Baylor University has recently released a study, titled “American Piety in the 21st Century”, on American religious affiliation and church attendance. This study has been widely described as casting doubt on previous studies which have found that religious belief is declining in America. In particular, the Baylor researchers have described their study as an explicit rebuttal to the much-cited 2001 CUNY ARIS study which found that as many as 15% of Americans are non-religious. In contrast to this finding, Baylor University professor Kevin Dougherty stated that 89% of Baylor’s survey respondents had an identifiable religious affiliation, and asserted that previous surveys have overestimated the numbers of the non-religious by incorrectly giving that label to people who actually attend some identifiable, non-denominational church.
“We find that barely one in 10 truly have no religious affiliation in America,” said Kevin Dougherty, Baylor University assistant professor of sociology. “Prior national surveys have concluded that 10 million people are not religious who actually are in church every Sunday – praying, believing in a God; 10 million Americans counted as religious nones.”
There are several things that can be said about this. First, even if we take the Baylor researchers’ findings at face value, their study has not shown a dramatic decline in the numbers of the godless and may not indicate any decline at all. The 2001 CUNY survey found that 14% of the population is non-religious; by contrast, the Baylor study claims to have found that the non-religious constitute 10.8%. The Baylor findings claim that “This three to four percent difference is significant” and that “researchers have previously over-counted the religiously unaffiliated by 10 million Americans” (p.10). And yet, their own data from elsewhere in this very same document state that the survey has a margin of error of plus or minus four percentage points (p.53). In other words, the finding that the Baylor researchers claim to be “significant” is actually within their own self-stated margin of error! It is scientifically inaccurate and academically dishonest to assert that a difference within the margin of experimental error represents a significant change. In reality, when error bars are taken into account, the Baylor data do not contradict the ARIS data at all.
Considerations of experimental error become even more important when one considers the relative amounts of data upon which the differing conclusions are based. The Baylor study had 1,721 respondents. By contrast, the 2001 ARIS study had over fifty thousand respondents (source). It is a basic statistical principle that, all things being equal, a survey based on more data is more reliable than a survey based on less. Unless the Baylor researchers can show some systematic error in the ARIS data, it is reasonable to assume that that data set is more accurate and more representative, and that the claimed difference in the Baylor study simply represents a statistical fluctuation caused by their low response rate.
For interested readers who wish to delve more deeply, Beware of the Dogma has a superb analysis of the Baylor study, including incisive criticisms of some very strange apparent gaps in its methodology. One such criticism that I found very important was that the Baylor study classified people as religiously affiliated if they described themselves as “seekers”, a highly vague and ambiguous term that could accurately describe many people who have no real religious belief.
The SBC’s president, Frank Page, says in the article that “his prayer is that more churches will begin offering Christian schools, both for families who can and for those who cannot afford such education”. Aside from the bizarre wording – does he really mean he believes that even families who cannot afford it should send their children to private Christian school? – there is a very revealing implication here. In essence, Page is admitting that young people who are exposed to perspectives other than his own tend to leave the church – that Christian belief cannot survive without intense indoctrination and careful sheltering from external sources of information that might encourage people to consider the issue from other sides. This, in turn, suggests a strategy for atheist evangelists: if we can simply expose believers to information about ourselves, our work may largely be done for us! It is very encouraging to think that religious indoctrination, as intense as it is, could be so fragile. There may be a world of potential atheists awaiting us, if we can reach out to them.
There is one final piece of good news for your consideration: an article from this summer, Catholic priests become sought-after Polish export. This piece informs us that religious belief in Europe is declining so steeply that the Catholic church can no longer find enough people who want to be clergy to fill the posts in most countries, and the Vatican is relying heavily on Poland, one of the few countries that is still a net exporter of priests. Even former Catholic strongholds such as Italy and Spain, the article points out, have only a 10-20% rate of church attendance. In this respect, it seems that Europe is leading the way; now only if we can make headway in Poland, the Vatican’s last European stronghold, the results might truly be worth seeing.