In Clark G. Gilbert’s April Conference talk, he referred to Ryan Burge’s Substack Graphs About Religion. Gilbert (see his note 11) was talking about people still being concerned with meaning and purpose in life even after leaving religion. Now, though, Burge has a post up that I’m quite sure will not be referred to in Conference, titled “The LDS Church Has a Retention Problem.” (Although it has already gotten a defensive response from a BYU professor in the Deseret News.)
Burge uses Pew Religious Landscape Study data to compare retention among Mormons and other religious groups and finds Mormons score poorly. He has a lot of interesting things to say in his post, and I definitely recommend you go read the whole thing if it sounds at all interesting. I wanted to focus on just one point he made, though, which is that Mormon retention is lower among younger people. Here’s one of his charts:
As you can see, retention is lower among younger people, as Burge discusses. But what I find striking is that this comparison actually understates the retention problem. The oldest respondents have had far more years of adulthood, far more opportunities to leave, but they still have higher retention rates than the younger respondents do after far fewer years.
Here’s a reworking of just the Mormon data from the chart above. I took the midpoint of each decade cohort (i.e., 1945 for 1940s or earlier, 1955 for 1950s, etc.) and calculated age in 2023, which is the starting year of the Religious Landscape Study.
If there were a similar exit rate for each of the cohorts, the pattern in the scatterplot would go down to the right, as the older cohorts that have had more time to leave would have had more people leave. (I admit here that I’m assuming a situation where people who leave hardly ever come back, although I admit it does occasionally happen, in stories often trumpeted in Conference.) But instead, the pattern goes up to the right, showing younger people having left at a far faster rate than older people.
Here’s one other way to look at the question, and a crude way to quantify it. In the table below, for each age cohort, I figured out what constant yearly loss rate would lead to the observed total loss rate for each age cohort. For example, for the oldest group, the midpoint age in 2023 is 78 (2023 – 1945). With a 68% retention rate, this means 32% have left. For simplicity, I assumed leaving starts at age 18. Then I found the yearly loss rate that would be required to yield a 32% loss rate after 60 years (78 – 18). This turns out to be 0.64%. This just means that (1 – 0.64%)60 = 68%, or in other words, if you reduce the starting number by 0.64% each year for 60 years, you end up with 68% of your starting number remaining. Here’s a table with the same calculation for each age cohort.
The yearly loss goes from well under 1% for the 1940s and 1950s cohorts to between 1% and 2% for the 1960s and 1970s cohorts to nearly doubling from the 1970s to the 1980s cohorts to more than doubling from the 1980s to 1990s cohorts. There are a lot of caveats: if you look at Burge’s original chart, the LDS error bars are wide because the LDS samples are small. I’m only showing loss per year estimates to a tenth of a percent in the table above because it’s needed to distinguish the values for the oldest cohorts. Of course these values have lots of error in them and are nothing like as precise as I’m showing. And like as always with Pew data, it’s the US only. But even given all this, I find the conclusion striking: the Church appears to be doing a really poor job of retaining American members born in the 1980s and later.