Midjourney: Descendants as the stars of the heaven and as the sand which is on the seashore, in the style of Van Gogh
A recent piece of mine about how many more children US Latter-day Saints are having was recently published by the Deseret News. The TLDR is that we are still having more children than the average American, but here I will take advantage of the added flexibility of blogging to derive some estimates about what that means for Church growth in the US.
How many more children are we having? It’s hard to know for sure given sample size issues, but a rough, reasonable estimate is about twice as many. However, this isn’t as much as one might think given that the US’ fertility rate has tanked and is now solidly below replacement-level (1.64 children per woman).
How does this translate into growth? One way of translating TFR into generation-by-generation growth is by converting it into what’s called the Net Reproductive Rate, which takes sex ratios at birth and mortality rates into account to derive an estimate for how many daughters each woman can be expected to have.
Why daughters and women? Basically, the math is simpler if you assume women reproduce asexually, and with a little intuition you can see that the NRR is equivalent to the proportion by which a population will grow from generation to generation. If the average woman has 1.1 daughters, then the next generation will be 1.1 times as big.
So, if we use a rule-of-thumb that Latter-day Saints have twice as many children, we can simply double the NRR for the US to derive an estimate for how much the Church could be expected to grow from generation to generation given no immigration (in this case, no converts and nobody leaving).
The latest US NRR is .798, which means that the US will shrink about .8 generation-by-generation (once population momentum has worked its way through the system–long story).
If we double that for members that’s about 1.6. That means that the Church will grow by a factor of 1.6 every generation if our current reproductive patterns were frozen in place.
In other words, a little over 1/3 of our children can leave the Church without it shrinking assuming no conversions.
If we add conversion to the mix, we can say that we can have a net outflow (people converting to the Church minus people joining) of 1/3 before the Church starts to shrink.
Of course, I have no idea what our precise, cohort-by-cohort conversion and leaving rates actually are. It can get complicated, since when they leave or join is important for growth.
I am the descendant of somebody who left the Church but whose family stayed in. That is going to have less of an impact on Church growth than somebody who left before they raised a family. Conversely, converts who join after they have influenced their children are going to have less of an impact on Church growth than those who join before they have their children.
So to summarize, the leaving and joining rates are the big unknown, but we can leverage extant data on Latter-day Saint childbearing to get a sense of what the “natural” growth rate of the Church is, and ultimately how much leaving and/or conversions we’d need to see for it to grow or shrink.
How does this factor in inactivity? My wards inactivity rate is about 70%. So that means 70% of the current number being counted as a member of the church in total numbers will be unlikely to pass that membership down.
Or does your 1/3 number of ‘leaves the church’ factor in these?
This is just based on self-identification, so presumably including the ones that at least still identify as LDS on a form.
I saw your article in the DN on this the other day and wanted to respond, because there are a few misconceptions in the article and there are a few things I think you’ll find interesting. I appreciate you covering this topic, as I think it is something that is going to have a big impact on the future.
The first potential issue is that you mention that Utah women are not having enough children to meet the replacement rate of 2.1. This is technically true, but suffers from issues relating to how the Total Fertility Rate (TFR) is calculated. Because we consider all women of childbearing age (15-44) when calculating the TFR, it can affected by changes in the age at which women have children. When the mean age at childbirth rises rapidly, TFR can be artificially low. This happens because fertility rates fall for younger women due to delayed age at birth, but this delay is not made up for at older ages because the women at those ages already had their children earlier. This is called the “tempo” effect, and results in a period TFR that is lower than number of children the women giving birth at that time actually end up having. The delay in childbirth is happening so rapidly in Utah right now that I think it is likely that the cohorts of women giving birth in Utah currently are likely to actually end up having enough children to meet the replacement rate of 2.1 even though the current period rate is 1.92.
In the article you mention that Utah fertility is declining because the LDS percentage of the population is declining. While this may be having some impact, if you actually break down the fertility numbers by geography within Utah, some of the biggest declines are happening in areas where the percentage of members likely hasn’t shifted much, like the area immediately adjacent to BYU. Furthermore, Utah has experienced absolute declines in births, not merely a decline in the relative rate of births. This means that the rate is not simply declining due to new people moving into the state who are not LDS. What’s more, the parts of the state with the highest percentage of Latter-day Saints have seen the smallest declines. For example, Salt Lake County, where church members are now nominally below 46% of the population, has seen a 26% decline in TFR, while Utah County, where members are nominally 80% of the population, has only seen an 18% decline. Interestingly enough, even though the Utah’s TFR has fallen slightly faster than the nation as a whole, thus likely meaning that the fertility advantage that Latter-day Saints have compared to the nation is shrinking, the fact that Salt Lake is seeing bigger declines that Utah County means that it is likely that the Latter-day Saint fertility advantage is actually growing within Utah. This is because it is not the absolute TFR that matters—it is the percentage difference between the two, which is growing.
Another topic you cover is declining family size. This is actually a popular misconception. Family size did decline in the US as a whole after the end of the baby boom in the 1960s, but it actually risen slightly since then. Changes in TFR in the last few decades are largely about changes in the level of childlessness, not the family size of those who have kids.
Utah family size (and thus, probably Latter-day Saint family size) does follow a slightly different pattern, however. Utah fertility declined in the 1960s but then had another small baby boom in the late 1970s/early 1980s. As a result, Utah families did not really get smaller until the 1980s. Utah family size has been basically flat since then, and is not declining, as shown below.
Instead of pulling a small subsample from surveys like the Cooperative Election Study (which I don’t really trust, for other reasons), I think a more reliable way to determine family size for Latter-day Saints is to look at Census data for areas where church members make up a very high percentage of the population, such as Utah County. Of course, even though the percentage of members is high, it’s only 80%, so these numbers are probably a little too low. Also, there’s a very real possibility that family size is different in Utah vs outside of Utah, but this is a good way to at least get a baseline and look at change over time. The census used to ask these questions directly. Since 2001, the Census Bureau’s American Community Survey (ACS) available from IPUMS has a large enough sample to look at this fairly reliably even at the county level. The specific question about children asks about children currently living at home, not children ever born, so it is a slight underestimate. However, the wording is consistent across time, so it can be used to look at change.
Since we are interested in the size of families, I restricted my analysis to married mothers (instead of all women). This prevents changes in the percentage of women that have no children from affecting the results. I exclude women who are not married because the question is specifically about children living at home, and custody arrangements may artificially lower that number for women who are not married. Since the average number of children peaks between the ages of 37 and 41, I have used that age range. So, looking at married mothers ages 37-41 in Utah County over time, we see the following for average number of children living at home at the end of mothers’ childbearing years:
1980 census: 4.05
1990 census: 3.93
2000 census: 3.65
2005-2009 ACS: 3.36
2008-2012 ACS: 3.38
2012-2016 ACS: 3.42
2017-2021 ACS: 3.41
As a point of comparison, here is the US as a whole:
1980 census: 2.54
1990 census: 2.19
2000 census: 2.23
2005-2009 ACS: 2.21
2008-2012 ACS: 2.24
2012-2016 ACS: 2.26
2017-2021 ACS: 2.28
Now, keep in mind, this is measured at the age when women are essentially done having kids, so it basically corresponds to fertility rates from 15ish years prior. So, the women measured in the year 2000, for example, would have had the bulk of their kids in the 1980s. What you can see in this data is a big drop in family size in the Utah County sample between Baby Boomers and Gen X. However, the women in the most recent years are the oldest of the Millennials, and they had slightly larger families than Gen X.
Since this is only looking at women who are basically done having children, it doesn’t tell you anything about the younger Millennials or Gen Z. However, other analyses I’ve done breaking it down by younger ages shows that there has been essentially no change over the last 15 years. Younger Millennials and Gen Z parents have roughly the same average number of kids at each age as did Gen X.
However, since 2007 there has been a huge increase in the percentage of women at younger ages that have no children, likely as a result of declines in marriage. I have collected the number of births per year by single age for each birth parity for the last few decades and used those to calculate cumulative first birth rates by birth year cohort for Utah County. For those born in 1980, 50% had their first child by the time they were 24. For those born in 1998, it was only 26% by age 24. Now, some of this is simply delaying. Many of these 24 year olds without children will of course have children, just at an older age than those in previous years. But if you look at those who are approaching the age where first birth rates really decline, it is clear that not all of these women delaying childbirth will actually make up for the delay at older ages. For example, the 1990 birth cohort has passed age 32 with only 72% having had a child, compared to 84% at that age for the 1980 cohort. The first birth rates that the immediate few birth cohorts preceding 1990 experienced show that the rates for the rest of the early 30s will not be nearly sufficient to reach much more than that 72%. There will simply be more women permanently childless. And that’s just the women born in 1990. Women born in 1998 have even higher childlessness rates.
All of this was pretty rambling and long, but my point is basically that LDS families have probably been about the same size for about 30ish years now and are currently not declining in size, despite lower fertility rates. Lower fertility is being driven by increased childlessness (likely resulting from declining and delayed marriage), rather than smaller families. I have a whole lot more data on this than what I have shown here but this is already way too long of a comment, so I’ll leave it at that.
Thank you for all that analysis! Very interesting. I was particularly interested in your finding that family sizes, once you remove the childless, have remained more stable than I would have thought. A few points:
1) I think the childless are part of the story, so when we talk about “average family size,” so in general I’m fine including people with zero children in that calculation, but your analysis is quite useful for decomposing how much of that decline comes from what.
2) I do think there is much to be gained from looking at high-LDS Utah areas as a sort of proxy for LDS fertility, and the much larger-N census data is much more comfortable to work with, but at the end of the day Utah members aren’t synonymous with members in the US in general, and Utahns are not synonymous with members. There’s enough of a non-overlap that it was nice to be able to just use data that directly asked about religious affiliation.
3) You’re right that I should have mentioned the possibility of tempo effects artificially downweighting TFR (as we saw during the “lowest low” fertility in Western Europe before it rebounded), but whether it’s a tempo effect-induced decline or an actual decline is always a bit of a mystery until women in the respective cohort age of of reproduction and we know for sure.
Great stuff William, and thank you for sharing it. I do have to concur with Stephen C’s concern about using high-membership counties in Utah to make inferences about church membership as a whole. We talk a lot about becoming a global Church, but it seems to me that even in the US we’ve become a much more national Church. A smaller proportion of members have any ties to Utah, and for many of those who do the ties are less direct (e.g. a generation or two in the past). That leaves plenty of room for Utah Saints to differ from Saints in the rest of the country on cultural matters. On the other hand, given the lack of good data on Church members nation-wide, it is another useful way to see through a glass, darkly.
(As for Utah County, one of my ancestors temporarily relocated his family from the Salt Lake Valley to the Utah Valley as Johnson’s army approached, and described the inhabitants in his journal as “strange.” This descendent enjoyed his youthful sojourn in Utah County, but can’t disagree with the description!)
Very true about potential differences in Utah vs outside of Utah. The 2008 wave of the American Religious Identification Survey found that LDS household in Utah were a little larger than those outside of Utah. I do wonder, though, how much of that is related to the much higher percentage of converts in the rest of the country, who are more likely to be single. I will say that the Mountain West had much larger declines in fertility rates over the last decade than the rest of the country, so it seems plausible to me that household size gap may have shrunk a little.
In terms of future growth, especially in the US, something that I don’t really ever see taken into account is the big drop in births in the 80s, which can be seen both for Utah and for the church overall. The church’s data on this is a little tricky to work with because of an apparent methodological change in the late 80s. Utah had almost twice as many births in 1981 as it did a decade before, but then it declined sharply and the 1981 total would not be surpassed again until 1997, when births again shot up as those big late 70s/early 80s cohorts began having kids. Keeping in mind the problems with using Utah as a proxy for church members (although the available church data does mostly mirror this), I think this goes a long way towards explaining why the church’s growth rate slowed over the last 20 years, at least in the US (of course, what happens in Utah has much less of an effect on the church internationally than it does on the US). The women giving birth in the 2010s were from birth cohorts that were either smaller in size or only slightly larger than the ones a decade before, after having previously experiencing cohorts getting much larger over time. I think we’ll start to see growth tick up slightly again as the larger late 90s/early 00s cohorts hit peak childbearing ages, despite lower fertility rates.
I do think a lot of the sharp decline in Utah in particular is a “regression to the mean,” “larger you are, the harder you fall,” etc. With the raise of the Internet the normative family size has probably become more homogenous., making to harder to maintain Mormon-specific fertility norms.