I kind of like the speculative theology that people who do not have the ability to raise children in this life, whether through death, accident, or lack of opportunity, will be able to in the next. 

The TLDR: Latter-day Saint women are less likely to be childless than non-Latter-day Saint women. Specifically, we’ve gone from a little under 70% of LDS women 18-45 having young children to a little under 60%, while for non-Latter-day Saints the decline went from about 55% to 45%.

The gap between us and non-Latter-day Saints has basically stayed the same over the past 17 years. Latter-day Saints are more likely to be childless than they were in the past, but so too are non-Latter-day Saints.

One of the most notable empirical findings in the religion and fertility literature (back when there was sociology literature about something else besides politics and “gender, race, and inequality”) was that the storied Catholic fertility advantage collapsed in the mid-twentieth century.  I’ve long wondered whether we’re going through something similar with Latter-day Saints. As I’ve noted before, we still do enjoy a fertility advantage, but it’s hard to know how much it’s grown or declined relative to everybody else simply because it’s hard to have a long enough time series of data that has enough Latter-day Saints to really draw a lot of conclusions from (the GSS has a longer time series, but it only has a few dozen Latter-day Saints or so in every wave, so once you start splicing different ages out it isn’t very helpful for this kind of analysis).

The Cooperative Election Survey does have a lot of Latter-day Saints across a longer timeframe, but it only has whether they were a parent or guardian of any child under 18. This isn’t ideal obviously, but as a rough proxy it’s the best we have. So I subset the data to just include women under the age of 45 and took the average number who answered that they were the parent or guardian of any child under 18. Again, it’s not ideal, there are a lot who just haven’t had their children yet, we throw out all the variation of how many children, and “parent or guardian” isn’t the same as children ever born. (And yes, I’m only using women, but that’s a standard operating procedure in demography for a variety of reasons). Still, as a proxy for Latter-day Saint pronatalism it works and, again, is the best that we have data for.

So what do the results show? Unsurprisingly, Latter-day Saint women under the age of 46 are indeed much more likely to be parents of young children than their non-Latter-day Saint counterparts. Furthermore, when we fit a simple trend line it doesn’t look like the gap is closing either. Both us and non-Latter-day Saints are becoming more childless in tandem (although the increase in childlessness isn’t as much as I would have thought off the top of my head). In 2023 there was one fluke year where we had higher childlessness than everybody else, but in the CES every other year has about half the sample size of the other years, so they’re worth less in terms of predictive accuracy. The 2023 numbers are based on only 69 LDS under-46 women, whereas when we have the larger sample in 2024 of 142 women the gap becomes big again, so I feel okay chalking that up to a statistical fluke.

---

 

Code:

library(dplyr)
library(tidyr)

library(pollster)

df <- readRDS(“LOCATION/dataverse_files/cumulative_2006-2024.rds”)
attributes(df$has_child)
table(df$has_child)
#Under 45 year old
df$has_child_num<-ifelse(df$has_child==”Yes”, 1, 0)
table(df$has_child, df$has_child_num)
attributes(df$religion)
df$LDS<-ifelse(df$religion==3, 1, 0)
table(df$LDS, df$religion)
table(df$year)
attributes(df$age)
table(df$gender, df$year)
table(df$LDS, df$year)

df<-subset(df, age<46)
table(df$age)
df<-subset(df, gender==2)
attributes(df$gender)

df_summary_wide <- df %>%
group_by(LDS, year) %>%
summarise(
avg_has_child = weighted.mean(has_child_num, weight, na.rm = TRUE),
.groups = “drop”
) %>%
pivot_wider(
names_from = year,
values_from = avg_has_child
)

X<-df_summary_wide

table(df$year, df$LDS)

#Check
afd<-subset(df, year==2015 & LDS==1)
wtd.mean(afd$has_child_num, afd$weight)