On my post last week about how many kids GAs have, Petra asked about what the numbers would look like for women in general Church leadership positions. To answer this question, I’ve looked up the number of children that women in the General Relief Society Presidency (hereafter, GRSP) have had. To match the dataset I have for the FP/Q12, I included only women called since 1920.
I limited my search to the GRSP, and didn’t look at women serving in other general positions, because the GRSP alone were difficult for me to track down. I started with the handy list of presidency members on Wikipedia, and I was able to find numbers of children for all of the presidents and some of the counselors there as well. But many of the counselors don’t have Wikipedia pages, so I had to look in the Church News archive, the magazine search at lds.org, the Google News archive of the Deseret News, and the archive of the Relief Society Magazine at the Internet archive.
Anyway, here’s a graph showing the results. GRSP members are red dots (with a red line for the best linear fit); FP/Q12 members are blue dots (with a blue line for best linear fit).
Matt W. and Jared were correct, I think, in pointing out on the other post that there really is a trend for the FP/Q12 members, even if it’s not dramatic. It’s a decrease of about 0.12 children per decade. For the GRSP members, the trend is strongly positive: an increase of 0.43 children per decade. This is truly remarkable, especially in the face of declining birthrates in the developed world across the same time period.
It seems straightforward to construct a narrative that fits these data. As the Church has become more involved in culture wars in the past few decades, GAs have found it more important to define women by their role as mothers. Consequently, they have made it a point to call women who have more children to prominent positions like GRSP to serve as an example for the general membership of the Church.
I’m sure there are many possible alternative explanations, including that this is a spurious relationship. Please feel free to share your favorite alternatives in the comments.
A few caveats:
- The GRSP dataset is even smaller than the FP/Q12 dataset: only 36 women. It’s not a lot to go on.
- I excluded the two RS counselors who were unmarried (Sheri L. Dew and Barbara Thompson). I thought it was of more interest to compare people who were married and supposed to be having kids to each other, but you might argue that they should be included. With them in the GRSP dataset, the increase falls to 0.32 children per decade.
- I tried expanding the GRSP dataset back a decade to 1910 just to see whether the steep trend would hold up. It didn’t: it fell to an increase of 0.12 children per decade. This is still perhaps noticeably different from the FP/Q12 decrease of 0.12 per decade, but it’s not nearly as dramatic as it was.
- One counselor (Gertrude R. Garff) had three children, including an 18-month-old, when she was called. It’s entirely possible that she wasn’t done having kids. I could argue that that doesn’t matter, since what’s important is how many she had when the decision was made to call her. But you could come back and say that the fact that she was still in the process of having kids maybe indicated that she was going to have more, and that must have been known by the men calling her.
- Another counselor (Julia A. Child) had 5 stepchildren, which I didn’t include in her total count.
I would definitely include the ones with no children in the average.
I don’t totally understand the “average” line. It doesn’t really go in a straight line does it? Are you just taking the average of 1920 and the average in 2020 and comparing them? Or are you finding an everage every decade?
Why does the women’s average go up?
I would guess that in previous years, the mothers of many children were overworked with the demands of running a household while raising that many children so church leaders were more likely to choose someone with a little more free time. Women used to even stay home from sacrament meeting with younger children. Women’s work was extremely demanding. Pregnancy was hidden and “confinement” was not that long before. Even my mother in 1967 felt she needed to quit her job when she started to show at 6 months because it just wasn’t appropriate to be a teacher while in her condition. So a woman with many pregnancies would have had fewer church callings than women who weren’t producing as many children and were out in public more often.
However, as time went on and modern conveniences changed things, a SAHM who had more children can be viewed as very capable and competent at the top of her field, much like a “CEO” or professional at the top of his field. The church tends to call very capable people to serve the Lord in higher demanding callings. Also, pregnancy and childbirth was nothing to be ashamed of, so women continued to go out in public during pregnancy and post-partum and perform the same callings and attend church etc.
Very interesting, Ziff! Could you also look at how the rhetoric about “multiplying and replenishing the earth” compares between the men and the women? Perhaps the data are too sparse for the women, but I am curious as to how often female leaders talk about those issues.
jks, each line is drawn to fall as close as possible to all the data points. More precisely it’s drawn using least squares, which means it’s placed to make the squared (vertical) distances between the points and the line add up to a value that’s as small as possible.
Regarding including the single women, why would you include them? I think they’re a different population because if you’re single in the Church, having kids is sinful, so their count of children means something different than it means for the married GRSP members.
Thanks for proposing an alternative explanation!
Thanks, Mike! That’s a great idea to look at the “multiply and replenish” rhetoric from women and men. I might already have that in my dataset. I’ll have to look.
This is very interesting as your posts usually are, Ziff. I imagine there is some truth to your explanation for these data but jks’s explanation also rings true to me. Womens roles were more limited in practice in the past and women with a lot of children would have a lot less experience with any type of service outside the home.
Whoa, who has 13 kids?
Regarding including the single women, why would you include them?
If your hypothesis is that the GAs are calling women with more children to serve as role models, you have to account for the unmarried/no kids cases. They could have filled Sisters Dew and Thompson’s positions with married women with a lot of kids, but they didn’t. That tells us something about their motivation in making the callings, and leaving them out of the analysis doesn’t accurately reflect that.
Last Lemming & jks – I think the argument against including women without any children is a strong one. In this case, the message is: “Women, you are mothers above all else, but when you can’t be mothers” (as in the case of single women) “you can still be valuable on your own”. So, you call single women to try to support the idea that single women aren’t outcasts, but among those who ARE having children, you call women with lots of children to point to them as examples of righteous living. It’s two different groups.
In other news, are there any unmarried GA’s?
Very cool analysis. What happens if you drop the 13 child person as an outlier? Still looks like the trend might be positive. Also would be interesting to see the pre-correlation trend compared to the post-correlation trend for RS. My guess is pre-correlation trend is negative, while post-correlation is positive.
Maggie, thanks for the support! I can see Last Lemming and jks’s point that it’s saying something that single childless women are called. But I still do think the argument is stronger to leave them out because they represent a different population than the group being held up as ideal role models.
Interesting questions, BerkeleySatsuki. If Linda S. Reeves (who has 13 children) is dropped from the dataset, the increase falls to 0.32 children/decade.
If it’s split pre-correlation/correlation (which I did at 1970, but there was nobody new called between 1958 and 1974, so the same result would be obtained from splitting at any of those years), the pre-correlation slope is 0.28 children/decade increase, and the correlation slope is 0.37 children per decade increase. Not as dramatic a difference as I would have expected.
Note, though, that in addition to the slope difference, there’s also a level difference: if the lines for pre-correlation and correlation are extended to 1970 to meet, the predicted number of kids for the pre-correlation line is 3.9, and the predicted number of kids for the post-correlation line is 4.6.
(The pre-correlation vs. correlation analysis has Linda S. Reeves back in the dataset.)
I find this fascinating. To me it represents the fact that when I look at the GRSP I generally often feel I’m seeing women with whom I have little in common. Mostly (with Cheiko Okazaki as an exception) these are women how have never been employed in their adult lives, who have the support of church-active spouses (except the two exceptions you noted), and who are white and from Utah (not always, but mostly). Not inherently bad things, but not representing the diversity of the women they are leading, either. I feel like the Church is holding up an ideal that is really out of sync with the real lives of its members.
So Ziff, are you saying there is a correlation correlation?
“Regarding including the single women, why would you include them?”
I disagree that you should only include the women you “think” are being held up as role models. This is a study with facts, but you are trying to skew the facts by deciding that the single women weren’t really intended to be role models
You say that “their count of children means something different than it means for the married GRSP members” but it doesn’t. They had no children yet they were called. Them having no children is as different as someone having fewer children because she married later, or was had fertility, or her husband was impotent, or she had to have a hysterectomy at age 30. By your logic, surely there is some reason to exclude the women with only two or three children as well as the ones with zero because they simply didn’t have the real chance to have more….or maybe weight them differently based on how long they were married during typical fertile years.
You might want to take a sociological approach. Data from the American Community Surveys suggests a subset of the Utah Mormon population has started having more kids in recent years, perhaps in response to falling birthrates around the country.
As of last year, 5.3 percent of families in Utah lived in 7-or-more person houseeholds. This represents an increase over 4.5 percent in 2005, but a decline from 6 percent in 2000.
Very interesting. I’d love to see this flipped from least squares to either 3rd order polynomial or moving average, with and without the single sisters. Looks like it would show it went up til the 80s then started coming back down for both cohorts (with the current outlier with 13 kids wow!)
Comparing the trend of the greatest number of children in insightful. For the men, 14 in 1935 down to 10 in 1970 to 5 in 2010.
For the women, the reverse is true.
I need to go read the original post, but I don’t know of any male GAs that are childless, compared to a significant chunk of women that are. Is this correct? Or intentional?
Ziff,
Holy smokes, I’m amazed at the research you did for this post.
The church really should hire you to do this stuff. It’s fascinating.
Or maybe you could start a Mormon stats blog. There are probably others (ahem, Mark and Nate) who would love to join you in putting stuff like this together.
Having a GRSP member with 13 children in 2014 is just bizarre to me. I don’t know what that information means to the church, but I can’t comprehend a family that large.
As far as excluding her as an outlier, does statistical theory support leaving her in the data set or not?
Thanks, Jessawhy! Actually, I’m trying to make ZD into a Mormon stats blog, but my co-bloggers keep resisting by writing more interesting stuff. 🙂
About excluding or including the one person with 13 kids (Linda S. Reeves), unfortunately, there’s no hard and fast rule. If there were a reason to believe she’s a member of a different population than the others, then having her in the data is distorting the results. However, if she’s part of the same population but just extreme in that population, then she should stay in.
It’s pretty much the same thing jks was sniping about above. I think it’s pretty clear that the single women are being called for a different reason than their number of children. She’s taking a more literal-minded approach and saying they’re part of the same population because they were called along with the married GRSP members.
Matt, if you’re interested, I’ve emailed you the data. I don’t think there’s enough data here to support a higher-order polynomial. Even if it fit well, that’s a lot of parameters to estimate with a small amount of data. But in any case, have at it and see what you find!