Notes from a 1940 Census Arbitrator

My name is Amy and I’m an arbitrator for the 1940 U.S. Census Community Project.1940 U.S. Census Community Project

Ok, you can stop throwing things at your monitor now. (And really, would your mother be happy to hear some of the words that just came out of your mouth?!)

If you’re not familiar with the project, arbitrators are those who referee between the sets of values from the two independent indexers. If Indexer A said the first name was David and Indexer B said the first name was Daniel, the arbitrator has to decide which one was right. (If neither was right, the arbitrator enters what he or she believe is the correct value.)

Since the 1940 census indexing project started, and particularly in the past three weeks, arbitrators have become, at best, persona non grata or, at worst, pariahs of the project.

Indexers can review their batches and see where the arbitrator chose a value other than theirs. This was intended to help indexers see where they’ve made mistakes and to help them be better indexers.

Since nobody knows who the indexers or arbitrators were of any given batch, the indexers don’t know who specifically to complain about. Consequently, indexers complain about arbitrators as a whole.

I gotta tell ya, the past few weeks have not been easy for some of us who are arbitrating the 1940 census.

Let me continue by saying this: There are some bad arbitrators out there. There are some who have not read the updated rules on the FamilySearch wiki, nor the update that appears every time they open the indexing program. There are some who don’t choose “<Blank>” for a 1935 if the person was under 5 years old. (Hello — if they were less than 5, they weren’t even living in 1935!) There are some who expand “R” to “Rural.”

Of course, what bothers people the most is when an arbitrator changes a name (either a person or a place) that the indexer knows is right. Hey, I feel your pain! Been there, done that! I had an arbitrator change my “Broyle” to “Boyle” (there was definitely an R in there) and change “Uhrichsville”, Ohio to “Yrichsville”, Ohio.

But before you go to string up the closest arbitrator by his or her toenails, I’d like for you to think of a few things:

  • Arbitrators are human. As such, they will occasionally make mistakes.
  • You don’t see how many times the arbitrator chose your value instead of the other indexers. Think about all records with strange names and bad handwriting where the arbitrator said you were right.
  • Ask yourself if the different value will really make a difference in someone finding the record. I just explored Broyle/Boyle by doing a search on FamilySearch in the 1930 census. Turns out that searches for John Broyle also gives me results for John Boyle and John Boyles. So even though the arbitrator changed Broyle to Boyle, it should still be discoverable. Similarly, changing that “R” to “Rural” isn’t going to keep anyone from finding that entry.
  • If the arbitrator changed the name to something with a wildcard, it is still discoverable. For example, if they changed your “Burns” to “B*ns”, it can still be found by anyone doing a search for Burns, Byrns, Benns, Borns, Bynns, etc.
  • FamilySearch keeps all of the indexed values: Index A, Index B, and, if applicable, what the arbitrator entered. They’ve said that they will eventually add a search option to go across all values; however, they have not announced a time table for this.
Yes, there are some doozies of names being changed and it is never fun to see your entries changed when you believe they’re correct. But remember that the change often does not affect the ability of someone to find the record….  and arbitrators are human, too.
Posted: April 29, 2012.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  • Amy,
    Thank you for your post. I appreciate your comments. (no, I didn’t have anything bad to say or think, and my monitor still works)
    Is there a place where an Indexer can go to, to address “how do I handle ….”.

    For example, I started 3 downloads this morning that I just couldn’t handle the names. So, I just returned the batch. That’s pretty simple, and would GUESS that is what I should do.
    But, how should I handle those “names” or places, that I just can’t figure out, using the Hints that are available to us. Rather than take a “bad guess”, leave it blank, or enter my best guess.

    I take it, that we should be aware of the Age of the person when dealing with the 1935 question.

    I have been entering the data, as was suggested many times, by column, so I have long forgot, and don’t look back to the Age column when entering the 1935 data.

    It sounds like a fine line between “enter what you see” vs entering a reply because the enumerator was wrong? If that is the case, on which side should the indexer fall. I guess the enumerator could be wrong, just as the indexer could be wrong.

    I may not have looked at the right place for those types of questions.

    Also, what is considered at level at which an indexer should “stop” indexing. Is it better to continue to index and let the process resolve the issues?

    Thank you,


    • Russ,

      There are a few different places where you can get answers to specific “how do I handle ” questions. FamilySearch Indexing has a Facebook page: I went there earlier today when I had a situation come up with a page number that the enumerator changed. Another resource is the “Unofficial Share Batch Group for FamilySearch Indexing” group on Facebook: They are awesome when you’re having trouble reading a particular name or you have some weirdness with a record.

      On Twitter, there is @FamilySearchInd, which is the official handle for FamilySearch Indexing. You can also ask a question with the hashtag #1940census. A lot of indexers and arbitrators follow that tag and are quick to help.

      I think you did the right thing returning the batches that you weren’t comfortable with. That’s the great thing about indexing. If you get a batch that you’re not comfortable with, just send it back for someone else to work on. No shame, no guilt, no pressure.

      I also index in columns. (It seems so much faster!) When I come across a person who is under 5, I go ahead and tab over to the 1935 city field and mark it “”, then go back and get the rest of the age column. Sometimes the enumerator did it right and left it blank, but there are times when they got ahead of themselves and marked everyone “Same House.” 😉

      As for what level should an indexer stop indexing? I’ve never seen a percentage threshold. I would say that if you’re not quite sure of your abilities, keep going to the “Review Batches” and see what the arbitrator changed. It really is a good tool for learning. (Confession time: that’s how I learned that I indexed the wrong column for the household in the first batch I did! Didn’t make that mistake again!)

      Also, try to stay in states where you’re more familiar with the names. I’m comfortable deciphering German surnames in Ohio, Michigan, and Indiana. I’m not comfortable deciphering, say, French surnames in Louisiana; therefore, I don’t index or arbitrate Louisiana.

      And remember — there’s no shame in returning a batch. There’s also no shame in taking your best guess. Sometimes, that’s all you can do!

  • A friend sent your article to me as she was tired of hearing me rant about the arbitrators! You have raised some good points, and I will try to be less contentious in the future :-). It IS challenging indexing where some of the first names seem to be made up on the spot! I suspected that some indexers (see above comment) were returning those difficult batches and leaving them for the rest of us to struggle through. Oh well, I still have a 98% success rate, so I will keep plugging along. I review all my results and strive to do the best I can. Thanks for the reminder that you and other arbitrators are doing the same!

    • I know what you mean about the first names, Marianne! People think that celebrities give their children weird names today. They should see some of the names in 1940!

      I’d say if you have a 98% success rate, you’re doing fantastic! Congratulations!

      Striving to do the best we can — that’s what all of us should be doing. Indeed, that’s all we really can do.

  • Excellent job Amy! Way to stand up for yourself! I haven’t really taken any arbitration that personally yet. I think part of it is that I know how hard it can be to decipher the handwriting. If the arbitrator says it’s Shmidt and not Schmidt… I tend to trust them. It’s not like these things are written in stone. Goodness knows that on Ancestry there are many names I’ve had to offer corrections to for past census records! Keep up the good work!

  • Is it possible to report the potential for a misspelled name. I looked at the 1940 census for my family in Allentown, Lehigh, PA, ED 39-36, 1st page. The last name is (Claude) Covely, but the handwriting makes it look like Conley or Corely.

  • Thank you for a great post! I just became an arbitrator last night and have six arbitration batches under my belt so far. During arbitration some decisions are easier than others. During an awesome webinar last night by DearMyrtle and her panelists, I asked a question about what we arbitrators should do if we can’t decide which indexer is correct. I’ve already run into that problem and took their advice to look at the 1930 census, etc. That has helped me make decisions, and sometimes both indexers were wrong. Thanks for the supportive words for us arbitrators out here!

  • Hi Amy, I am both an indexer and arbitrator. I don’t get upset by judgment calls in my entries. Of course I shake my head, but they are judgment calls. Often I agree with the changes. I do get upset when I get zinged for 9 children, on one batch, aged 4 and under where residence in 1935 is chosen over my “blank”, or worse yet, when an arbitrator chooses to enter the county and state for almost an entire page of people who either lived in the “Same Place” or “Same House” in 1935.

    People should not be arbitrators who don’t know and follow guidelines, period. It is not fair to indexers. Worse yet the arbitrary arbitration confuses less experienced people. If they can keep accuracy percentages for indexers, there should be a check for arbitrators.

    It also feels like we are being told “they’re working on it” but nothing is being done.

    As an arbitrator, I would be willing to discuss every decision I’ve made with anyone who has any question. I’ve spent hours worrying with some batches, have even had to sleep on them, because I wanted to be as accurate as possible. I realize that there are many, many good arbitrators but the bad ones are like the rotten apple that spoils the whole barrel.

    There is accountability in place for indexers, there should be for arbitrators also. Until there is hostility will exist.

  • Thanks for showing us the other side. I have been indexing the 1940 Census and have gotten frustrated when the arbitrators are ignoring the rules that are visible in the help menu. I am enjoying indexing and keep doing it. My results are in the 99% range so we must be agreeing more than most! P.S. Do you ever arbitrate Texas?