1939 register - closing an open identity

[Nick_Ips]

I would also add a further note of caution.
If too many people complain about records of living people are being released the government will do as they did with the 1920 Census Act.

Exactly the point I was making above. Government departments rarely need an excuse to make data secret, people complaining loudly about living people's data being published in error just makes it easier to justify all the data being kept hidden. Or else adopting the kind of approach taken with the 1939 Register data for Scotland...

[Liz_in_Sussex]

anyone had success at closing an open identity without said signed document?

I hope you have been successful in getting the record redacted.

A few weeks ago my mother (who is in her 70's) was talking about her childhood neighbours so I went to the 1939 Register to check some names for her.  I was very surprised to see her name open for all to see - she was only a baby at the time and her two older sisters were still closed. 

At my mother's request (and I do have POA) I emailed FindMyPast asking them to cover her record as they would be unable to prove to me that she had died. 

We were ready to send them a letter from the family solicitor, depending on the reply to the email, but were pleased to receive an almost immediate reply from FindMyPast apologising for any distress this had caused my mother and a promise to cover the record - which was duly done.

Perhaps when she is in her 90's she will feel a bit more cheerful about seeing herself openly online!

Liz

[Nick_Ips]

Liz, it is good to hear that FindMyPast were so responsive to your request.

But just to illustrate some of the points made earlier, what you've been sucessful at doing is concealing a small piece of information which was only available to people who have registered (and probably paid) to view - and is one name in a list of millions. The chances of a scammer finding it are slim. FindMyPast probably log details (IP addresses or something) of people viewing each image, so if some kind of fraud resulted there may be a trail which would allow detection.

However, what you've posted above is potentially far more valuable to those with ill-intent than the information incorrectly opened on the 1939 Register. Happy to explain by PM if you'd like.
[I've edited this paragraph because I wasn't comfortable with the original version]

I don't mean the above comment as criticism of you, and I don't mean to alarm you, but it does illustrate the point that hiding a 1939 Register entry is not that effective unless you also ensure other useful bits of information (the kind often posted on FB) are kept private.

If on reflection you decide to edit/delete your post then I'm happy to do the same with this one

[Trishanne]

I have given up on getting the transcript of my family details corrected. All details are transcribed one line down, so our household name has been given as that of the man who was the last entry of the house above.
 Our surname has been transcribed incorrectly too, but that does look confusing on the original record.
 My mother is not on the transcript, she is on the original,
Because all details are one line down, my brother who was born in November 1931, has been given my sisters details so his birth date is shown as September 1933. My sister has my details, so instead of September 1933, she has my birthdate April 1936.
I am still alive, I'm positive, but I am there. I was under school age on the original, but I have been given the details of an Aunt who was living with us. On the transcript I  have been given her birthdate and I am down as a Domestic Science teacher.
We had a guesthouse at the time, so there are about 6 more people with wrong details, if relatives have looked on the records for these people, I don't think they will have had much success.
There is not a lot of room on the transcript to make alterations, I tried when I first saw it, so now my fathers details are correct, but he is still the second line down, the name of the gentleman from next door still is shown as head of the household.
All other details are still in correct. I have tried again just this week, to make alterations, so I wait  to see what happens this time.

[katerimmer]

I am very interested to see this thread, as I reported this problem to both Findmypast and The National Archives not long after their update of the 1939 Register in November 2016, when they opened 2 million more records.

  As I understand it, when the Register was originally launched online, records for people born less than 100 years ago were closed unless they were marked deceased on the Register itself.

 Some time after that, FindMyPast ran software which checked the closed records against the pre-1984 GRO death indexes, and the 1984-2007 GRO death indexes, which they have copies of on their site, using certain rules agreed with TNA to determine whether a match was unique or not, and if it was, the record could be opened. (They also open records regularly where birth dates become more than 100 years old.)

 Then they used an outside agency to carry out a similar matching process against the post-2007 GRO death indexes, which FindMyPast do not have a copy of but which is supplied to certain companies.  The 2 million records opened in November 2016 were the result of this last matching process, and as far as I know, FindMyPast staff have no way to check whether there is actually a death index record for people whose records were opened then, because they do not have access to the post-2007 death indexes - unless they visit one of the main libraries which hold them on microfiche.

  I don't think there was a significant problem with the earlier updates based on matching against the pre-2007 indexes, only the post-2007 one. Both FindMyPast and TNA have denied that there was any problem with this update - see the most recent comments on TNA's 1939 Register blog for the discussion which I and a few other people had with them, until they closed it to further posts:
http://blog.nationalarchives.gov.uk/blog/1939-register-census-census/#comments

Most of the points I would make are in my comments on there, so I won't repeat it all on here as it would take ages and make this post much too long!  But just to add that I have made a complaint to the Information Commissioners' Office and am waiting to hear the outcome. 

Oh, can I also add that if you contact FindMyPast support to have a living person's record closed, they may not know how to do it properly and they may either redact it on the image or the index but leave one or other of those open, so it is best to go via clicking the button to have the record closed, but if you can't supply the ID proof which it still asks for, upload any image just so it will go through (e.g. a screenshot of the 1939 Register record or whatever) and explain in the text box how you know the person is alive.  They have said that they don't require the ID proof any more, but for some reason they have not removed it from the webpage, so it doesn't go through unless you upload some kind of image.
 

[Nick_Ips]

Kate, I've just read your comments on the TNA blog with interest. Have you read the whole of this Rootschat topic? If not could I politely ask you to have a look at my reply #64, and some of the other comments people have made about:

1) The actual nature of the data that may have been incorrectly opened and the other sources for the same information (such as dates of birth) which are readily available elsewhere. This is not 'secret' information, just information people might prefer is not is the public domain, but in reality they have no control over.

E.g. Michael Heseltine's DoB is available from a simple Google search - his 1939 record being visible would be of no consequence. And I doubt Mr Heseltine is going to take legal action against wikipedia.

2) The more people complain about this, the more likely it is the next time we want a dataset containing some living people opened then the response from the government will be a no, justified on the basis people were terribly concerned about the 'small' number of 1939 records incorrectly opened.


If someone wants to find out your date of birth they will. It is not hard to do. If your bank still uses your date of birth as a 'security question' then you need to change bank.

Someone planning to use people's information for some fraudulent activity are unlikely to trawl through the 1939 register looking for the incorrectly opened details of a living person in order to get their date of birth. There are far more efficient ways of doing so.

In the nicest possible way, I do hope the Information Commisioner investigates your complaint but concludes that the approach adopted by TNA/FindMyPast is completely reasonable in the circumstances. Reading the responses from the TNA it seems clear to me that they have carefully balanced the issues of protecting living people against providing access to historical records.

If we were only able to see the 1939 records of people who could be proved to be dead (and that means the over 100's too) then it would be a much diminished resource.

Edited to add:
Also, I suspect your suggestion that the 1939 Register search should be protected by a 'reCAPTCHA' or similar anti-bot device would be deeply unpopular with the many thousands of people using the site for family history research. I also fail to understand what you think anyone might achieve by doing bot searches of the data, as the record has to be unlocked before you get any meaningful information.

[katerimmer]

Yes, I have already read the whole of this thread.

1) I cannot stress enough that I don't care whether Michael Heseltine's  (by the way, he is a Lord nowadays, not a Mr!) privacy has been invaded, but the only easy way to get a big enough sample of verified living people to check was to choose famous people.  Dates of birth for ordinary people are not so easy to get, though I realise that you can get someone's birth certificate - but you have to pay for that.

2) The first thing I did when I realised that records had been wrongly opened was to contact FindMyPast and advise them that they should immediately roll back the update until they had figured out what went wrong. After some discussion with them and getting nowhere, I contacted TNA and it was only when I finally realised that neither organisation was ever going to admit that there was a problem that I contacted the ICO (which, by the way, TNA advised me to do.)  I would have hoped that the 1939 Register would set the standard for future databases which contain closed records of living people which can be automatically opened after their deaths, but since there is such an obvious error in processing of the 1939 Register, it needs to be sorted out now before it can be used as a basis for any future projects.

It is not true that a record has to be unlocked before you get any meaningful information, as data such as date of birth can indeed be obtained without even logging into FindMyPast.  I would not like to go into details of how I think people might use a database made up of data copied from the 1939 Register.

[vrvt]

FindMyPast sent me a kind response and closed the record of my living relative even though I did not have any documentation. Obviously the system is not perfect, but they do a good job with the funds and capabilities they have. In an ideal world, they would have someone researching every individual to determine whether they are living or not, but the cost of that would have to be passed on to the customer, and I suspect would make the venture not worth the risk.

Just as a point of reference, the 1940 US census was released in full years ago, which contains details of living people.

[Nick_Ips]

1) Lord or Mr, it makes little difference. Our information is out there whoever we are. For some it is more readily available, but nobody should think that a determined fraudster is not going to obtain a date of birth with relative ease. A birth certificate does cost money, but hey, there are expenses in every business. Also bear in mind the 1939 Register birthdates are not great when it comes to accuracy. If I was planning a fraud (which I'm not  ) then I'd rather pay £9.25 to get the definitive birthdate instead of risking the cheap option and a couple of years at her majesty's pleasure.

However if news reports are correct, personal information about people can be obtained online for a matter of pence per item - that is where the business is, not trawling through the 1939 Register.

2) You say there is an "obvious error in processing of the 1939 Register", but is there? The response from the TNA says

The risk for the 1939 Register has been adjudged by the Government Actuary Department, based on the likelihood of someone living in 1939 continuing to be alive at age 100, to be 1.5%. Following discussion and agreement with the Information Commissioner’s Office regarding this dataset it was decided that this was a tolerable level of risk.

To me this reads that some very clever people from different organisations have sat down and looked at the problem and agreed that whilst there is a risk of living people's data being published, the level of risk is acceptable (given that any living person can request take-down as well).

The only way of absolutely ensuring that no living person's data is published is to simply not make any of the 1939 Register available until every person known to be born before the Register was taken has died. Is this your preferred option?



I'm actually more concerned that the specified level of accuracy in matching the death records means there are many records for people who have died that will remain closed with no apparent proposals to address this moving forward - until they hit the magic 100.

Ok, yes, it is possible to obtain an idea* of the date of birth without unlocking a record. But if you remember, there were certain adjustments made to the site in the early days because FindMyPast had detected unusual activity. I think it safe to assume that although FindMyPast may not be monitoring the site use to protect our privacy, there will be sufficient monitoring to detect large-scale 'probing' of the data.   *But see the above comments about accuracy of this DoB

It comes back to the 'why' - why would a fraudster go to the trouble of developing a bot to obtain free of charge dates of birth of people who might, or might not, be living? Much less risk for a fraudster to identify a 'target' person, go online, pay £9.25 and get a nice accurate birth certificate.