The following is Stephen Cranny’s third guest post here at Times & Seasons. Stephen Cranney is a Washington DC-based data scientist and Non-Resident Fellow at Baylor’s Institute for the Studies of Religion. He has produced over 20 peer-reviewed articles and five children.
I was surprised at the Church’s seeming statement (I’ll discuss why it was “seeming” later), echoed around the standard General Conference post-game analysis that half the Church in the United States was single. The surprise came from the fact that self-response surveys on religion and marital status consistently show that a solid majority of adult Latter-day Saints are married. For example, the largest recent self-report survey (Pew 2014) reports that 66% of self-identified adult Latter-day Sainta are married. I checked the General Social Survey, an omnibus survey taken every year with a wide variety of variables, and found similar results. Now, the GSS only has a couple dozen Latter-day Saints every year so the results are much more sporadic (as can be seen below). However, by combining years you basically get the same picture painted in the 2014 Pew report, with Latter-day Saint marrieds in the low-to-mid 60s range as of the mid-2010s (2018 is also available in a different dataset, but I dashed this off in between errands; presumably that one year won’t radically affect the take-away).
So why the discrepancy? I can think of a few reasons, some more convincing than others.
1. Surveys use self-reported members, while the Church may be drawing from Church records.
Researchers have found significant discrepancies between the numbers reported by the Church and numbers reported by the census in countries that have a religion question on their census. That’s not to demean Church record-keeping; as anybody who has systematically visited every house in a ward list can tell you, a significant portion of people on church records are people who were baptized when they were younger but who have, for all intents and purposes, lost contact with the Church and would not identify as Latter-day Saints on a survey. Presumably it’s difficult for the Church to know how to handle this segment of the Church population because, as any ward clerk who has transferred records from decades ago from another country can tell you, sometimes they do resurface after years and start attending their local ward.
However, the Church does often use a more refined denominator for in-house purposes. For example, in a private setting (it was a very random occurrence and my professional background had absolutely nothing to do with why I happened to be there) I heard a General Authority once note how much the temple-recommend holding membership had grown in the past decade (you would think I would have remembered the details, but I don’t; I vaguely remember being moderately optimistic). However, whatever the condition of activity required to make it into the denominator, it would probably be stricter than the limit set by surveys of simply identifying as a member. Since more religiosity almost certainly means a higher likelihood of being married, the Pew and GSS results should be taken as a very conservative estimate, since the bar for being included as a Latter-day Saint is so low. Consequently, assuming that the Church’s analysis is more sophisticated than just using the raw Church records, I doubt this explains the discrepancy.
2. The latest survey results are several years behind the Church’s in-house numbers. The discrepancy reflects changes since then.
The Pew survey was conducted in 2014, while the GSS trendline here goes up to 2017. Elder Gong noted that the crossover happened very recently. It is technically possible that the gap is due to rather radical changes between the surveys and latest round of Church data collection. However, I am highly dubious about this option, since these kinds of demographic indicators change very slowly for a population. For example, between 2007 and 2014 the percent of self-identified Latter-day Saints that were married in the Pew survey dropped five points. Assuming that trend continued for the next seven years until now, about 59% of self-reported Latter-day Saints (which again, is a very conservative estimate as it includes at least some disappeared ones who don’t make institutional contact with the Church in a way that would register with their data collection) would still be married.
3. The discrepancy reflects random error from the survey.
Any survey that does not include the entire population has a margin of error, and technically it is possible that Pew and GSS were both shifted up simply due to the fact that they were using a sample (Pew’s sample size of Mormons was 664). However, even with this sample size going from 66% to 49% would be a highly unlikely. However, random error could possibly combine with some version of the issues above in a way that would explain the discrepancy.
4. The Church did not actually say that a majority of members were single.
The statements from General Conference regarding percentage single:
President Ballard:
We long to help all who feel this way. Let me mention, in particular, those who are currently single. Brothers and Sisters, more than half of adults in the Church today are widowed, divorced, or not yet married.
Elder Gong:
Also, the majority of adult church members are now unmarried, widowed, or divorced. This is a significant change. It includes more than half our Relief Society sisters, and more than half our adult priesthood brothers.
A significant proportion of divorced or widowed people are currently married, although in surveys the options are usually presented as mutually exclusive categories. It is likely that they were referring to people who don’t meet the Latter-day Saint cultural archetype of consistently married. One could argue that President Ballard implied that the “widowed or divorced” group only included the not-currently-married, but he wasn’t clear here, especially since he used the term “unmarried,” and not “never married”. Ultimately, I believe that the remarried category is the most likely explanation for the discrepancy given the issues with the other explanations noted above, and that therefore the statement that “most Latter-day Saints in the US are single” is false, although I reserve the right to be wrong.
On a related note, in discussions I’ve had with others after conference it is clear that the image some people have in their head as a result of this conference is that this is a Church of 20-40-year old never-marrieds. However, the Pew results show that only 19% of self-identified Latte-day Saints fall into this category. (Of course, the fact that in the minds of many “single” automatically conjures up images of late-20-year old never-marrieds in a student ward, and not a single mother with kids, is another issue, one that I believe the General Conference speakers addressed). Whatever the case with the discrepancy, it is clear the never-married category is still a minority in the Church (a significant minority, but a minority nonetheless).
Conclusion
Are half of all self-identified Latter-day Saints single? I doubt it. Does this change anything about the messages given in response to these statistics? No. Whatever the specifics, the case is that a significant portion of the Church is single, and no point that was made in conference hinges on whether we’ve actually crossed the “majority single” threshold yet.
Does it potentially change people’s assumptions about the US Church’s makeup based on this conference? Potentially. This post was mostly an attempt to respond to certain assumptions I’ve seen people making as a response to conference; it is not an attempt to criticize the Church’s in-house social scientists (some of whom I’ve had the pleasure to know personally) or methodology, nor is it an attempt to needle the powers-that-be to get more specifics. The Church will probably remain silent on technical details leading to this number, and if option # 4 is true, I’m sure there are people at South Temple Street amused at the incorrect assumptions that have led to the “majority single” soundbite.