The Pew Center on Religion & Public Life released the results of a survey yesterday that made headlines by revealing that more than a quarter of Americans have left the faith in which they were raised, and that figure is 44% if one counts movement from one Protestant sect to another. Perhaps the bigger news in the survey is that all Protestant sects combined make up only 51.3% of the population.
Or do they? "Unaffiliated" Americans account for 16 percent of the population, breaking down as follows:
Secular Unaffiliated: 6.3
Religious Unaffiliated: 5.8
It's probably a fair bet that slightly over half the "religious unaffiliated" Americans are unaffiliated Protestants, which brings the combined number of American Protestants to more like 54%.
Maybe the biggest news is that over 10% of Americans are not religious at all. That's roughly equal to the number of Americans who are ex-Catholics. Indeed, there is likely considerable overlap in those groups.
What the survey may miss is the difference between religious and cultural identity. For religious groups with strong ethnic components (e.g., Judaism, Greek Orthodox, Native American religions), one can well imagine a survey respondent self-identifying in one of these groups without affirming any of the religious tenets thereof. A secular Catholic or Jew, for example, could self-identify as Catholic or Jewish, even though she should also count as an agnostic (say). But because the percentages sum to 100%, the survey wouldn't register this respondent as an agnostic. That, in turn, may suggest that the aggregate proportion of non-believers and doubters is larger still.
Finally, note that the survey results by age show that older people are much less likely to be atheists or agnostics than younger people. It would be interesting to learn whether the current younger generations are simply less religious than earlier ones, or whether this fact reflects a change that occurs in lifetimes: If the latter, that would mean that as people get older, they worry more about their mortality, and are more likely to place Pascal's wager.
Posted by Mike Dorf