Genealogy from the perspective of a member of The Church of Jesus Christ of Latter-day Saints (Mormon, LDS)

Tuesday, March 6, 2018

Please find the babies!

Every so often, I address a topic that I think is so important to genealogy that I post it on both my blogs. This is one of those topics.

The reverberations of the recently announced study conducted by staff and others will continue to affect the way we do genealogical research for a long time. Here is the citation to the article.

Kaplanis, Joanna, Assaf Gordon, Tal Shor, Omer Weissbrod, Dan Geiger, Mary Wahl, Michael Gershovits, et al. “Quantitative Analysis of Population-Scale Family Trees with Millions of Relatives.” Science, March 1, 2018, eaam9309.

One of the most shocking findings of the study involves the number of babies missed in doing research. My son Jared Tanner, a neuropsychology professor at the University of Florida, has been reading the publication and we have been writing back and forth about some of the startling conclusions that can be drawn from the data supplied with the article. Here is a comment he recently sent to me about the babies missing from our genealogical research. 
For every 100 people born in the U.S. in 1900, there should be (on average) 10 who died before age 1. Of those 10, based on this article, we expect only 6-7 are listed in a family tree. The other 3-4 are missing.

If there are higher infant mortality rates (e.g., in 1800 [there are not great records]), there are more missing babies but at the same rate (50%). So if infant mortality was 20%, there would be 20 deaths for every 100 babies born. 6-8 of them would not be listed in a family tree.

Infant mortality rates could be much higher or a little lower in any given location and race/ethnicity. In the U.S. black babies died at higher rates than white, on average. So there are more unknown black than white babies.

At a minimum, family trees on average are missing 3% of infants born around the year 1900. That percent is higher (up to 10% or even more) for other times and locations.

My interpretation of the statistics could be wrong but, as you said, there are many missing babies. Someone is watching over these fallen sparrows.
Here is an earlier statement Jared made:
One interesting finding is the predicted mortality for infants dying < age 1 is 50% lower than expected. This means the number of included names of infants who died before age 1 is about 50% lower than it should be. If we use a conservative estimate of 10% of children dying before age 1 (in the U.S. in 1900;, for every 7 included names of children dying before age 1, there are at least 3 missing. It's likely quite a bit higher (1700 and 1800 mortality rates were likely higher). There are a lot of missing babies in the family trees.
For me, this is a very emotional issue having just gotten back from taking care of a newborn grandbaby. But also because of my effort to digitize the Mesa, Arizona Cemetery Records. This collection, now on, has records of hundreds of babies who died in the early years of the settlement of Mesa, Arizona. While I was digitizing the records, I had to avoid reading them because I got too sad and could not work. If you want to see the images of the records, here is the link and citation.

"Arizona, Maricopa, Mesa City Cemetery Records, 1885-1960." Database with images. FamilySearch. 27 January 2017. Mesa City Cemetery.

This study, sponsored by will eventually have a tremendous impact on how and possibly why we continue to do genealogical research. 


  1. Just wondering: since so many children in so many family trees have just a birth date and no marriage or death information, how did the study determine if these children lived past a year of age or not? How the researchers handled this issue would have a profound effect on any statistics of missing children.

    1. I think the study focused on the "missing" children. As we research back in time in families, we can see where there are likely missing children. They probably used the statistics showing the average number of children in a family during each time period and compared that number with the number shown in family trees. That is a really good question.

  2. My heart breaks when I find a family where it is obvious there must have been a baby born (too many years between known children) but there are no available birth records, death records or burial records for the location and date. If I could find at least one of those record sets, I could do further research and maybe fill the "hole".