Years ago I discussed the highly plausible possibility that Latter-day Saints are disproportionately gay (at least for males), because our large family sizes mean we have a higher chance of having older brothers, and older brothers, or the “fraternal birth order effect” has been shown to have a significant influence on male homosexuality. 

At the time I didn’t come up with an estimate for how much more homosexual the Latter-day Saint population is, but I have since done so in a back-of-the-envelope kind of way, which suggests that men raised Latter-day Saint have about 9% more homosexuality than the general male population.

(And yes, I know that referring to “homosexual” as a noun is generally rude, but in this context we are very specifically dealing with the attraction element of sexual orientation and not the identity necessarily, so I think it’s apropos in this case). I think my numbers are right given the premises, but as always with things like this it’s nice to get another pair of eyes and I’d appreciate any insights anybody has. 

Wonk alert start

First, we need to determine about how many more older brothers, on average, Latter-day Saint background men have. We don’t have this kind of data directly anywhere as far as I know, but we can estimate it from the number of children their mothers have. 

Specifically, Preston’s Equivalence (1976) gives us a formula for converting number of children to number of siblings of the children’s generation. (Personal aside, Sam Preston was wrapping up his career when I was a graduate student at UPenn. He didn’t advise me, but I remember once coming into his office with North Korean data for this or that project back in my ah-shucks, naive youth when I thought that all data that came out of the UN website was gospel truth. After generously working with me for a bit he asked where the data came from, and then laughed and told me not to bother). 

WordPress doesn’t render equations very well so I’ll paste it in (also see here for an explainer).

Where sibship size=F, average number of children born to parent generation is G, variance of number born is the numerator, and average number of children born to parent generation is the denominator.

So it’s a function both of the average and the spread. To come up with an estimate of the average and variance of children ever born I grabbed these numbers from the recently released 2024 wave of the Cooperative Election Survey for Latter-day Saint women between the ages of 45 and 65 (after they’ve completed their childbearing, but still within the same generation more or less). The 124 women we have yields 2.74 children on average with a variance of 3.9. For non-Latter-day Saint women we have 1.76 with a variance of 2.29. These numbers are a bit lower than we’d expect given the TFRs of around 2.0 in the 2000s, but I’m cutting the data up so finely that I’m ignoring weights here, but this should suffice as a point of comparison. 

Plugging these numbers into the Preston Equivalence yields a Latter-day Saint having an average of 4.17 siblings, while the average non-Latter-day Saint American has an average sibship of 3.06, so the average Latter-day Saint has about 1.11 more siblings. 

But how many more older brothers does that translate to? First, since about half of the siblings will be sisters, we cut that number in half, so .555. However, this will yield the number of total brothers, when the fraternal birth order effect only matters for older brothers obviously. If we estimate that on average half of all brothers will be older and half will be younger we cut this number in half again to yield an estimate for the number of additional older brothers specifically, which is .2775. 

So the average man of Latter-day Saint background has about ¼ more older brothers than the average non-Latter-day Saint background man. Assuming linearity, since one additional older brother increases the chance of male homosexuality by 33%, if we multiply .33 by .2775, that yields an effect of 9%. 

If we do a robustness check and increase the non-Latter-day Saint TFR to 2.0 with the same variance, that equals 1.02 more siblings, which equals .255 more older brothers, or 8% more likely to be gay. 

Wonk alert end

So, even with some fudge factor we can say that Latter-day Saint background men are probably around 8-9% more likely to be gay than non-Latter-day Saint men in the United States. (It’s worth noting a common misinterpretation of “increase of X%.” This means that if the baseline rate of male homosexuality was, say, 1%, then a 9% increase would increase it to 1.09%, NOT from 1% to 10%).