Why is the 1900 US Census showing that children were born after the first census they appeared in?

Question

I'm really scratching my head trying to figure this one out.  I'm looking at the children of Bernard and Mary Lee who lived in Atchison County, Kansas.  They appear in the 1865, 1875, and 1895 Kansas State Census, as well as the 1880 and 1900 US Census.

The estimated birth year based on age varies from year to year, sometimes wildly.  But the 1900 US Census gives an exact month and year of birth instead of their age.  This would be great, if the date given didn't interfere with them being alive for the previous censuses.

This occurs with several of the Lee children, but let's look at one for example.  Annie Lee, according to the 1900 census, was born in February 1882.  This can't be right, as she was alive and present in the 1880 census, where it states that she was 3 years old at the time.  I'd attribute this to a child dying in infancy and them naming a second after it, if it weren't for the fact that it also occurs with her siblings, Daniel, Bernard, and Margaret.  It also occurs with her sister Alice, but coming from the 1895 Kansas census as she wasn't present in 1900.

Can anyone explain why the census that gives an exact month and year of birth instead of an age is so logically wrong?

Comments

Although I frequently use the month and year from the 1900 census, it is not exact   It depends on three people.  First, the person who gave the information to the census taker, second to the census taker, third to the indexer.  You can eliminate indexing problems by studying the image.   I make a note in the source if I find significant indexing issues, like the whole family is indexed under the name Johnson but the image shows Zipperer.  True story.  Additionally, we usually assume that the census information came from a person in the family  but legally it could come from a neighbor.  In the 1940 census  the person giving the information is marked.

Answer 1

D.,

Remember that the information in the census was added by a census worked who usually did not know the details of the family.  (I have records where the census taker was the family head so there are exceptions.)  That means that the census taker asked someone - hopefully someone that knew, but not always - older adults may have been away working or taking care of a family member at another location.  And in extreme cases a neighbor was asked for the information.  I have family cases where the father lied about ages of children because he had them out working - this was verified by family stories.  So - what to do?  In most cases I suspect that the earliest record is closest to the correct date, not always but if I am certain that it is the same individual that is alive in the 1880 census and 1900 and 1880 has an earlier date, then I go with the 1880 date.  Certainly no later than 1880.  Remember that the census taker is demanding a month of birth, the individual has no idea - it is better to make something up than upset that government man.  Or the census taker made something up because his boss, (almost all census takers in early records are male - if they had used the local school teachers, almost always female, the handwriting might have been better.) demanded that all of the boxes be filled in so something was added.  I worked on the 2000 census and can tell you that every problem that you see in all of the old census records will continue to be true in the 2000 census - I could go on but I think this makes the point.

Comments

Doing a quick check of my own grandparents and great-grandparents, they were all absolutely correct in the 1900 census.

I'm shocked! It seems like I've seen plenty of cases where that data is wrong. I consider it a poor substitute for what I might call a "real" source, for all the reasons described here There's just to much that can go wrong, and if it CAN go wrong it definitely WILL. It's just a question of "how often"?.

Answer 2

The problem might lay in the number of Annie/Anna Lee’s in Kansas. Granted some are Lee by marriage, but a number are Lee by birth.

Comments

This one fits on age and parents' names, though not much else

https://www.familysearch.org/ark:/61903/1:1:MMY6-K81

So clearly there were two couples in Kansas called Barney and Mary Lee.

---

RJ, the Anna you linked lived in Reno County Kansas, halfway across the state from Atchison.  While she does fit the bill better date wise, that is a long distance away.  Also it shows her parents being from New York while my focus Annie Lee's parents were born in Ireland.

Edit:  I see now that you addressed this in your comment.  Carry on.

Answer 3

This happens because the person providing the information to the census taker simply isn't fully aware of the birthdates. The months may be right but the years may be "best guesses."  I have one record where the ages are off by 25 years. give or take.