This is in reference to Explaining unknown person found in family listing in 1940 Federal Census of Iowa?

Here, the questioner is asking about how to interpret an unusual finding in a US Census record from 1940. The poster happens to know that some of the people on that page are still alive, because they are her/his family. But any of us could surmise that on any given page of the 1940 Census, there is likely to be at least one, and probably multiple, people listed there who are still alive.

The 1940 Census is publicly available online to anyone with access to the internet or who goes to a library or genealogy/research center. Census records are sealed for 72 years after the head count for the sole purpose of maintaining privacy, but there is no expectation that all the people listed are dead.

Many other records with living people are also publicly available. From phone books to survivors in obituaries to birth and marriage indexes (or sometimes even certificates).

We really need to see the Census record (and surrounding pages) the poster is asking about in order to give an informed opinion on her/his question. It's possible to give general advice without it, but it does make a difference.

Should we allow links to images of publicly available documents, even if one or more people listed in them is alive?

  • Here is another example: genealogy.stackexchange.com/questions/15142/… There is a link to a document created in 1923 for a 2 year old. She was born less than 100 years ago and could possibly still be alive. Looking at the document was vital for answering the questions about spelling of the hometown and who filled out the document. And this question has not been closed (I don't know if it has been flagged). Feb 3 '19 at 17:18
  • it's a LINK not an image -- there isn't an answer to this question that has ruled out links.
    – user104
    Feb 3 '19 at 17:44
  • 1
    @ColeValleyGirl My question specifically asks if links are okay and not everyone is saying yes. Feb 3 '19 at 17:58
  • 3 out of the 4 answers say: yes. @JanMurphy is questioning the whole basis of the question. and whether links are useful, not addressing whether they should be allowed.
    – user104
    Feb 3 '19 at 18:01
  • @ColeValleyGirl so we don't have consensus. Do we still have an okay for including links to docs that might be of living people? Can I ask the author of the question about the 1940 Census to include a link to it and give us hints about which family it is, without using names? Feb 3 '19 at 18:03
  • As far as I can remember, we've never disallowed links in the past... just images and plain text identifying information. Maybe wait for @JanMurphy to respond to my request for clarification?
    – user104
    Feb 3 '19 at 18:06
  • I think you should strike "or images" from "Should we allow links or images of publicly available documents, even if one or more people listed in them is alive?" unless you meant "Should we allow links of images of publicly available documents, even if one or more people listed in them is alive?" or "Should we allow links to images of publicly available documents, even if one or more people listed in them is alive?".
    – PolyGeo Mod
    Feb 3 '19 at 23:55
  • @PolyGeo I meant would we allow either links to images or the images themselves. We do allow images on this site. But if your answer is links only, no images if the people in the document are <100 years old, then that's what it is. Feb 4 '19 at 1:47
  • As per my answer at genealogy.meta.stackexchange.com/a/3411/19 I think links to public documents are OK in an instance such as this, but a copy of all or part of the image from the link would not be unless it had been scrupulously redacted of all potentially living person details. Copying all or a large part of an image into our questions is problematic at any time.
    – PolyGeo Mod
    Feb 4 '19 at 2:17
  • @PoyGeo do we need some guidance from the SE powers-that-be -- our original privacy policy was formulated (in co-operation with them) when their ToS said "Subscriber represents, warrants and agrees that it will not contribute any Subscriber Content that ... (c) infringes any intellectual property right of another or the privacy or publicity rights of another ". But it doesn't say exactly that any more. We may decided to stay with the policy we have, but should at lest know the current constraints?
    – user104
    Feb 4 '19 at 15:29
  • What is the next step? Should we create a SE Meta post and then link to this one? @ColeValleyGirl Feb 4 '19 at 16:12
  • Or ask the mods to consult via their channels (that's what we did originally).
    – user104
    Feb 4 '19 at 16:13
  • Okay, @PolyGeo what do you think? Feb 4 '19 at 16:15
  • I'm comfortable with our Privacy Policy in the help center and the clarification proposed in my answer here but I'm also consulting via mod channels to see if that stands up from an SE perspective.
    – PolyGeo Mod
    Feb 4 '19 at 23:28
  • And also presumably to see if it can be relaxed (as there are varying opinions around which a consensus might form if we have different constraints). @PolyGeo
    – user104
    Feb 5 '19 at 12:47

Our policy is to not publish the names of living people. We shouldn't illustrate questions with links to un-redacted images. We don't have the permission of the living people on the image, so why should we expose their names indirectly by linking to it? I'd rather see someone clip out a part of the page and redact the names, not unlike what a presenter might do for a webinar where DNA matches are being shown on screen, and the emails and kit numbers and names are blurred out.

See Guidelines for Sharing Information with Others Recommended by the National Genealogical Society, a PDF download, for one example of recommendations for good practice on how to share information with others.

We really need to see the Census record (and surrounding pages) the poster is asking about in order to give an informed opinion on her/his question. It's possible to give general advice without it, but it does make a difference.

I don't think we need to see the image in order to answer this question. The question asks:

Can you offer any insights, or suggestions of sources I could check?

There are two ways to answer a question like this: a "little picture" solution and a "big picture" solution.

Our own personal insight into how often a census household doesn't compare to one we would expect is limited -- unless we are conducting a One-Name study or a One-Place Study, we've only looked at a limited number of pages for any given census. We don't have the in-depth knowledge of someone who has done population studies, or studied the community extensively, or has done the statistics to determine whether or not the census has under-counted the population.

Taking the "little picture" view, I can offer examples from my own research in other census years where people have been enumerated in ways that make Ancestry assign them to the wrong household. I can talk about all the clues I found that made me step back and take a second and third look at the census pages to see what was going on -- but that isn't a satisfactory answer to the question about whether errors in the census are common.

From a "little picture" view, I could advise someone to do newspaper research to see what they could find out about the enumerator for that district. In England, I found several long articles about how the Bishop had chastised the local vicar in one of my study places for not doing his job. It was a 'lucky dip' -- those articles directly answered my question about why I was seeing irregularities in the baptism register.

For a "big picture" view, a researcher can look for population studies that addressed under-counts and other irregularities in the census.

In some of the 'what's going on here' censuses, I discovered that the 'extra' person actually belonged to a nearby household.

In my opinion -- unlike the other genealogy groups you refer to in the comments, it's not our job to solve the specific problem that a person asks in the question. It's our mission to show how to solve that kind of problem, using the example in the question as a case study. We need to think like a teacher and a genealogy presenter as well as being problem-solvers.

So in this case, if I can get the time to find the examples from my own research, I could use as examples in my answer the case where an aunt was mistakenly put with the wrong family because the enumerator wrote the people down out of order, or the nephew who belonged with the family on the next page. I could link to examples in the answer which met our site guidelines of not talking about people known to be living.

The issue for the site as a whole is -- how much will my answer be improved by showing links to the specific examples, instead of just telling you about them? Will it be more help if I snipped out the examples, and included the images? Or would a "big picture" approach be of more use to answer this question?

Yes, we're all human and curious, so we like to see what people are looknig at. But if you aren't familiar enough with a record set to answer a question without looking at the specific image mentioned in the question, it's okay to pass on a question and not answer it.

See How to decode / read US Veterans Bureau Form 7202? as an example of a question where the specific person's information was blurred out. I was able to identify the source and answer the question without seeing the specific person's name on the card.

  • 2
    So many of our questions are exactly about helping someone with a specific problem. Not just pointing them to resources on tax records in Kentucky. We already don't get a lot of questions, if every question on the same topic were closed as a duplicate, because we were ignoring specifics, we'd be down to almost none (especially for DNA). If I can look at the actual page, I might be able to figure out in seconds what the deal is, instead of spending 10 minutes outlining all the possibilities. We can decide which way we're going to on on this SE, but it should be consistent. Feb 2 '19 at 20:41
  • Yes, but how much would your quick answer, given after looking at the document for several seconds, help other people who visit the site? We're here to teach people how to fish, not to feed them.
    – Jan Murphy Mod
    Feb 2 '19 at 21:14
  • Then why do we allow 500 variations of "DNA tests show a close relative, what could this person be to me?" Feb 2 '19 at 21:17
  • Maybe we shouldn't.
    – Jan Murphy Mod
    Feb 2 '19 at 21:25
  • 1
    @- to be accurate the original question asked THREE questions, not just the one cited above.
    – BobE
    Feb 3 '19 at 3:41
  • 1
    @Cyn "Then why do we allow 500 variations of "DNA tests show a close relative, what could this person be to me?" That's why you and I are working on the canonical DNA Q&AS! So we can point all these 'same questions' to one answer.
    – user104
    Feb 3 '19 at 7:23
  • 1
    @ColeValleyGirl Is it? My understanding was that the purpose was to have it so we didn't have to repeat ourselves with the basics, but not so we would close all questions that referenced it as dups. Feb 3 '19 at 16:45
  • @Cyn, why would we not close them as dupes if they are dupes?
    – user104
    Feb 3 '19 at 17:03
  • 1
    Jan, can you clarify -- do you think links should be disallowed on privacy grounds -- it's a different question as to whether they're useful.
    – user104
    Feb 3 '19 at 18:01
  • @ColeValleyGirl I've added that to the top of the answer.
    – Jan Murphy Mod
    Feb 4 '19 at 3:18
  • @ColeValleyGirl I've also added a link to the NGS Guidelines for Sharing Information with Others. If I can find comparable documents from the Society of Genealogists etc. I'll add them.
    – Jan Murphy Mod
    Feb 8 '19 at 8:43
  • Good document, Jan -- not sure the SoG has anything similar.
    – user104
    Feb 8 '19 at 8:59

Yes, such links should be allowed, provided that the placement serves the purpose of providing useful information to ask or answer questions that are appropriate for the Genealogy & Family History site.


I think including a link to public documents that contain potentially living people (those born less than 100 years ago with no evidence of decease linked to) may be acceptable under these circumstances:

  • that no part of that image itself is included within the post unless ALL potentially living persons and their details are redacted (blacked out).
  • you do not refer to any potentially living person in that image by name. Instead say that the linked image may be consulted for additional background to the question.
  • Moderator discretion will be used to remove and redact links if their use in a particular question is assessed to risk violating the privacy of any potentially living individual. Redactions always require the approval of a second moderator which guarantees that this can never be done unilaterally.

I asked SE staff and moderators from all sites to review our Q&A here, and the only comment so far, from an SE staff member (not lawyer), has been:

I don't feel like I have a particularly useful opinion on this; y'all should settle on whatever policy avoids trouble for your community; if it looks like something is, can be, or may be used for nefarious purposes, shut it down.

My feeling is that we should continue to seek consensus here on our Meta and in the meantime work to our current privacy policy. Since that does not specifically include or exclude links to public documents that contain potentially living people I suggest we use them cautiously. If I see any that concern me I'll redact them, and redaction always requires approval by a second moderator.

It should also be noted that we already condone/encourage posting of links that may contain living people's names within our Help Center where we say:

Acceptable evidence of death needs to come from some official or semi-official source such as an entry on a death-registration / grave / cemetery / obituary site. A website link (URL) should be provided wherever possible.

  • 2
    Pointing to people by reference instead of by name doesn't help you avoid data protection law in Europe at least - it's still personally identifying information regardless of how it's expressed. Feb 5 '19 at 13:34
  • @GaneshSittampalam Does the GDPR apply only to data provided to a data custodian for a particular purpose, or does it apply to any data found anywhere on the net -- i.e. if it's out there in legally published historic documents? I can't work it out...
    – user104
    Feb 6 '19 at 9:21
  • @ColeValleyGirl It applies to personal data anywhere - even if it's already on the net, if you do something with it you'd be processing it and need to have some justification. It would often be easy to find that justification though. Feb 6 '19 at 9:47
  • @GaneshSittampalam Even if you're just linking to a document elsewhere? (I agree, copying images is processing). And the consent was given 70 years ago or more before the Internet existed? Or the documents are legally allowed to be open to the public?
    – user104
    Feb 6 '19 at 10:07
  • @ColeValleyGirl I'm not an expert, but I'd guess linking on its own doesn't count. But I think linking + saying "the four year old in the second family in the image" counts just as much as linking + saying "Bob Smith is in the image". Feb 6 '19 at 10:18
  • Maybe we should insiste that proof of death is printed as a redacted image instead?
    – user104
    Feb 9 '19 at 7:36
  • @ColeValleyGirl I'm not sure that anyone has ever gone to the trouble of providing a link as proof of decease so I think the chances of training our user base to redact images accurately is small.
    – PolyGeo Mod
    Feb 9 '19 at 7:55
  • @PolyGeo So be it -- we just remove the names of anyone born less than 100 years ago if there's no acceptable proof offered.
    – user104
    Feb 9 '19 at 9:25
  • @ColeValleyGirl I can always live with redacted images but I don't think that's practical for more than a small part of our user base. I think links are much easier even if rarely used. That's how they came to be mentioned in the Help Center.
    – PolyGeo Mod
    Feb 9 '19 at 10:11

I no longer believe this is the right solution -- I'm leaving it here for the common trail, but support Jan Murphy's answer instead.


Good question.

The "100-year rule" (as per the discussion at Should we reduce the 100 year privacy rule to something a bit more flexible?) is based on the privacy legislation for most of the world (the US is an anomaly in using 70-odd years) -- it's simpler to have a single rule rather than try to work out what legislation applies in each circumstance, so 100 years errs on the side of caution as well as consistency.

However, different countries have different rules for 'publically available info'. In England and Wales anyone can order a BMD certificate if they know sufficient basic information about a event up to 50 years ago (year/quarter, name, district -- and all those can be accessed via search engines such as FreeBMD. (Full information is needed for events less than 50 years ago). And because the certificate is 'Crown Copyright' they can legally publish it wherever they wish. However, the 1921 census is under 100-year embargo (the 1939 register is only available because of a legal loophole that doesn't apply to the census, but is only published with the names of the living and those who aren't proven dead redacted).

By contrast, in the States, it can be impossible for anyone other than a relative to get hold of a certificate, but the 1940 census is available online.

Should records for a living person (even the ones visible elsewhere online) be published here? I would prefer not -- we're searchable via Google etc. (while the records at the genealogical providers are not) so anything placed here is more widely visible. Perhaps a compromise would be to allow record references to be published so that people with access to the 'walled sites' can go and look if they wish -- it isn't as convenient as having (part of) an image inline but it may be the best we can do other than banning them altogether. Again, consistency will be important -- we don't stand a chance of enforcing a rule that varies by country especially when we're a worldwide site.

To clarify: detailed links to 'walled gardens' (site that can't be searched via Google etc) that don't name living individuals might be acceptable, but images or details posted in clear text should not be allowed -- if only to keep us in line with SE Privacy Policies (and -- perhaps -- the GDPR).

  • What if we allowed a link to the document (on a site that lets you look at adjoining pages) and/or a screenshot (preferably both) but allowed the text to give only first names? This would allow answerers to give informed answers but also make it impossible for someone Googling the name to find the discussion? Jan 30 '19 at 15:00
  • @Cyn, I think it would be easier to moderate links than images. I think allowing the text to give only first names will make the question less explicit so wouldn't favour that -- however, if somebody was born less than 100 years ago and not proven to be dead they shouldn't be named at all so the completeness of the name is a moot point. (Using first names for living people when their surname is readily identifiable from other details in the question is IMO a no-no).
    – user104
    Jan 30 '19 at 15:28
  • Well the entire point is for those of us who are answering the question to see the document the OP is talking about. We can't do that if there are 20 families on a page and we don't know which one it is. Or which child is the "extra" child. If the census page is for an apt building (flats), the addresses would be identical. Jan 30 '19 at 15:34
  • @Cyn, I don't think there's an ideal solution. I suggest however that if it's a census record identifying the household (by name or line number or address or whatever is most appropriate for the document in question) as part of the reference would be acceptable... it still wouldn't identify living individuals to anyone using Google who didn't also have access to the document. We'd be skating close to 'identifying possible living people' (name, age, possibly relatives, address in 1940...) but none of that would be available to casual googlers.
    – user104
    Jan 30 '19 at 15:37
  • I will say that every other genealogy group I'm part of online allows discussion of public documents regardless of if the people are living or not. Those groups, like here, don't allow discussion about contacting people right now or finding people who might not want to be found. Honestly, I've never found any group that wouldn't allow a discussion of a 75 year old Census record or other records that we all routinely use in our genealogy (passenger records, obituaries, death records, etc) that may mention living people. I bet you anything that there are already tons of posts here that do. Jan 30 '19 at 15:42
  • 1
    @Cyn, You might find discussions about such documents here, but we've been pretty strict about policing the 100 year rule/not identifying potentially living people in the text of Qs and As on this site or associated images. Just because other groups do it doesn't mean it's acceptable here. You can discuss the documents and how to work with them, you just can't name the people.
    – user104
    Jan 30 '19 at 15:45
  • 1
    @Cyn It may be material that we're subject to SE's Privacy Policy (in their terms of service): see genealogy.meta.stackexchange.com/questions/1874/…. Other sites you participate in may not be subject to the same type of constraints.
    – user104
    Jan 30 '19 at 15:51
  • @ColeValleyGirl - IANAL, and can't say that I've digested the SE's terms of Service, however it does occur to me that a Q&A that contains "Donald J. Trump, 72 years old, currently living at 1600 Pennsylvania Ave, Washington DC, current occupation is President of The United States" would be forbidden on the Genealogy site. Seems a tad ridiculous
    – BobE
    Feb 3 '19 at 4:05
  • 1
    @BobE -- forbidden for a number of reasons -- identifiable living locatable person who is a celebrity being a few of them. Doesn't seem ridiculous to me at all.
    – user104
    Feb 3 '19 at 7:21
  • @ColeValleyGirl - would the same statement be forbidden on Politics SE or Skeptics SE? (keep in mind I'm suggesting that the statement might be made in an appropriate context of a larger question). My point is that the identification of living persons is fairly common place on other SE pages and I've never seen the SE TOS being invoked in those circumstances.
    – BobE
    Feb 3 '19 at 16:02
  • @BobE, if Trump was mentioned as part of a question about Politics, that's fair game -- he's chosen to be a public figure and his politics are on-topic over there. But being a public figure doesn't put his Family History up for public grabs any more than (say) yours is. We have one rule for living people. Full stop. We also forbid 'celebrity genealogy' about dead people (see genealogy.meta.stackexchange.com/questions/3403/…) for various reasons.
    – user104
    Feb 3 '19 at 16:15
  • So a mention of his father and that they both have a history of questionable leasing discrimination, would be disallowed because his father's name is mentioned, therefore making it a matter his families history. My point is - if the GFH page chooses to constrict itself by being restrictive, well that's fine, but don't shift the responsibility of that restriction to the SE's TOS. And, as far as my family history is concerned, I personally have no issue with having the same information on these pages that can be found by googling my name.
    – BobE
    Feb 3 '19 at 17:41
  • 1
    @BobE, Also what matters more is protecting the privacy of people who might not even know this site exists, let alone realise they might be discussed on it.
    – user104
    Feb 3 '19 at 17:58
  • 1
    @BobE but not names.... which is 'personally identifying' and kind of key around here.
    – user104
    Feb 4 '19 at 15:12
  • 1
    @BoBE I'd be dubious about anything that associates a phone number with any part of a name if it could lead to the identification of the individual.
    – user104
    Feb 4 '19 at 17:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .