Zelophehad’s Daughters

Implied Statistical Report Graphs

Posted by Ziff

Over at T&S, Kent Larsen wrote an interesting post based on the Church’s statistical report from Conference. He compared this year’s data with statistical reports from 5, 10, and 25 years ago. Since I find this kind of speculation so entertaining, I searched lds.org and found statistical reports all the way back to 1973 to fill out the data set a little. To make the resulting data easier to look at, I’ve put some of the numbers Kent and the commenters discussed into graphs.

I apologize in advance: I am sure I will be repeating at least some of what was said on Kent’s thread, probably at times without even knowing it. I will endeavor to make at least a few original comments in addition, though.

Here’s a plot that was discussed on the T&S thread: number of members per unit (ward or branch). Please note that I have reduced the vertical axis to only the 300-500 range, so this plot exaggerates changes over time.

members-per-ward-branch

It looks like the number of members per unit is higher now than it’s been before (at least in this 35-year period). I don’t know if this reflects changing inactivity rates or deliberate policy decision (suggested on the T&S thread). The big dip around 1980 is also interesting. I wonder if this was related to the introduction of the consolidated (3 hour block) schedule. Once that was in place, then multiple wards could share a building, so more wards could be created.

Here’s a plot of missionaries per 1000 members. It looks largely similar to a plot of missionaries per unit, but I thought it might be interesting to ignore the effect of unit size. Note that it only goes back to 1977 because this number wasn’t included in statistical reports before then.

missionaries-per-1000

There are at least three processes going on here. First, there’s the declining birth rate that’s making Church members on average older, and reducing the number of missionaries per 1000 members (or at least reducing the number of young single missionaries). Second, there are changes in stringency of criteria for young people to be eligible to serve a mission. The recent “raise the bar” effort is the most obvious example, but on another thread (before the T&S one) I think Ray pointed out the earlier drop in missionaries, around 1980. Regarding this, I found a talk–I think by Mark E. Petersen–from the mid-1970s, where he mentioned how the number of missionaries had tripled in the previous decade or so. I wonder if the early 1980s downturn in missionaries was the result of a similar process that yielded “raise the bar.” Perhaps the General Authorities decided that in trying to encourage everyone to go, they were ending up with too many unqualified missionaries. The third trend that might contribute to changes in number of missionaries is changes in activity rate, but I find it difficult to tell how. (This assumes of course that the general activity rate in the Church has been changing over this time period. I’ve seen this suggested a lot, but I haven’t seen any data on it.)

Here’s a plot of convert baptisms per 1000 members and 8-year-old baptisms per 1000 members.

baptisms-per-1000

The convert baptisms are all over the place. They’re low now and they were in the early 1970s, but there were peaks over twice as high as the troughs through the late 1970s and early 1980s, and then again in the late 1980s. Since the late 1980s, though, it’s been all downhill. There may be a very slight recent recovery, one might hope associated with raising the bar, but given the size of the variations, it’s most likely just noise. To the degree that this recent decline represents an effort to only baptize people who are really interested in the Church, then it seems like a positive trend. If it just represents a decline in number of people baptized across the board (both interested and not interested but easily persuaded) then of course it is more negative.

The 8-year-old baptism rate is mostly just one long slide, with a little recent uptick. By the way, I agree with Kent on his T&S thread that the 2008 figure is way out of line with the last few years, and seems like it must be a mistake. The general trend is of course attributable again to declining birth rates, although there might be some activity level changes hidden in there too.

Finally, here’s a plot of a couple of values I came up with to try to estimate how well the Church is doing at retaining children and teens. One approach to this is to look at the number of 8-year-olds baptized in a particular year and then ask how many later served a mission. Another approach is to look at how many children were blessed (a figure no longer included in the statistical report) and then ask how many later served a mission. In this plot, the red line gets at the first question. It’s the number of missionaries in a given year as a fraction of the number of 8-year-olds baptized 11-12 years prior. (Everyone is counted twice because they’ll be counted twice as missionaries in a two-year mission.) The blue line gets at the second question. It’s the number of missionaries in a given year as a fraction of the number of births 19-20 years prior. Again, there isn’t data for this latter value beyond 2007 because “children blessed” was last included in the statistical report in 1988.

missionary-fractions

This analysis ignores at least four groups of people. First, women who serve missions don’t serve until they’re at least 21, so looking back 19-20 years to birth or 11-12 to baptism isn’t going to include them. Second, converts who serve missions won’t be included in either the blessing or baptism numbers. Third, men who serve missions but leave when they’re older than 19 won’t be included. Finally, senior couple missionaries will of course be missed.

Given that the numbers of people in these groups is not zero, and is not likely even near zero as a proportion of the total missionary force, it would be wrong to think that the fractions in the plot could be taken literally as fractions. For example, the leftmost red point does not actually suggest that 30% of children baptized as 8-year-olds in 1973-74 went on to serve missions in 1985.

If, however, we make the weaker assumption that the number of people in these four missed groups is fairly constant across time, then the relative sizes of the values in the plot might be meaningful. Having offered all those caveats, I actually don’t have a whole lot to say about the plot except that I’m a little surprised at how flat these values are across time. The baptism to mission retention line (the red one) even appears to have been increasing through the late 1990s, although it’s flattened out and declined since then.

Finally, here are a few random thoughts that came up while I was putting the data set and plots together.

  • I am surprised at how sloppy some of the record keeping appears to be. I am far from the first person to notice this, but you can estimate the number of members lost to death, resignation, and excommunication each year by looking at the change in total membership versus the number of baptisms. For several years, this is a negative number, which should be impossible. And not just a little negative. In 1989, it was -186,000. You might blame this on badly trained clerks, and I guess that’s possible, but my guess would be that if clerks were off, they would tend to be off in random directions, and would therefore cancel each other out. To make an error this big really requires mistakes at the top of the record-keeping chain.
  • The yearly statistical report used to contain a lot more interesting information. Look at the 1973 Statistical Report, for example. It has birth rate, death rate, and marriage rate. It tells how many members held each priesthood office, as well as Relief Society, Sunday School, and primary enrollments. It has counts of temple ordinances performed. Think of the fun we could have speculating with numbers like that now!
  • Related to the previous point, it appears to me that parts got cut out of the statistical report as they became bad news. For example, take the birth rate. It was increasing from 1973-1977, leveled off in the early 1980s, fell dramatically in 1983, and hasn’t been reported since. Or the number of children blessed. It increased through the early 1980s (with convert baptisms fending off the flattening birth rate), but then fell from over 120,000 per year to the 90,000 range. After a couple of years in the 90,000s, it was gone. Also, the number of temples was reported through 1985, when it had nearly doubled in three years, and then wasn’t reported again until 1998, when the most recent temple-building boom began.
  • I mentioned that the death rate used to be reported. I applied it to the total membership count and tried to match that up with the estimate of number of members lost mentioned in my first point. Some years it matched up quite well, which you would expect, I think, since likely far more members are lost to death than resignation or excommunication. For example, in 1983, the estimated number of members lost was 23,419 and the estimated number of deaths was 21,600. But that was an exception. 1980 was more typical, with an estimated membership loss of 77,000, but only 18,088 estimated deaths.
  • This is getting back to the issue of sloppy records, but I wonder how many of the oddities in these data are the result of re-definitions of the categories. I know this is an endless problem for social scientists doing longitudinal (over time) data collection, and even more particularly when they look at archival data. Change the definition of things and the counts or measures will jump all over the place, even if the underlying phenomenon remains constant. This year’s jump from five years in a row of 90,000-100,000 8-year-old baptisms to over 123,000 seems like it must be the result of some definitional issue like this. Did they decide to count 9-year-old children of members who were baptized as child of record baptisms? If I recall correctly, in my mission at least they were counted as converts once they were 9, regardless of whether their parents were members or not.
  • The Church has hugely detailed membership records, and should be able to do all kinds of interesting studies about what makes for active adult members. In fact, I’m sure they do those studies but they just don’t report them to us. For someone like me, they know, or could figure out, how big my ward was as a kid, as a teen, and as a young adult. They could figure out how many other kids were my age, how far I lived from the church building, how active my parents were, and perhaps even through matching records from different sources, how often I went to seminary. I would find it interesting to be able to poke through all these data and see what kinds of relationships turned up.

Okay, enough from me. I’d love to hear any interesting information or speculation you have on these numbers or any of the questions I’ve raised. And if you want to see more plots, let me know and I’ll try to put them up.

20 Responses to “Implied Statistical Report Graphs”

  1. 1.

    Yay! More Ziffian analysis. I found the visual presentation of the charts very helpful.

    I haven’t studied these threads closely. But the two things I was most struck by were, first, my impression is that the number anomalies are probably based on policy decisions for how categories are defined, not sloppiness in record keeping. For instance, my understanding is that people in the massive address unknown file are presumed to remain alive until age 110, which is obviously a tremendous skew from actual reality.

    The other is your point, with which I agree, that reporting is in large measure a function of what is good v. bad news. The classic example is the Church ceasing public financial reports when it got mired in really embarrassing deficit spending. And the sky didn’t fall when they stopped the public disclosures, so I think they learned from that to just not disclose bad news. I personally really, really disagree with that particular institutional mindset.

    Thanks again for this.

  2. 2.

    These are great!

    A few things:

    re: the number of members lost to death, resignation, and excommunication each year by looking at the change in total membership versus the number of baptisms It is my understanding that this number would also include children under 8 that are children of record. I get this from the quarterly report which includes children of record under the age of 8 as part of the total membership. It seems consistent that the church is doing the same in conference as it is doing on it’s quarterly report for wards and stakes.

    As Far as active members, the quarterly report works like this:
    1. We track total membership vs total sacrament meeting attendance. This doesn’t connect to an individuals name. This is done weekly, but the quarterly report takes the week that the report is done only (as opposed to an average or something.) Many Clerks round up on this.
    2. We track new converts attendance, this does attach to specific names. (This is something like “who attended once in the last month?”)
    3. We track attendance in Priesthood, RS, YW, YM and Primary, this is a number not attached to specific names for reporting, although the ward could use attendance sheets to do so if they wished. (I know none that do). (Again, this is “Who attended at least once in the last month?”)

    Hope this helps.

  3. 3.

    Ziff, I agree with you that much of what we hope to extract from statistical reports suffers from GIGO. We just don’t have quality data to start with, and the problems occur both with ward clerks and COB. Two examples: Every time I move, some other Mark Brown’s records get sent to my new ward. On my most recent move, I was called right away as membership clerk in the ward, so it became my task to try to convince HQ to take back the other guy’s record and send mine. After five weeks and four attempts, I finally succeeded. And when our son submitted his missionary application, it was initially rejected because some brainiac had decided he was the father of three children (at age 19!), and those childrens’ records were attached to his. Again, the process of getting that corrected was only slightly less complicated that organizing an excursion to the moon.

    I hope you don’t mind if I include the text of a letter that once appeared in Sunstone. (I’d just link, but I can’t find the link now.) It is funny, but also captures the frustrations a clerk experiences.

    Attention: Records Processing Division

    Brothers and Sisters:

    Inside you will find a bunch of membership records with addresses in the Bronx, New York. A whole bunch. Now, this shouldn’t be. I’ve told you about it before. Go to the Bronx. There you will find Yankee Stadium. You will find the New York Botanical Garden. You will find Van Cortlandt Golf Course, Pelham Bay Park, and the Bronx Zoo. But you will not find the Manhattan First Ward. You will not find any Manhattan Ward at all. The Bronx Ward and the Manhattan First Ward are different units; different universes, practically. Do you hear me? The Bronx is uptown and to the right.

    I know what’s causing your confusion. The Bronx Ward used to be part of our ward. But it isn’t anymore. Please believe me. If you could see my face, you would see that it is an honest face, one that could not lie. It is also, though, a tired face. I have seen much in my day, and most of it has been membership records from the Bronx. Week after week they roll in, like apples into the cellar bin at harvest; and week after week I roll them back to you, but always a couple fewer go out than came in, with the result that I am slowly smothering here in New York. Everywhere I turn there is someone from the Bronx. Beaumont, Bayshore, Bradford Park and Boscobel; Delafield, Dryser, Duncan, Debs: the rhythm of this list of streets has taken possession of my mind like a mantra, and they’re all in the Bronx, every last one of them. But the Bronx is not in Manhattan. It’s not in our ward. We’ve sent you maps. We’ve sent you notices. We’ve sent you pleas and threats. It’s the third or fourth time I’ve written. What more do you want? What more can we give? I’m a young man. I should be out tonight, on the town. I should be at a Broadway play tonight with a beautiful girl on my arm, the shriek of the city in my ears, and the double beat of summer love in my heart. That is why I came to Babylon: to live the shining and perishable dream itself. And what am I doing? I am stuffing envelopes. I am stapling. I am writing an inane letter. It’s getting to me, you see.

    Please, please. Send the Bronx membership records to the Bronx. A radical idea, perhaps; requiring no doubt a major reorganization of the Presiding Bishopric’s Office, but is it so much to ask? We have
    all we can handle in the mail we’re supposed to get, without people in Salt Lake sweeping everything off their desks into envelopes addressed to the Manhattan First Ward. I know I’m just one clerk in a
    whole religion of clerks, but your prompt attention to this matter would settle my mind and simplify my life.

    Sincerely your brother,

  4. 4.

    FYI– I don’t believe that the membership numbers are reduced for resignations and excommunications. They ought to be, but the numbers don’t add up unless you assume that these folks are still counted as members.

  5. 5.

    Mark, you left off the best part of the letter! The letter written 21 August 1980, is signed by Randal K. Quarles, who went on to serve as the U.S. Under Secretary of the Treasury for Domestic Finance. He is currently the managing director of the Carlyle Group (last I checked).

    We’ve had it over on the LDSClerks wiki for some time.

  6. 6.

    Great, great stuff!

    1. I agree with Kevin…most of the anomalous numbers are probably due to policy changes in how things are counted, rather than sloppiness.

    2. Of course, it’s hard to know, since the church seems to be quite secretive about everything. I wonder if we’ll ever find out what happened with the weird uptick in “children of record” this year? (Can’t Peggy Fletcher Stack get a mole inside the church statistical department??)

    3. I’ve heard rumors that the church does lots of inside studies based on the huge amount of data available. But I doubt the studies are all that good, since they are produced in an environment where criticism is difficult or impossible, both because there are no outside eyes looking for weaknesses, and because church doctrine and culture makes almost any criticism of policy taboo.

  7. 7.

    This is an extremely depressing post appreciated, but to me indicative of many of the problems in the Church.
    I have for many years studied the final years of the communist empire and the Present Church is redolent of that era. Report only the harvests where output exceeded forecasts. When reality does not conform to expectation just alter the reality or claim attack from without. Party functionaries are never allowed to criticize party bosses. Data are withheld from the masses so as to avoid discouragement or criticism of the system and certainly party officials. I won’t press the analogy too much. But is this a Church or an empire or a business? Does obfuscation extend from history to finances to statistics as well? The sad truth is yes.

  8. 8.

    Hey Wade, just keep it in perspective. Most other religions don’t do this at all.

  9. 9.

    I’ve kept my own spreadsheet for a few years now–I wish I knew this were quite common so I could have stolen someone elses rather than type everything in–including a trip to the ward library because the numbers at lds.org are missing for one year, one category.

    The 8yr old baptisms/1000 members is a good graph and can be compared to CoR/1000 members to get a view of birth rate. I’m amazed at how low the Church birth rate appears when looking at child adherents. The rate looks more like the US rate than the international church would lend itself to. Maybe the increased birth rate of other countries is negated by larger inactivity rates abroad?

    Another interesting trend besides the decreasing missionaries is the slowing growth rate of the number of stakes/wards. This rate has fallen to 1-2%–the lowest in recent years. The growth rate of stakes even went negative in 2000 or so with the large constriction in Chile. As Pres. Hinckley and recently Elder Anderson said, the biggest problem the church faces is its own growth. I think this is why we have heard so much about revitalizing missionary work, retention, ward teaching, and leadership.

  10. 10.

    I think “CoR/1000 members” would be a severe underestimate of the birth rate, because there are LOTS of people counted as members who are totally out of contact with the church, often with whereabouts unknown. Whenever one of those people has a child, it doesn’t get counted in the numerator, but that person still shows up in the denominator.

  11. 11.

    I agree the CoR/1000 members ratio would be an underestimate of the birth rate–kinda like multiplying the birth rate by a guess of an overall activity rate like kodos states. However the year-to-year change would be the important factor–especially if you could correct for the inactivity rate.

  12. 12.

    One reason for the uptick in 8 yo baptisms could be the new trend to send out wedding announcement-style invitations for children’s baptisms. (complete with adorable photo and invitation to a big party)
    If it’s cute and trendy, maybe more people are doing it :)

  13. 13.

    Thanks for all your comments.

    Regarding sloppiness or redefinition of categories causing the numbers to jump around, Kevin and kodos, I think you’re exactly right that it’s more of the latter than the former. I guess when I said “sloppiness” I was kind of thinking of that including both sloppiness and redefinitions. I guess I was thinking kind of sloppily (which I will now redefine as “flexibly”).

    Matt, are you saying children under age 8 who were born to members who are active enough to bless them in church (or at least make the children’s birth known to their ward/branch) are also included in membership counts? That makes sense. So what you’re saying is that the natural increase each year is based on births to active members and not on baptisms of 8-year-olds? Do you know what happens to kids in this category, though, if they’re never baptized? Are they kept on records until they would be 110 like Kevin mentioned for people who have been baptized, or since they were never officially affiliated with the Church are they dropped off at some point? I guess this makes more room for categories to be defined and redefined and for numbers to jump around.

    Mark, that’s pretty bizarre that your records keep getting mixed up with those of other Mark Browns. Perhaps you should ask the Church to correlate your record with your Strengthening the Members Committee record (or whatever it’s called) and then when you moved to a new place you could tell them, “I’m Mark Brown. The known troublemaker, not one of the regular ones.” Also, thanks for that letter from the frustrated clerk–I hadn’t seen it before.

    MoHoHawaii, I’ve heard that point made before too, but it appears to me that the numbers don’t add up even if you did take resignations and excommunications into account. I’m guessing they’re relatively few in number compared to deaths as a reduction in total membership, and it doesn’t even look like, based on the year-to-year numbers, the deaths are getting counted properly (given numbers like the 186,000 member error). It looks to me like with the deaths sometimes counted oddly, it’s hard to tell what’s going on with resignations and excommunications against that noisy backdrop. I may be wrong–is there a better way to tease these apart (perhaps based on Matt W.’s point)?

    kodos, good point about the internal studies probably not being all that rigorous. I suspect that you’re right that they’re likely done with their conclusion in mind, and if the conclusion doesn’t match what was expected, they’re just discarded.

    Wade, interesting comparison. I’m hoping the Church comes to a better end than the Soviet Union did.

    jose, I’m glad I’m not the only geek to like stacking numbers up in spreadsheets. So what was the missing number? I was too lazy to even look it up in print. What was it–the number of priests one year, I think?

    Jessawhy, I must be out of touch. I’ve never even heard of such a thing. So are people sort-of-but-not-really soliciting baptism gifts with those invitations, do you think? Is it because we’re more affluent now and have fewer kids that we want to make a bigger deal out of baptisms? I read somewhere–must have been Ardis’s blog–about a Church president chastizing people for being too slack in getting their kids baptized when the rivers were too cold. And now we have heated fonts and special invitations. My older son missed out, but perhaps I can plan such a thing for my next. :)

  14. 14.

    Great stuff. been doing my own version of this for some time. other stuff I’ve noticed:

    2007 was the only year in about three decades that overall membership didn’t grow at least 2%. That bounced back up this year.

    Excluding this year’s suspicious “boomlet”, the birth rate in the church is half what it was in the 1970s.

    The number of converts has remained relatively stable for the last 20 years (following 15+ years of steady growth).

    Those two figures are largely responsible for the church’s annual growth sinking from 5-6% in the 70s/80s to 2-3% in recent years.

    wish I knew how to post graphs…

  15. 15.

    I love it when you do the stats thing. Thank you. I only wish I could do it too.

    The issue with the number of deaths reminds me of a discussion I read in a missionary journal from 1880.

    One Mr. Holley, a local Methodist preacher, rather sneeringly spoke up to Brother Spencer, as he was talking on the healing ordinances and said, “Say Mister, I suppose if you have this power the people do not die out in Utah.” Before Brother Spencer had time to answer, Mr. Garrett spoke up quickly and said, “No, the people never die out there. They live until they get so old life is a misery to them. Then the old people will gather together in large companies and go over into Colorado in order to die.”

    Maybe this is skewing the data.

  16. 16.

    I’ve been reading Edward Tufte (sp?) all afternoon and I just wanted you to know that I have a new appreciation for how clear your graphs are, Z. Not “graphicjunk” at all.

  17. 17.

    Thanks, CWC! That’s high praise indeed. I love Tufte’s writing (and figures).

  18. 18.

    I can tell you why the numbers dropped around 1980. Kimball kicked Sonja Johnson out of the church for supporting equal rights for women, and I went with her.

  19. 19.

    I am surprised at how sloppy some of the record keeping appears to be.

    Part of this is due to the variety of quality of record keeping at the ward and stake level. I was membership clerk in two wards, back to back. In the first, the previous clerk had done a marvelous job, though there were a few things I had to clean up. In the second, the previous clerk had done a poor job, though not nearly as bad as what the finance clerk had done when I got switched to that position.

    That said, the church itself is extremely paranoid and secretive about the numbers. They were deeply embarrassed when someone from the U of U correlated the 2000 Utah census with various published numbers and found that church activity was far lower than the church had been claiming (though quite inline with what most of us clerks had long known.)

    My gut feeling is that activity rates are very dismal right now and dropping fast. If this is true, it has huge implications on church building programs and missionary programs. (From a pure numbers perspective, the missionary program has become a largely ineffective right of passage.)

    RE: The members per ward numbers.

    The 1980 drop was due to a theory pushed by Packard that smaller wards would push members to do more missionary work. (Not just speculation, this was stated at least in California.) This was taken to an extreme in Silicon Valley in the late 1980s where many wards had less than 200 members.

    RE: Uptick in baptisms

    This is likely entirely due to what’s called the echo boom. The baby boom largely ended between 1958-1964 (I claim 58, but other demographers prefer 62/64.) Due to boomers having kids, there was an echo boom/spike in the late 1960s. Many towns which had been closing elementary schools suddenly had to open them again. My youngest brother was smack in the middle of this and he is now baptizing his kids, thus it’s arguably an echo of an echo.

  20. 20.

    I wonder how much of the changes in stats has to do with the way the Church keeps records, i.e. putting them into computer software. It’s much easier to keep track of people when you’re not dealing with a lot of paper that may get lost in transit.

    On a related note, the church is slowly releasing this new Family History program/Web site (often called new FamilySearch) that actually incorporates Membership Records into a common pedigree. One of the things I keep running across is this problem where some of my ancestors have more than one membership record or who have incorrect information on their membership record (which may have caused the creation of a duplicate). With this system, Church members are actually allowed to request changes to the membership records of their ancestors. Maybe that will help clear up some of the numbers problems from the past.

Leave a Reply