Stats and Titles: A Cautionary Tale

2014-08-18 08:47:47

Excerpt: The actual conclusion to which these statistics lead us is that people today, at least those who think beyond the daily grind of life, are much less likely to march in lockstep with a large group affiliation, whether political or religious.

There is a report of the change in religious affiliation with the following title: As Protestants decline, those with no religion gain (source). If you look at the chart there, it shows Protestant as having 48% affiliation, while the next two affiliations are Catholic at 22% and “nones” at 20%. In other words, Protestants have more than all other groups added together, excluding the Nones. Yet the subtitle would lead the reader to believe that the Nones now have the largest percentage: “Protestant” is no longer America’s top religious umbrella brand. It’s been rained out by the soaring number of ‘Nones’– people who claim no faith affiliation.

Clearly the subtitle is very misleading and gives a different impression than the title. But the title is also misleading, since nothing is said about the decline of any of the other affiliations such as Catholic and Mormon. Have they declined as well? If so, then the Protestant decline is not as significant as the decline in all religious affiliation… which is what the rise in Nones means. So why the juxtaposition with only Protestantism? It’s possible that only that affiliation has declined, but no statistics are given to support that. A much better report would be to list the change in percentage of all the groups.

But perhaps more importantly, we should note that while the survey was done by a reputable polling organization, the sample was much too small to be meaningful. The caption on the chart indicates, “Number of respondents: 9,443 in 2007 and 17,010 in 2012.”1 Currently there are about 300 million people in the US, so this represents a mere .004% of the population. Yet even if we reduce the total population to include only adults, assuming an average of 1 child per adult, the percentage only rises to .017. This is not adequate or a fair representation.

But even if one accepts the sample size, the fact remains that this is not at all about the decline of Protestants but of all religion. In the body of the article we find this admission: “Where did they go? Nowhere, actually. They didn’t switch to a new religious brand, they just let go of any faith affiliation or label.” So a much better title and subtitle would be to emphasize the decline in all religious affiliation, and to give statistics for all of them combined as compared to the Nones.

The link for further reading does little to clarify, only adding political affiliation trends among the Nones. But it does manage to throw away the reason for the article (“The big news about… the Nones, isn’t that they’re No. 2 now in the USA…”)2 and claim that what really matters is the “diversity” among them. So the “real news" here is not about Protestantism or even the decline of all religious affiliation, but the “diversity” among those who have no religious affiliation.

The few anecdotal quotes are from people who seem to have turned to atheism or simply don’t know if anything is true, rather than those who don’t vote according to their religious affiliation or lack thereof. That is, though the goal of the two articles seems to have been to raise alarm over the declining ability to control large voting blocs, nobody was quoted about that. And none of this addresses whether people who still identify with major religions vote as a bloc.

The title and subtitle gave the impression (to the reader who merely scans titles) that this was all about Protestants losing faith, the article itself showed (to fewer people) results from a small sample that showed this wasn’t really only between Protestants and Nones, and the second article (that much fewer people were likely to read) worried about voting blocs and didn’t even ask whether this was only an issue among the Nones. The real news here is that people today, at lest those who think beyond the daily grind of life, are much less likely to march in lockstep with a large group affiliation, whether politically or religiously.


  1. The second article goes on to say that the sample size was “46 million people… in a new analysis by the Pew Forum on Religion & Public Life, out Tuesday.” Does the new sample show the same percentages? We aren’t told.
  2. It also contradicts the chart, saying that the Nones are “No. 2 now in the USA” when they’re clearly No. 3.