a variety of apocalyptic mis-interpretations of polls.
Many self-defined "nones" attend church every Sunday.
Surveys don't measure what we think they measure.
As one friend writes, "If we don't have studies, how can we have panicked overreactions to things?"
In most surveys, respondents have to fit into the specific categories and methodologies of the survey.
In the recent Pew study, respondents were required to report just one religion. But many people are multi-religious.
Many of us complete surveys aspirationally, describing ourselves as how we hope to be.
Many of us complete surveys strategically, describing ourselves the way we want to be perceived.
Despite the media articles that the Pew report generates, the data tells us very little beyond changes in how people are willing to present themselves to anonymous surveyors.
“The facts are that the world is probably much more religious than it was a century ago,” Stark stated. “It may in fact be more religious than it ever was.”
In point of fact, religion in America is quite stable.
The "state" of religion is always super positional and only becomes known when the observer measures the "state" and defines positions. Like the famous cat in the box. In some ways, though different, like Heisenberg's Uncertainty Principle.
In other words, much of the press on the Pew religious life survey may create heat, generate click-throughs, gather attention and benefit advertising revenue.
It just won't be true.
For more, and theological reflections on polls, see http://gregorywalter.blogspot.com/2012/04/injury-as-signs-of-times.html
Most people want to know what is going on in their societies, communities, and cultures. The worst way to do this is a poll. Polls, unless crafted very carefully and subjected to considerable interpretation, make remarkable assumptions about the way that human beings perceive themselves. First, polls (as opposed to extended interviews and observations) presume that I, when taking a poll, am transparent to myself. We hardly are aware of all of our motivations and the various ways in which our self-reflective observations are self-deceptions and half-truths. We need help to understand ourselves and our families and situations. The short account we give of ourselves in a poll does not disclose much.
Second, polls threaten religious claims directly because they skew what people think is true based upon their situation, no matter whether they are conscious of it or not. Polls are sometimes thought to generate or reveal what is generally plausible or believable in a society. They are thought to present to us what are called "plausibility structures." A person usually appeals to what is plausible by appealing to a general sense of what a society accepts as true. For instance, many Americans can imagine environmental disaster. What they cannot imagine is a world governed otherwise than by a free-market capitalist economy.