We’ve been having some discussions behind the scenes about jigsaw identification here at TP recently. Not very transparent I grant you, but the problem is that our discussions have centred upon some judgments that one commenter was anxious might have given away a little too much specific information  to the point that within the family’s local community the parents and children were likely to be readily identifiable. So what did we do about that comment and the points it raised?

Well, firstly, we asked the commenter to write a post about Jigsaw ID here. And secondly, we contacted the Judicial Office to pass a message to the judge as follows :

We have received two comments on the Transparency Project website raising concerns about the possibility of inadvertent jigsaw identification of children involved in care proceedings through two judgments of HHJ XXX, published recently on Bailli. We are anxious not to publish the comments and thereby increase any risk of jigsaw identification further. We wondered if you are able to relay those comments to the judge concerned for her information and if appropriate they can be amended. If we are being overanxious (obviously we don’t know the circumstances of the case) then perhaps you will let us know it is ok to go ahead and publish. Any feedback gratefully received. I’ve pasted the comments below for ease of reference.
Alternatively if you can point us in the direction of a more appropriate person to contact about this that would be great. I am reluctant to approach the Judge directly, and in any event I don’t have her email address.
We had to wait a little while for a reply due to judicial leave and the fire in Holborn knocking out various computers / offices, but we’ve now got our response. The judge, we are told,
…was grateful for the feedback and the judgment with the children’s names in it has been amended. The judgment with the XX father will be removed and Judge XX will see if it can be further anonymised.
So, there we are. But this does leave us with two problems :
Firstly, how do we transparently deal with the original comment, which raised really important points about the specifics of jigsaw ID, but which obviously contains fact specific information which itself might identify the family?
And secondly, what about the bigger picture?
Looking at the question of this specific instance first – below is an edited version of the original comment (which was sent to the Judicial Office in full) – as with the identifying details in the email extracts above we’ve removed just enough to stop this post being used by means of jigsaw ID :

This case [CASE CITATION] raises the question of jigsaw identification. I do not criticise the judgement. It is in my opinion a beautifully, considered child centred judgement. I googled the judge , [who] … just may not be aware of the local demographics. [CITY] is strange for a UK city , in that it is predominately white, traditionally there has been very little migration let alone immigration. It is changing but certainly at a slower rate than other areas of the UK.  This is what stood out for me:


The XX community in [CITY] is minute , in fact the judge notes that the father is isolated. Not only that but I am guessing that the child is mixed race given her mother is XX. I suspect that they will stand out like a sore thumb locally, and because it is a close knit community news travels!

I am the last person to derail the transparency debate, anyone who knows me on this and other forums would have worked out that I am rather keen on opening up of family courts. To me though this is what has to be grappled with. Do initials of children need to be randomly generated, should the father just have been identified as [ETHNICITY] or should his ethnicity been left out all together. I reiterate again I am not criticising the judge, no one person can be aware of everything. Any thoughts?

This person also subsequently raised a second judgment from the same judge where the children were inadvertently named (first name only).
Obviously this issue is far easier to debate if one can give specifics – it is rather disembodied in this anonymised comment. But one might surmise, for example, a case in rural cornwall, where a Somali family with eight children between the ages of two and seventeen, six of whom were boys – might be readily identifiable to large numbers of local children, parents and professionals.
We have wanted to publish the comment, and through this post have done so as far as we are able. We cannot, for obvious reasons, link to the judgment that remains or the one that has been removed – and so we cannot fully convey to you the discussions we have had about how this is a great case study for the dangers associated with the anonymisation of judgments.
As for the bigger picture – well, it is unsatisfactory. On this occasion, we have happened to have the issue drawn to our attention by a conscientious individual with some knowledge of the system but who does not work within it. We have acted in what we hoped was a responsible way by making enquiries and those have come good (albeit that the upshot is one less judgment available to view by the public). But the system can’t operate in this ad hoc way. And importantly, this is not an isolated incident. To my knowledge there have been numerous slips, errors and (in my view) inadequate pieces of anonymisation which leave children – and adults – vulnerable to unnecessary and inappropriate identification. Whilst it worked in this instance, most people would not happen to have the email address for the judicial office at their fingertips (or even know where to look for it) and in any event the judicial office could not (I imagine) cope with vast numbers of such indirect queries for judges – and no doubt some judges might not take kindly to such interference from random small charities!
Hand in hand with a system where judgments are published must go a system of proper checks in respect of anonymisation. These things cannot be left to chance or to the hard pressed judge or transcriber or local authority lawyer. At the moment the lines of responsibility are unclear – in my experience something of a collaborative effort is made to get a judgment in properly anonymised format – it is often a bit of a ramshackle and inefficient affair and is ripe for mistakes to be made – and there is often a lot of buck passing about who will do the grunt work and who will take time to check it (because nobody actually gets paid for it!). This is bad for two reasons – firstly because it exposes families to a risk of harm or unnecessary upset through identification that ought not to happen and secondly because it undermines the transparency agenda either because errors of implementation obscure the cogent arguments for greater transparency in principle or because the resource implications for hard pressed professionals and judges are a disincentive to publishing judgments that ought to be published and could be published.
What is to be done?