Superforecasting the Church for 2023

Note: After this post went live and the organizer reached out to me, some of these specific predictions were added to an actual prediction market at Manifold Markets. 

In the past public predictions usually took the form of some pundit making a prognistication about an event that was going to happen years in the future, and by the time the prediction was falsifiable everybody had either moved on or the prediction was so vague as to be non-falisifable. However, recently the “superforecasting” movement has turned armchair theorizing into a systematic science, with predictors being graded on their accuracy after making predictions that are clearly definable and falsifiable. One manifestation of this movement are prediction markets, where people literally bet money on clearly defined events happening or not happening, and with actual money on the line, people do really deep dive research. 

So I thought it would be fun to do my own superforecasting competition for the Church in 2023. The rule is that the question has to be defined clearly enough to be demonstrably true or untrue within the time horizon. Here I’ll give percentages. I don’t claim to be a superforecaster, I’m not spending a ton of time on quantifying trends and the like, and no money is on the line, but I’ll check back on my accuracy around this time next year. 

  • The Church’s membership, on-the-books growth rate will be below 1% as reported in the April 2023 Conference. 

My probability: 80%

The Church’s year-on-year growth rate has been below 1% for the past couple of years, so getting above 1% would require reversing the long-term slowdown trend. 

However, temporary reversals sometimes happen even if the long-term trend stays the same, since growth trends can be bumpy; additionally, conversions might see a post-COVID uptick, so a growth rate above 1% isn’t completely beyond the realm of reasonableness, but it’s unlikely. 

  • There will be another scandal involving local level Church leadership and sexual abuse, either perpetrated by or confessed to the local leader. 

My probability: 80%

Given the sheer number of local leaders in the US, combined with the tragic ubiquity of sexual abuse, along with the increased awareness of this problem, I think the chances that there will be another scandal involving a local leader to be rather high–it seems like 2022 had a new one every other month or so. 

  • There will be a major, sex- or sexuality-related Latter-day Saint story carried by more than one major news outlet (excluding Deseret News and Salt Lake Tribune, and excluding sexual abuse addressed in the category above). 

My probability: 70%

For some reason people have always had a prurient interest in Latter-day Saint sexuality (The Guardian and Newsweek in particular appear to have a particular, ahem, fetish), and this doesn’t show any signs of abating, whether it’s Layton swingers (a real thing) or “soaking” (not a real thing). Additionally, the Church seems to occupy an outsized place in the sexual minorities discourse in the country, so this is one I’m reasonably confident will happen during the year 2023.  

  • There will be a significant policy change. Here “significant” means that it is addressed or referenced in more than one general conference talk.  

My probability: 50%

While President Nelson’s tenure has had a lot of these, they seem to have tapered off recently, so I’d give this one an even chance.  

  • A temple will be announced in Mongolia. 

My probability: 55%

A Mongolian temple has been a perennial favorite as a prediction for a new temple. Given the distances involved, the Church’s growth over the past couple of decades, and the Church’s movement towards building more temples to cut down on distances involved in temple travel, I put this as slightly better than even chance. 

  • The Church will open proselytizing missions in China. 

My probability: 2%

Despite being the perennial rumor, I just don’t see a political pathway to this happening in the short run. 

  • The Church will open a new country for missionary work. 

My probability: 75%

Given that a new country is opened about every year or so, and given that there is still some low-hanging fruit before we open a Pyongyang mission, I’m somewhat confident that a new country will open in 2023 for missionary work.  

  • A general authority will be excommunicated. 

My probability: 3%

This happens every couple of decades or so, so I’m giving it a low single-digit probability. Possible, but not likely.

President Nelson will pass on and a new President will be sustained. 

It would be morbid to compete on who gets this right, so I won’t be making a conjecture and I won’t count this in the overall score. The one thing that I will point out is that, according to the latest US Life Tables, somebody in President Nelsons’ demographic category (95-100 years old, white, non-Hispanic male), has an expected lifespan of 2.5 additional years. Additionally, rumor is that President Nelson is an extremely abstemious, clean eater, so I wouldn’t be surprised if he makes it to 100. However, the survival curve within the 95-100 range is pretty steep, so the 2.5 years might be an overestimate given that President Nelson is currently 98. Still, I would bet that he will be our first centenarian President of the Church. 

  • President Nelson will resign as President of the Church. 

My probability in the next year: 2%

My probability overall: 5%

I don’t think so, but nobody thought so with Pope Benedict XVI, and the idea of President Emeritus isn’t unheard of. (Somebody correct me if I’m wrong, but I believe George Albert Smith entertained this possibility?) President Nelson did live through several cases of Presidents of the Church who weren’t very functional near the end, so maybe. 

13 comments for “Superforecasting the Church for 2023

  1. Due to President Nelsons healthy lifestyle, he probably does have a few years of not being functional. On the other hand, perhaps he’ll be blessed with only a few months of not being functional. But he has seen the church operate with a non-functioning President, so he might not think of it as something that needs correcting.

  2. Hi Stephen C: I wanted to ask about your demographic prediction. Do you have any insight into how well the church is doing at documenting deaths for its inactive members (estimated at about 2/3rds of total membership)? I think this is an important question given the elevated mortality rates of the last few years.

    For instance, the church might update its membership system with death information from a database known as the GenealogyBank Historical Newspaper Obituaries, which covers the United States for the years 1815 to 2011 and was one of the earliest sources at FamilySearch to be indexed by computer.

    Since not everyone’s death in this country has been noted in an obituary, I assume the church has also used the Social Security Death Index as a more comprehensive tool for determining when members have passed away. One limitation of the SSDI is that the file available through FamilySearch covers only the years 1935 to 2014 — a cutoff that precedes the pandemic by several years.

    I also crunched some numbers and found another limitation. I joined annual death rates to yearly population totals and found about 152 million Americans died between the years 1935 and 2014. However, the SSDI for these same years contains around 94 million records, according to Ancestry. The coverage rate for the SSDI improved over time, according to the FamilySearch Wiki, from about 50 percent between 1962 and 1971 to about 85 percent from 1972 to 2005.

    Now let’s estimate how many deaths have occurred over the last eight years among the inactive LDS population in the United States. If that population averaged 4 million, and the death rate averaged 900 per 100,000, then over 8 years that would be around 288,000. It seems likely that the majority of those deaths have not yet been recorded in the church’s membership systems, given the database coverage cutoffs mentioned above.

    Do you find this demographic analysis plausible? If 288,000 is 4% of the current 6.76 million members in the United States, what implications would these unrecorded deaths have for the kind of growth rate projections you cited above? And what are the odds that the church will procure updated death records in the next year or so that would allow it to do a better job with its membership system of catching up with the deaths that have occurred?

  3. @jader3rd
    A common misperception is that living longer via better health or economic development or whatever leads to people living longer, but with more issues since their body is limping along at older ages, when my understanding is that health not only helps life expectancy, but disability-free life expectancy (https://www.nber.org/system/files/working_papers/w22306/w22306.pdf), which is a separate metric that they measure. So if President Nelson has a pound of broccoli every morning before running windsprints around temple square (or something), he not only would be expected to live longer, but to have fewer “disabled” years. However, if the issue is senile years, offhand I’m not sure how much senility plays into the “disabled” measure or whether senility as a distinct phenomenon is associated with healthy eating (I assume it is since healthy eating is positively associated with everything it seems).

    @Sterling: My understanding is that the people on the missing address file stay there until they are 110, after which the Church assumes they are dead and they are presumably removed from the Church numbers. This is one of many reasons why the membership numbers have problems.

    I don’t have a lot of experience with the NDI, but I don’t see why the Church couldn’t use it to prune the records so all the dead 105-year olds that are still on the records could be removed. Surveyors sometimes use it to see who in their longitudinal samples has died (for example, the General Social Survey did this so researchers can directly measure whether, for example, going to church means it’s less likely that you’ll die: http://web.skku.edu/~socioadmin/wp-content/uploads/2015/03/Kim_2014_Religious.pdf), so presumably the Church could do the same (the NDI has numbers up to 2020–https://www.cdc.gov/nchs/ndi/index.htm).

    However,

    1) maybe there’s a potential false positive/negative problem with using the NDI that would be infeasible to correct at scale. If all you have is the date of birth and name across the whole US, matching identities could become difficult at scale (you’d probably have to use some sort of “fuzzy matching” too for slight name differences–I have a son whose name on the Church records is slightly different from his legal name).

    2) Still, even given raw membership’s problems as a metric it more or less tracks the trends of other indicators that are probably a more reliable indicator of growth such as new units, so I’m okay using them.

    3) Finally, for internal purposes I doubt the Church would be that interested in using resources to refine the raw membership members, since I assume internally they on more refined and valid indicators (such as sacrament meeting attendance), although using the NDI might be helpful in cleaning up ward lists (in the US at least).

  4. @jader3rd

    A common misperception is that living longer via better health or economic development or whatever leads to people living longer, but with more issues since their body is limping along at older ages, when my understanding is that health not only helps life expectancy, but disability-free life expectancy (https://www.nber.org/system/files/working_papers/w22306/w22306.pdf), which is a separate metric that they measure. So if President Nelson has a pound of broccoli every morning before running windsprints around temple square (or something), he not only would be expected to live longer, but to have fewer “disabled” years. However, if the issue is senile years, offhand I’m not sure how much senility plays into the “disabled” measure or whether senility as a distinct phenomenon is associated with healthy eating (I assume it is since healthy eating is positively associated with everything it seems).

    @Sterling: My understanding is that the people on the missing address file stay there until they are 110, after which the Church assumes they are dead and they are presumably removed from the Church numbers. This is one of many reasons why the membership numbers have problems.

    I don’t have a lot of experience with the NDI, but I don’t see why the Church couldn’t use it to prune the records so all the dead 105-year olds that are still on the records could be removed. Surveyors sometimes use it to see who in their longitudinal samples has died (for example, the General Social Survey did this so researchers can directly measure whether, for example, going to church means it’s less likely that you’ll die: http://web.skku.edu/~socioadmin/wp-content/uploads/2015/03/Kim_2014_Religious.pdf), so presumably the Church could do the same (the NDI has numbers up to 2020–https://www.cdc.gov/nchs/ndi/index.htm).

    However,

    1) maybe there’s a potential false positive/negative problem with using the NDI that would be infeasible to correct at scale. If all you have is the date of birth and name across the whole US, matching identities could become difficult at scale (you’d probably have to use some sort of “fuzzy matching” too for slight name differences–I have a son whose name on the Church records is slightly different from his legal name).

    2) Still, even given raw membership’s problems as a metric it more or less tracks the trends of other indicators that are probably a more reliable indicator of growth such as new units, so I’m okay using them.

    3) Finally, for internal purposes I doubt the Church would be that interested in using resources to refine the raw membership members, since I assume internally they on more refined and valid indicators (such as sacrament meeting attendance), although using the NDI might be helpful in cleaning up ward lists (in the US at least).

  5. I’ll play!

    Membership growth: 2023 will be the first fully post-pandemic year, so missionary work should rebound some. I’m not optimistic about long-term growth in “the West” but 1% is a low bar. I’ll say there’s a 40% chance growth is below that.

    Sexual abuse: I agree with you overall, but I think the probability of a scandal where a church leader fails to act on information given to them after the Arizona tragedy made the news is only 10%.

    Church policy change: My sense is that President Nelson came in with a set of changes he wanted to make (and maybe an unusual willingness to make changes) but now he’s done and the rate of change is back to baseline. So I’m going to say 10% chance of a major change this year. The interesting question is whether a flurry of changes with a new prophet will be the new normal. I think President Nelson’s changes have been inspired, so It may be a question of whether the members are ready to receive further light and knowledge at a greater pace.

    Mission in China: This goes so strongly against “Xi Jinping Thought” that I suspect it would require him being removed from office. That’s not as unthinkable as it was a few months ago, but still very unlikely. And a new regime would probably have bigger issues to worry about in its first year than relationships with the Church. I give it 0.1%.

    President Nelson passing on: His good diet, activity level, and apparently pretty good health right now could mean his risk of dying in the next year is half that of the general population of 98-year-old American men. But that’s still a high risk. Anything could happen.

    President Nelson stepping down: I could see him being the one to introduce emeritus status for the First Presidency and Quorum of the Twelve under the right circumstances. Say, he had been diagnosed with early stages of Alzheimer’s years ago so he became prophet with full capacity but could feel it slipping away. But there’s no sign of that. At this point he doesn’t have to worry about long-term problems. Any mental incapacity would have to be fairly sudden, like a stroke, and wouldn’t give much opportunity for responding authoritatively. (Recall President Hinckley’s lament about the counselors in the First Presidency under an incapacitated prophet having responsibility without authority–President Oaks is not going to make an incapacitated President Nelson emeritus.) Note that President Nelson arranged to have his successor be the one who would take charge anyway. Probability: 1%(ever)

    One of my own, though I need to come up with concrete criteria for it: a major break between the Church and the Religious Right. There’s conflict over abortion (the Church’s abortion policy does not satisfy the hardcore Pro-Life) and the Church endorsing the Respect for Marriage Act, but I think it really comes down to Christian Dominionism vs. the 11th Article of Faith. If DeSantis becomes the GOP nominee for president vaccines could become a major culture war issue, and once again the Church would be on the other side. That doesn’t fit into the Dominionism vs. Pluralism framework though.

  6. Oh darn, I was hoping based on the picture that one of your predictions would be that the Church would disavow polygamy! Can I add that to the list? The likelihood is probably single digits, but this zeitgeist seems to be picking up momentum since Carol Lynn Pearson’s book Ghost of Eternal Polygamy was published. I recently published a thorough analysis of the doctrinal underpinnings used to justify “many wives and concubines, called “An Enemy Hath Done This: the Seed and Weeds of Polygamy.” If the prophets continue pleading with the women of the church to speak up with our impressions and insights, polygamy will not be in the canon for long.

  7. The church disavowing polygamy would start a firestorm of “what else did the past leaders get wrong” with the members that the church would never recover from. Everything would be in question… more than it already is. I guess you would get those who will just follow the new belief/change/policy and not question, like the word mormon deal happening now. How the majority of the members go along with that when every “prophet” before didn’t have an issue with it, is beyond belief and a little scary to me.

  8. On President Nelson: Being a former heart surgeon, he knows how to eat to maintain his health, but he also has very good genes. His father lived to age 93, and his mother lived to a couple of months short of 90. Considering his obvious health and vigor, I really can’t imagine him not making it to 100. Maybe 105?

  9. In my mission we had branches with hundreds of members as a result of baseball baptisms. They didn’t know they were members and would be starting to die off.
    When all the baseball baptisms come off the records the membership numbers should go backwards, but there seems to be a culture that requires area authorities to report growth, even when the reality is otherwise.

  10. @ RLD: You reminded me, I should have explicitly made the probability of a major change conditional on President Nelson staying on as President of the Church. If there’s a new President of the Church in 2023 presumably the probability of major change would increase.

    In terms of the “missions to China,” we tend to psychologically cluster small numbers together, as if 1% is the same as .1% when one is really 10x the other, so one way of thinking about this is how many times would we need to re-run the year 2023 through before we get one case of what we’re talking about. In my 2% case, we’d run the 2023 simulation 50 times, which does seem a bit low, so maybe a 1% (100x?), but I do think it’s more likely than once in a thousand.

    According to some ways of measuring it we could say that break has already happened. On two issues (recognizing the validity of the 2020 election and coming out in favor of vaccines) the Church has taken an official stance that directly conflicts with signature issues on the “far right,” (I don’t know about “religious right,” maybe).

    @Gwendolyn: Well, one thing we do agree on is that the probability for such a change is pretty low.

    @Tom: That’s interesting, I didn’t know about his family history. If his father as a (at least sometimes) drinker lived to that age, how much longer would his teetolling son live? 105 is pretty rare, but not unheard of for General Authorities (https://en.wikipedia.org/wiki/Eldred_G._Smith).

    @Geoff: That’s why the fact that membership numbers kind of track new branches is kind heartening given all of the shenanigans that warp the meaning of the raw membership numbers.

  11. Making it concrete: a major mainstream news outlet will run a story about two or more significant leaders of the religious right claiming you can tell we aren’t Christians not only because of our messed up theology, but because of our positions (or general squishiness) on culture war issues. 15% in 2023; 35% before the 2024 presidential election.

    And yes, I think a mission in China in 2023 is such a low-probability event that if we could rerun the year 1,000 times we’d expect to see it happen once.

Comments are closed.