This blog post is my summary and analysis of a new report on young people and family court judgments by Julia Brophy, funded by the Association of Lawyers for Children

Since 3 February 2014, judicial guidance has required certain judges, hearing certain family cases (including care proceedings), to ordinarily send their judgments to be published (usually anonymously with respect to children and families) at (a small, charitable, public legal information website).

On 31 October 2014, the President of the Family Division’s retrospective consultation on the effects of this guidance (and consultation on possible further transparency reform steps) closed. Further announcements are still awaited.

On 12 October 2015, the Association of Lawyers for Children announced that it had been funding research over the past year as follows:

  • A (complete) pilot investigation by young people into how easy it is to ‘jigsaw’ identify children from judgments, with the benefit of local knowledge and also when using the internet. See Family Law website for the published report: A Review of Anonymous Judgments on Bailii: Children, Privacy and Jigsaw Identification, Brophy, Perry and Harrison, published by the ALC and NYAS in October 2015 (though the ALC website says it was sent to the President in August 2015);
  • A (virtually complete) review of existing avenues for accountability in public law children cases and models for inspection and accountability in other public services such as health and adult social care. This report will go initially to the President by late November or early December 2015. A multidisciplinary seminar on both reports is anticipated.

Dr Brophy also makes suggestions for future research in the report (para 14.10), including further analysis of the risks of jigsaw identification; a trial of short summaries on Bailii while full judgments lie on file, so that details about abuse and parenting are largely redacted; a review of anonymisation of the official Law Reports; a study of better anonymisation practices, including judicial guidance / anonymisation protocols and dedicated teams (with resource and training implications); and whether useful parallels can be drawn with the new s.33 Criminal Justice and Courts Act 2015 offence of disclosing private sexual photographs with intent to cause distress.

The jigsaw identification research itself

Nine young people (with local knowledge where possible) analysed 21 judgments published on Bailii for ‘identifying’ information and then cross-referenced information found in those judgments for social media and media reports. They were asked:

1. Whether judgments themselves contained specific information which would permit children to be identified by someone from their peer group or a member of their local community who was reading the judgment ?

“13 of 21 judgments contained specific information which would permit children to be identified” by peers or those local in the community” (paras 12.1-12.2).

Information that can narrow location
They (the young people ) first identified types of information that carried a high risk of narrowing down the actual location of a child beyond the wider local authority and/or court area and details of some professionals that are routinely provided. They included any “area information including naming a town; information about a school or school issues; gender and age of children; information about extended family members; and information about religious/cultural customs within households’ as ‘high risk in-borough geographical markers”. They concluded that 6 of 21 judgments had at least 4 of 5 of these markers and that this “placed children at high risk of being identified by peers at school and in local communities”.

Other information
They also audited judgments for details of abuse of children and parenting failures and difficulties. Considering these with the above locality high risk indicators above, they reached the conclusion above on 13 judgments.

My thoughts about this
The actual risk posed by individual pieces of narrowing information recorded for the research as a ‘high risk geographical marker’ seems to vary considerably. For example, in one of 21 judgments. the town the family currently lived in was not redacted, which is clearly a serious anonymisation error. In another judgment, two towns to which the family had moved due to domestic abuse were mentioned, but this didn’t seem to involve a family’s current address so arguably created little risk of identification even with other markers. In one judgment, “information about a specific school attended” was included. But there’s no suggestion the school was named or that it was the current school. It would be helpful to refer to the actual judgments and see the completed semi-structured schedule, rather than just the empty question template at Appendix 3. This would give a fuller picture of the reality of the claim that 6 of 21 judgments contain 4 of 5 ‘high risk in-county geographical markers’ in terms of objective risk of identification. (Safely referring to the judgments might mean some further redacting). Unfortunately the research does not identify or make it possible to identify the judgments analysed.

There is no suggestion that the young people in the exercise were able to identify children from the information in the judgments alone, even with their local knowledge. Nevertheless, any indication that young people see a likelihood of a child being recognised by their peers from the information in the judgment (if that peer were to ever read that judgment) is a concern to be addressed.

What the study also suggests to me is that it is extremely unlikely, as things stand now, almost two years on from implementation of the new guidance, that ‘ordinary’ young people (peers) know that judgments exist; or are anonymously published from care proceedings; or that Bailii exists; or how to navigate its confusing array of databases; or have the motivation to find out. Even these young people – who are representative of those with direct experience of care proceedings and who have extensive knowledge of how information is shared from proceedings arising from their involvement in earlier studies – plainly had no idea that judgments were published in this way. (In particular they “had little/no idea of the content of judgments on Bailii [and felt shocked and let down by] the “difficult, deeply embarrassing, shaming and damaging information about children’s lives”).

2. The young people also asked whether someone with local knowledge and/or intent to identify a child could do so from information in the judgment combined with social media and media reports online.

They found that searching specifically on the information from judgments enabled them to find coverage in on-line local and mainstream newspaper sites and on social networking sites and to identify children. (Paras 13.1 & 13.6)

“Materials on social networking sites (e.g. Facebook pages etc.) identified children and other family members and some also contained photographs of children” (para 13.6)

Young people found mainstream media reports (including online family law content) for 24% of the judgments and social media site coverage for 33% of them.

My thoughts about this
The report doesn’t specify how many children were actually identified but a look at the more detailed information (para 13.4) suggests it may have been only one child, who was named in several places on a grandparent’s facebook page. Again it would help to have the fuller information.

Nevertheless this suggests that significant numbers of children are at risk of being identified by anyone likely to find and read the judgment and then go on to search the internet, based on that information, if families continue to post details of their children on social media sites without effective action to prevent this, and detailed judgments continue to be published.

Obvious contenders are journalists, legal bloggers and, much more seriously, those with a vested interest in finding particular children, such as organised paedophiles and perpetrators of violence who are seeking families who have moved to confidential addresses.

This potential for linking what is firmly intended to remain an anonymous judgment, in all its graphic detail, with an actual child, identified by name, poses a significant risk to children’s privacy (and a serious risk to safety in some cases).

Similar risks arise from any press reports that identify the child directly or indirectly, especially where the adults are identified in criminal proceedings.

Identification of the child on social media sites can sometimes be perfectly lawful at the time it is done, since the statutory prohibition on publishing information that may lead to identification of children subject to proceedings ends when the proceedings end. (Though other prohibitions may apply, including a standard ‘rubric’ restriction when a judgment is published, that ‘in any published version of [it] the anonymity of the children and…family must be strictly preserved’ (para 21 Practice Guidance). Alternatively, putting information on social media may be done with disregard for the law or in ignorance of it. The law is complex, and may involve relatives like grandparents who often can’t access legal advice even in care proceedings, and who may be aggrieved at a perceived or real injustice, particularly where they have limited and perhaps one sided information about what has gone on in proceedings.

This is an issue that has been talked about for some time. It has also been written about on the TP website here, here and here. Often, when a wildly incorrect press report has surfaced, this generates discussion amongst TP members about whether to link the judgment to it (to counter the misinformation) and risk offending the standard rubric or some unknown further reporting restriction or more importantly identifying the child. A major difficulty is that there is no reliable publicly available record of whether a reporting restriction is in place or what it says. (The recommendations of the Law Commission for managing this in relation to Contempt of Court Act S4(2) Postponement Orders offer an interesting parallel. See also commentary here).

So far, we have seen instances of responsible media representatives (and others) proactively informing the courts where they have accidentally been able to identify a child (though little transparency about any systemic learning from each incident). I have yet to see any instance of irresponsible linking by the media to the detail within judgments or gratuitous naming of children, though it may have occurred. If anything, we see the opposite – that the media rarely, if ever, link to any published judgment, sometimes with significant consequences for public misinformation. We do however see instances of the media publishing the details of family members, in cases where judgments have been published subject to the standard ‘rubric’ (reporting restriction) that make it a contempt to identify family members ‘in any published version of’ the judgment’. (See here).

Redacting the details of abuse and parenting by reducing parts of the judgment to a short summary, as Dr Brophy suggests, is one possible solution to this risk of identification via the internet.

For me, the research reminds us just how crucial it is that the right orders / rubrics are made to accompany judicial decisions to permit publication under the guidance, in anticipation of s97 ordinarily ending on conclusion of proceedings; that attention is paid to service and explanation of those orders, particularly in relation to self representing litigant family-members; and of the need for a systematic mechanism for reporting and enforcing breaches or errors (as well as removing judgments temporarily where required).

Still no coherent systems for safe anonymisation?

In my view, the report is right to express grave concerns that, nearly two years on from implementation of the guidance, we appear to be no nearer to a transparent, systematic process for minimising unnecessary risks to children by anonymisation errors or jigsaw identification.

It seems to be the case that the judiciary have some internal guidance on anonymisation. This hasn’t been made available to other practitioners in the family justice system but in any event doesn’t seem to include any formal or standardised process for anonymisation. Without a protocol of some kind for safe anonymisation, children’s practitioners have no bottom line standard they can expect of the judiciary or local authority solicitor, from which to ensure judgments are adequately checked on behalf of children and young people before they are sent for publication; or from which to talk confidently to young people about publication of judgements, safeguards and any applications warranted in the particular case.

I’m not aware of any statutory guidance, case-law*, professional practice guidance or protocol, let alone training for social workers, local authority lawyers, guardians or lawyers for children on:

  • What must routinely be redacted (and then carefully double-checked): names of family members; towns; and school names. I agree that dates of birth should also be removed.
  • What may need to be redacted to prevent unwarranted risk of ‘jigsaw identification’ and the types of details that can become identifying in cumulative contexts even if not in isolation (such as details of disability, ethnicity, religion or culture, details of large siblings group ages or highly specific details of abuse or physical consequences like scarring):
  • Particular scenarios warranting higher levels of consideration, including children specifically at risk of being located by violent perpetrators or organised paedophiles; those where parental abuse has lead to, or is likely to lead to, reports of criminal proceedings; and children evidently already at high risk of identification on social media.

There is room for substantial disagreement on ‘best practice’ here but surely it’s not impossible to draw up an agreed minimum ‘something’ in preference to a defeatist ‘nothing’ for nearly two years, that leaves children at risk?

The Anonymisation of Judgments Quick Reference Table from the Australian jurisdiction seems a useful example of a systematic practitioner check-list that could be adapted for this jurisdiction. There seems much to learn also from their use of a centralised Judgments Publication Office operating on standardised guidelines. (Written about at TP here). It’s not impossible that the costs of a small, well-run, specialist unit could be off-set by wider savings including to judicial time. Links to Dr Brophy’s 2009 study of international comparators and some other relevant international research are here.

* Wigan Borough Council v Fisher & Ors v Fisher & Thomas (Rev 1) [2015] EWFC 34 (21 April 2015) for example (written about here on TP website) doesn’t guide on basic standards of anonymisation or redaction to avoid jigsaw identification, as opposed to considering the right balance in the very specific case circumstances that applied.

So few reported applications on behalf of children

Perhaps also surprising is that (nearly two years on) we have so few reported cases about altering the new starting presumption imposed by the guidance, to redact details of abuse and parenting failures (or oppose permission to publish the judgment at all, for particular children and young people). Such applications might be brought on the basis of evidence of particular risk, particular vulnerability or particular individually held wishes and feelings on the principle of privacy, notwithstanding anonymisation.

If the young people cited in this research are remotely representative of children and young people in proceedings, this may indicate that children’s practitioners are simply not routinely advising young people or taking their instructions. In fact the report (page 29) suggests anecdotally that this does seem to be the case.

Those who read the earlier report on behalf of young people by Dr Brophy in 2014 will recall young people expressed worry (page ii executive summary) that they hadn’t been told of the small chance that the press might attend their hearings (with how it would be handled if it arose). It may be that this problem for young people has simply extended into not being told about publication of judgments on Bailii either. This wouldn’t be surprising, given the complexity of the law; concerns about talking to young people on this subject; pressures of time, funds and court listing; and the dearth of guidance or training. Research on this seems important.

In the meantime, applications on behalf of certain children and young people to derogate from the starting presumption of anonymised publication and redact, not just identifying features, but perhaps certain details of abuse and parenting where warranted, or even to oppose permission to publish at all, may be exactly what is required to trigger more of a road-map for children in this arena. (Parallels might be drawn with the recent case of Louise Tickle v North Tyneside Council, written about on TP website: here, in the similarly uncharted waters of applications by the press for access to otherwise confidential documents from proceedings.)

Next Steps?

I agree with young people, researchers and others, including responsible members of the press, that adequate systems are not yet in place to adequately manage risk of anonymisation errors and jigsaw identification; and that this is unacceptable for children and young people.

It’s also a huge barrier to persuading any (let alone many) practitioners that ‘next steps’ are practically achievable, whatever the merits in terms of the right starting balance between children’s privacy and open justice.

While this report rightly focuses on the essential requirements of children’s privacy and safety, a wider research strategy is needed to encourage dialogue on improving reporting accuracy as well as public awareness and confidence in family justice.