For the next couple weeks we are going to dig into Jon Hein's WELS Statistical Report for 2023. This week we'll discuss some observations on trends and I will provide my 20-year synodical projection based on those trends. I'll present this with minimal commentary, and in the upcoming weeks we'll discuss several specific interest areas in the report including fertility, school enrollments and of course ECE's.
First, let's consider communicant membership The report provides data for the years 2000-2023 and does not attempt to project communicant membership into the future. I only make two assumptions in the projection of the data. The first key underlying assumption I am going to make is that membership is Lindy. If it is Lindy, then the local slope of our membership decline will persist in proportion to the slopes existing persistence. The longer-term trends that predate the near-term trends will predominate the future. It's not perfect, it's only a model, but one could argue it's optimistic in the sense that things only get 'better' on the basis of our current rate of membership decline.
In this plot and all subsequent plots: blue is data extracted from the statistical report, orange and green are my calculations or projections. My model predicts us leveling out in 2045 around 210,000 communicant members.
The changes in membership year over year are due to three main causes: child confirmations, adult conversion and death. Figure F and G (page 7) of the statistical report has historical data for all 3 causes but it's difficult to project adult conversion, so I didn't try.
Figure J displays WELS fertility rate in terms of births-per-fertile-woman. Recasting into a proper fertility rate and using the provided number of fertile women we can back-caclculate births per year. Figure X provides births-per-year from 1989-2009, and the fertility information from 2000-2023 overlaps nicely for that ten year span. What wasn't shown directly in Hein's report is the precipitous dropoff from 2010 to 2019 - a 30% decline in children over ten years. Then we see in a single year from 2019 to 2020 a decline of 37% with no signs of recovery as of 2023. Since a child spends nine months in the womb gestating it's difficult to assign that drop in fertility to COVID, which was not in full swing until March. The trailing indicator of COVID influence on births would be November 2020.
If we take the fertility rate and assume the same fertility window that the report uses (18-34, 16 years) then we can figure out if we are achieving replacement rate (2.1 births per mother). That is shown in the following figure. I would agree with you if you quibbled that most WELS girls aren't married at 18 and some women have children past 34, so pick your 16 year window - 25-41 sounds reasonable.
We lost replacement rate in 2020 but were well on the path prior to that. My assumption is that the fertility rate in 2023 (roughly 11%) is held constant for the next twenty years. This is extremely conservative on my part. It is clear from the data from 2000-2019 that we were on a downward trend regardless of the disrupter in 2019. The trendline is a second order polynomial where I excluded the data from 2019-2023 and removed several outliers. It's optimistic: The fit is not great (regression coefficient 0.8), but cranking it up to a third or fourth order (with a regression coefficient of 0.98 - a pretty excellent fit) shows us not having babies as of 2023. Polynomial fits can be problematic because of knots in the data that aren't apparent to the eye, so take it with a grain of salt.
Births are one factor, the next factor is do we keep our kids. The gate we track is confirmation which typically happens in eighth grade. The following figure shows our decline in confirming our babies with a pretty tight linear fit showing a 40% retention rate by the time we hit 2045. This is the second of two assumptions I mentioned earlier - I am assuming a linear decline in confirmation.
It should be noted carefully that the reduction in births and the reduction in confirmations is not a linear effect but a geometric one.
Knowing that our fertility rate has fallen off a cliff and knowing that we confirm fewer and fewer of those fewer children, and acknowledging that our age distribution is likely top-heavy, I'm increasingly convinced my opening figure is exceedingly optimistic.
It should also be noted that there is no tracking of baptized infants to confirmations. Hein is tracking those two gates separately. This means that there are some confirmations which occur on children whose families converted after they were born - likely due to their ECE or grade school involvement. So in reality we are retaining fewer children than confirmations indicate because the group of baptized infants who are confirmed is a subset of the confirmed. This means our actual retention rate of WELS baptized infants is even lower than the trend provided.
The actual confirmations per year are given in Figure X and I provide a projection in orange based on actual births reported, and in green based on fixed 11% fertility rates, which as I said previously are likely optimistic. By the early 2040's we will confirm less than 1000 children across the synod. I still recall my own confirmation class of 50 as being large-but-not-atypical in the not-too-distant past.
Of course we lose a lot of kids after confirmation. Speaking from my own experience, of my children's confirmation classes half of the students are no longer attending a WELS church, and perhaps quarter of them are infrequent attenders.
Finally, then what does this mean for MLC? I extracted MLC freshman enrollment from Figure Y and assumed a "conversion rate" of 4% of the confirmands from the window 4-7 years prior. 4% is basically the average of the green line in that plot. The projected freshman enrollment takes into account the actual birth rates and projected confirmations based on the linear regression performed above. The MLC demand signal is created using Hein's projected 3% annual growth of called workers shown in Figure Z, and assuming the average WELS called worker has a 30-year career.
The demand signal then is the deficit between 3% growth year-over-year and the attrition of 1/30th of the workforce any given year. When the orange trend crosses the black trend, you are not providing enough called workers to maintain 3% growth. I don't like lumping all called workers (pastors, high school teachers, grade school teachers, ECE teachers) in the same bucket but there isn't enough underlying data to tease it apart.
You may question the 30 year career since many WELS pastors and teachers will act in that vocation their entire 40+ year career, but recall also the number of young women who achieve the 'Mrs. degree' at MLC and decide to retire from teaching to be a stay-at-home mom because it is a more blessed vocation (they are right!). I think 30 is a fair composite. Even if it isn't, running the analysis with 35 years or 25 years only shifts the line by less than a year - the steep decline in freshman enrollment between 2035 and 2040 makes it pretty insensitive to demand signal.
In ten years, quite possibly sooner, it is entirely likely that the synod will not have enough called workers.
Comments
Post a Comment