Why are Census records so frustrating?

Question

One of the great frustrations I have with doing research is being unable to find census records for some individuals. Logically, if a census is complete (and they generally would be), then every person should be in the census record. Yet, when I search for certain people, I come up with bupkiss.

Is this because no one has completely indexed the records?

I'll give you an example. I'm researching a man named Joseph B. Scott. His parents were Walter H. Scott and Laura Virginia Brown of Petersburg, Virginia.

I found the 1900 Census, but I can't find any more. I know, from genealogybank that he was still alive when his mother died in 1925 because he's mentioned in the obituary. (no location is given.) So, I ought to be able to find him in the 1910 Census records as well as the 1920 Census records.

Yet I can't. This happens far too frequently and is very frustrating.

Is there a solution to this? Or are we just stuck with a less-than-perfect system?

I found an online census site that has all the Petersburg, Virginia census records, but they're gifs, not expandable, and not indexed. The prospect of going through tons of gifs with a magnifying glass in the hope of finding him doesn't exactly thrill me.

ADDENDUM: I want to thank everyone who has contributed to this discussion. It has resulted in a rich array of possibilities for research that I will be referring to repeatedly in the future.

Comments

For some Canadian censuses, there is also a free site called Automated Genealogy, they have some interesting ways of looking at the records. You can pick the location and then search for a particular last name and the results will show all people with that last name in that area.  http://automatedgenealogy.com/

And in Britain there is also  https://www.freebmd.org.uk/,  again completely free, they have censuses and baptism, marriage and burial information for most places and dates. And they have a WT citation that can be copied.

---

And in Britain there is also  https://www.freebmd.org.uk/,  again completely free, they have censuses and baptism, marriage and burial information for most places and dates. And they have a WT citation that can be copied.

commented by M Ross

.

FreeBMD is just births, marriage and deaths.  I think you mean FreeReg.

---

---

Yes, the home page link I gave also has the links to both FreeCen and FreeReg

---

I don't think I have ever seen the home page for FreeBMD, as my bookmark is set for the search.  

Answer 1

I don't think it's safe to assume Censuses are complete in the sense that they count everyone.  I know for a fact I was not counted in a few of them. In addition, people sometimes didn't live where you look for them, or they were counted but the database is indexed incorrectly.  For the indexing problem, I find it helpful to look more than one place, say Ancestry and FamilySearch.  If you find them on one but not the other, then search the faulty one for another name on the same page.

Comments

That is a great tip. I never even thought of that. Thanks.

---

Also, try using a wild card * for vowels or the ending in a name since they can be incorrect, as part of misspellings

---

Thanks for the star Paul!

---

Ancestry are (or were) definitely missing a section of the 1851 census of Cornwall, which included one of my ancestors.  I alerted them 3 times about the missing section, I even could tell them exactly which part that was, but as far as I know they never rectified the issue.  I did give up checking so they may have fixed it by now.

So Herbert's comment to check in a couple of different providers is good advice.

---

It is also worth checking the list of missing England and Wales census entries.  Some of them have simply been lost or damaged. https://www.findmypast.co.uk/articles/census-for-england-wales-and-scotland-missing-pieces?

Answer 2

Indexing can, indeed, be bad (and often is.) The enumerators made errors in spelling. Some families actually did not get enumerated in a census. So many in my family were not recorded in the 1870 United States Census. I wish you all the best.

Edit...I am finding scads of Joseph Scott's in the 1910 US Census. Do you know if he was married and what his wife's name may have been? Are you sure he would have been in Petersburg, VA? Could he have been elsewhere? Perhaps with some more information, I or someone on WikiTree might find him.

Comments

I swear I responded to this earlier. I guess I forgot to click on Add Comment.

Yes, there are tons of Joseph Scotts in Virginia. While I was searching for my Joseph P. Scott, I found LCOL Joseph V. Scott. I ended up fleshing out his record on familysearch and on findagrave, because I found it interesting.

But my Joseph P. Scott? I don't know if he married. I don't know if he moved from the area. All I have is a 1900 Census when he was still living at home (at 22), and a mention of him in his mother's obituary in 1925.

---

So, he might actually be on a census, somewhere, but it may be difficult to determine which one right now. Steve Lake, below, had a good idea--research the siblings and other family members. He's right, doing that often does yield information that helps one narrow down the possibilities for the person who is an elusive mystery.

Good job on eliminating one possibility through your research. That is actually progress.

All the best!

Answer 3

There are numerous reasons that someone wouldn't be on a census record or you can't find them. 1950 was the very first census where people could fill out their own information instead of having someone else fill out the information. Should be interesting to see the 1950 census once it's released next year.

Anyway, prior to 1950 there were people hired to go around and fill out the requested information on the census. Since this person is highly likely an outsider that doesn't know the family, errors are pretty common. I've seen some people change genders between census records. One time I saw 8 year old Johnny listed as a male. But 18 year old Johnny after hitting puberty became female because 10 years later it was pretty obvious that Johnny was a female. 8 year old Johnny it wasn't as obvious. Fun fact, before about 1950 it was common to put young boys in a dress to make them look cute. This is how I'm sure errors are made among gender because the person filling out the information gets a brief visual on the person and assumes gender based upon name.

Also, if no one in the family was home when the person who was working the census, then they would simply move onto the next house. It wasn't worth the time to wait around for someone to come home. They may try again of course but I doubt people working the census would go more than a handful of times to the same house when no one was home.

Moving is another indicator. If someone packs up and moves during a census year they could either be missed completely or even counted in the census twice in both locations.

And yes, I feel your pain on census records. Take a look at my ancestor Butcher-4229

Born in Missouri he married, had a few children, then moved to Louisiana around 1902 or 1903. The family should be in the 1900 census in Missouri. Yet they are nowhere to be found. His wife supposedly lived until after the 1930's. She should be somewhere in the census records. Yet she cannot be found in 1930.

Then with my last name, Kerbow, we run into multiple other problems. There are numerous ways to spell Kerbow and they all come from, according to family legend, 3 brothers who came to America from France. Kerbo, Curbo, Curbow, Kerbi, Kerby, Corbeau, Corbo, Corbeaux, Kerby (though I'm not sure if Kerby is part of the ancestral line) and Kierbow are common to bounce around between census records from one decade to the next.

Comments

It was actually a less fun fact regarding little boys in dresses. Potty training was easier. You can google it!  I imagine census takers knew that.

---

It was quite the ceremony for some boys.

Answer 4

There is also the possibility that the original pages were dropped in a mud puddle, fell out of the bag they were carried in or other similar circumstances.

I realise that you are looking for United States records. In Great Britain there are documented occasions of such accidents happening to census records.

Comments

The United States lost an entire Census, except for a very small number of records. 

History Lost: The Tragedy of the 1890 Census

---

Makes it fun when someone was born in 1881 and died before 1900. Ghosts in census records.

Answer 5

My strategy has been to work on every member of a family, not just one individual.  Sometimes this comprehensive approach turns up a completely unexpected spelling or place in a search for one or another family member.  The soundex component of FamilySearch sometimes helps find more results.  Wildcards sometimes works.  When a record looks like a suspicious possibility, I look at the original image, which has more info than the indexed fields.  That usually reveals whether the indexing was accurate, or not, and the extra details help verify relevance.  In smaller counties, I have actually read the entire census originals, forwards then backwards.  I often use the plus-or-minus ten-house rule to look for relatives or associated families and have found a few families that weren't indexed.  Can't attach a census unless it is indexed, so I have downloaded the image to use as a source.  Experience teaches lots of different ways to attack census problems.  Thanks for asking the question and lots of good suggestions have shown up.