The other day I realized that Duke University’s National Congregations Study, which includes about 87 randomly sampled LDS wards, has information on the race and ethnicity of the “person who is the head or senior clergy person or religious leader in your congregation,” which I assume in the Latter-day Saint case is the bishop, so I decided to see if we can glean any information about how racially representative leadership is relative to membership.

Upfront, statistically this is very much seeing through a glass darkly, but frankly I think it’s the only information that we non-COB employees have on this subject, so it’s worth taking a look.

The NCS had four waves: 1998, 2006, 2012, and 2018. The NCS piggy backed off of another survey that is taken almost every year, the General Social Survey, with some individuals asked more detailed questions about their religious congregation. (For the wonks; weights with small cell sizes can get squampous, plus the GSS that the NCS is based on is a relatively self-weighted survey already, so for my purposes here I’m not going to worry about the weights). If we simply look at the racial/ethnic composition of Latter-day Saint bishops by year we have the following table. (For some reason WordPress is cutting off the image; apologies.)

As you can see, we only have 23 wards/branches in 2018, 19 in 2012, 35 in 2006, and 9 in 1998. However, this isn’t nothing. Obviously, the vast majority of leaders are non-Hispanic white, but so are the members (for now), so the question is whether leaders are more or less diverse than we’d expect given the composition of the membership in general.

The latest Pew numbers in 2014 reported that 8% of members identify as Latino, and 1% identify as Black. Given the incredibly low number of Black members, it is statistically difficult to be able to say that Black people are underrepresented in leadership positions. We’d need a much larger sample size to say that, but of the 44 wards surveyed in 1998 and 2006, none of them had a Black bishop, so make of that what you will. Only one of them had a Hispanic bishop. I don’t know what the percentage of members in the US who were Hispanics was during that time period, but it is likely that, in this sample, Hispanics were underrepresented among leadership positions.

However, things are very different when we look at the more recent waves. Of the 42 bishops during the more recent waves, six of them were Hispanic, or 14%. However, one of them was included in a Hispanic congregation oversample, so If we remove that ward we have five Hispanic bishops for 41 congregations. If we take the Pew numbers at face value and perform a simple statistical test (for the wonks, a cumulative binomial probability test), we find that, if there was perfect representativeness among leadership for Hispanics at 8%, that there is a 77% chance that four or fewer bishops out of a random draw of 41 would be Hispanic, which suggests that given their baseline population Hispanic men are actually more likely to be called as bishops.

For Black men, if we set the baseline percentage at 1%, the chance of there being only one or no Black bishops in our sample is 94%, so once again it looks like there is a good chance that Black men are, nowadays, actually more likely to be called as bishops than their non-Black counterparts.

Now, there are a lot of different assumptions baked into these numbers (e.g. what you identify as is the same as what your congregation identifies you as, the GSS’ particular cluster sampling for that year isn’t throwing things off, etc.), so I’m not married to this position, but preliminarily at least, given the numbers we have, it looks like while local-level leadership may have been disproportionately white back in the day, more recently bishops are actually more likely to be racial/ethnic minorities than the average member.

*Asians are often combined with Pacific Islanders (except when they’re not), and they were measured differently in Pew and the GSS, so given the particular salience of the latter group in the Latter-day Saint community and the low numbers overall, I didn’t think running the numbers on that category would be accurate enough to provide any kind of a plausible estimate.

**Normally I post my R code on my Github page, but this is simple enough I’ll just post it here in case anyone wants to play around with the numbers.

library(dplyr)

LDS = filter(Cong,DENOM==”Mormon”)

table(LDS$CLERRACE, LDS$YEAR)

table(LDS$WT_WAVE3_NOHISP_CONG_IGN, LDS$CLERRACE)

pbinom(4, size=41, prob=0.08)

pbinom(1, size=41, prob=0.01)