CAFCASS and Women’s Aid have collaborated on research about allegations of domestic abuse in child contact cases. We published a guest post by psychologist Sue Whitcombe on this here : Looking beyond the headlines: domestic abuse allegations in family proceedings.

We also thought it would be helpful to look further at what this research does and doesn’t cover. Before we do, it is fair to point out that our Chair, Lucy Reed, has in the past been critical of some of the work of Women’s Aid : broadly speaking as to the evidence base for aspects of their campaigns (See here and here and associated links for examples). That said, she also recently ran a workshop at the Women’s Aid conference on behalf of The Transparency Project, to help those working with survivors of domestic abuse to support their clients through the court process (including helping them to understand the forensic process and to come to terms with the fact that allegations are just that – allegations – until tested and proved). The Transparency Project is an educational charity and our main objective is the provision of balanced and accurate information. We don’t have an agenda beyond that, but we are not afraid to disagree with either side of the argument. We try to present things in a neutral way.

What is the Cafcass-Women’s Aid research about?

It ISN’T about rates of domestic abuse. It’s about the prevalence of ALLEGATIONS of domestic abuse and the responses to them. That is to say – if those allegations were true, do the system’s responses to those allegations look safe and appropriate?

It ISN’T a study about parental alienation. It may be (as Sue Whitcombe suggests) that some unproven allegations of domestic abuse are false or exaggerated to further an agenda of alienation, but that isn’t something the Cafcass study sets out to consider. It might be a topic ripe for some further research – but it isn’t something this study aims to tackle.

Domestic abuse or allegations of domestic abuse?

The data analysed is about how many allegations are MADE, not admitted nor proven. In most places the report is very careful to distinguish between allegations and abuse, though there are a few slips, and when the report is talking about the impact on children it talks as if a) abuse is established and b) abuse is the cause of any presenting distress / issue (as to which see Sue Whitcombe). It is fair to assume that a reasonable proportion of those allegations come from people who have actually experienced what they allege they experienced. Some may be wrong, exaggerated or false – but any argument about precisely how many would be sterile. Some of the people making allegations in this data set (and their children) needed protection. Some of the people accused in this data set (and their children) needed protection from false allegations. This study can’t tell us how well they were served, but we can extrapolate some points.

What does the data show?

About 2/3 of cases involve allegations of domestic abuse. That isn’t a new stat – it’s often said to be around that figure, so this confirms that trend. And, as we already know, dads are more likely to be the subject of allegations than mums.

The sample size is relatively small (216 cases, of which 40 were subjected to qualitative analysis) and the study is based on incomplete data, as it was drawn solely from Cafcass’ files, which are known not to be a complete record of everything (in particular orders are often not kept). For example, in about 1/3 of the cases covered, the final outcome (court order) was simply unknown. This has real potential to distort the stats. It is a shame that the study was not larger and more robust in its methodology, and whilst we appreciate CAFCASS are working on limited resources, we do think this is the sort of topic which deserves a rigorous treatment and the prioritisation of resource (see here an example of a larger study on a related topic for comparison). Sue Whitcombe has set out some of the limitations of this research in her post, so we’ll try not to repeat that.

However, some patterns emerge which seem likely to be replicated more widely :

  • There were 126 female alleged victims and 40 male alleged victims. (The report does not state whether the cases included same-sex couples.)
  • Where women made allegations, they made a higher proportion of allegations of coercive control type abuse (and sexual abuse) than men who made allegations – that is to say, almost all (84%) of allegations about women were of physical abuse, whereas only just over half of allegations against men were of physical abuse.
  • Where domestic abuse was alleged, at first hearing stage the court was most likely to make ‘no order’ about contact (42%), with unsupervised contact ordered in 23% of cases. In cases without allegations, by contrast, the majority (55%) of orders were for unsupervised contact. (This finding differs from previous research by Hunter & Barnett in 2013, who found courts reluctant to make a ‘no contact’ order at interim stage.
  • At first hearing where domestic abuse was alleged, about 1/3 of recommendations in ‘schedule 2 letters’ (brief initial safety reports) were for no contact, just under 1/3 were for supervised contact and just over 1/3 for unsupervised contact.
  • Where there were allegations of abuse, it was less common for unsupervised contact to be ordered (39% in cases with allegations against; 48% without).
  • Cases featuring allegations of abuse were more likely to conclude with an order for no direct contact (19%) than cases without (11%).
  • Cases featuring allegations of abuse were more likely to conclude with conditions on contact.
  • Cases featuring allegations of abuse were more likely to conclude with contact that was supervised or monitored in some way than with contact that wasn’t.
  • Referral rates to Domestic Violence Perpetrator Programmes seem low. (They are often unavailable or a perpetrator is deemed unsuitable if he doesn’t accept findings; courses are less likely to be available to women).
  • Nearly 20% of cases involving allegations ended up without an order for direct contact, in contrast to the wider picture of 88% of fathers succeeding in contact applications in the 2015 Harding & Newnham research. That was a more in-depth study of 174 cases in 2015, in half of which there were allegations of domestic abuse. Earlier research by Hunt and McLeod of a sample of 300 cases showed a success rate of 80% applicants having deirct contact order or agreed.

So, this report suggests that making allegations of abuse has an impact on the outcomes of contact applications. However, what these stats would look like if you break down proven or admitted allegations as against unproven allegations is simply known. In our view, this is a question that really needs to be answered, because the impression given is that the mere making of an allegation makes it more likely a parent will be able to restrict the other parent’s contact at the end of the case. That may or may not be shown to be the case if we had the proper detailed data.

A legitimate criticism might be the decision of CAFCASS to prioritise research which is only able to answer a very limited set of questions, and which is inevitably going to beg almost as many questions as it answers. There is a pressing need for more data about this – it is a shame that CAFCASS did not decide to commission or participate in some academically verified research project that incorporated both CAFCASS records and court files in order to produce more robust results.

Leaving aside these issues, the study doesn’t seem to entirely support the proposition by Women’s Aid in their Child First campaign last year that the family court operates on the basis of “contact at all costs” – where allegations are made the initial response and the outcome are likely to be more cautious than where no such allegations are raised – whether they are treated sufficiently seriously or not, these allegations seem not to be being ignored.

Useful insights from this research include :

  • The apparent low frequency of fact-finding hearings compared to the frequency of allegations. This is difficult to interpret, but is likely in part to be as a result of the incomplete data – there may have been fact-finding hearings that did not show on a CAFCASS file, or allegations may have been admitted or proved via criminal conviction – or may have been rolled up with a family court hearing. It’s worrying if things haven’t improved since the Hunter & Barnett research.
  • Unsupervised contact seems to have been ordered at about a quarter of First Hearing Direction Appointments where domestic abuse was raised. These are likely to be mainly cases where contact was agreed and / or where unsupervised contact had already been taking place (85% of the unsupervised contact cases had involved previous unsupervised contact)
  • The report notes that “In discussions, Women’s Aid cautioned that this may not always equate to an ‘agreement’ about contact arrangements, and may be indicative of a context of coercion.” This is a fair point, and there is existing judicial guidance about ensuring that consent orders are truly consensual rather than coerced. However, we don’t actually know whether a significant proportion of these unsupervised contact arrangements were coerced – in some cases, parents do take the view that notwithstanding abusive behaviour a child’s best interests do require unsupervised contact. Perhaps in some cases they are yet to fully appreciate the impact of abuse on a child, perhaps in others they are making an informed decision and feel strong enough to manage handovers for the benefit of the child. But whilst this study incorporates a legitimate caution about potential coercion it doesn’t provide evidence about its incidence in ‘agreements’.

Responses to the research

We’ve not spotted much in the way of response to this research other than from fathers’ groups (and Sue Whitcombe as above).

For example, CYP Now report : Fathers group criticises domestic violence study. The father’s group in question is Families Need Fathers, who are reported as saying that  “unfounded allegations were resulting in children being “denied time with their dads for many months”” and that “the findings promoted the belief that “fathers are too dangerous to be trusted with their own children””.

The question of how family courts can deal effectively with allegations that turn out to be false without damaging a child’s relationship with its father, whilst those allegations are considered, is difficult, and one which FNF are entitled to raise. But the complaint is not so much that this research has failed to tackle the problem – it plainly doesn’t, but rather that it isn’t an issue that seems a priority for research (or thought) for CAFCASS. We’re not sure that it is fair, however, to suggest that the research promotes the belief that fathers are too dangerous to be trusted with their own children. The report does consider the prevalence of allegations against parents of either sex, and makes clear that, in a majority of cases, contact does continue notwithstanding the allegations (albeit that it may be restricted in some way). The research gives us a limited insight into what happens when allegations are made, without telling us whether they are true and without telling us what ought to have happened in any individual case.

It is clear that CAFCASS’ priorities have been nudged in this direction by the impact of prominent campaigns like the Child First campaign and the Women’s Aid Homicides reports. Whilst the Transparency Project agrees that this is an entirely legitimate area for study it is a matter of concern if research priorities are driven by media campaigns that themselves are based on a flimsy evidence base. We think that the important topics of child homicides and family annihilation justify more robust research treatment than hitherto, and would welcome further research in this area that can help keep parents and children safe at and after separation.

Ex Injuria writes that the collaboration of CAFCASS and Women’s Aid is An Error of Judgement. Their objection is not simply about the quality of the research but the decision to collaborate with Women’s Aid at all :

For them to be working cheek-by-jowl with an openly anti-male, feminist propaganda organisation such as Women’s Aid is a profoundly retrograde step and a regrettable error of judgement by their CEO, Anthony Douglas.

One might say that if this proposition were correct it would almost certainly also then be correct that CAFCASS ought not to engage with fathers’ rights groups. And we don’t think that can be right. We think it is unhelpful to refer to Women’s Aid in such derogatory terms. They are a campaigning organisation whose focus is on the needs of women, just as other organisations have their own client groups too. We don’t think that means they have nothing to offer or that it prevents CAFCASS from working with them. We do think that an organisation like CAFCASS ought however to be mindful of the perception created by working with particular interest groups in ways which may be perceived as being to the exclusion or detriment of others. It is important that CAFCASS should do all it can to be and to be perceived as unpartisan.

The rest of the Ex Injuria post makes some legitimate points about the inherent limitations of the research and also about the quality of CAFCASS recordings (these are interesting but no link to source is provided so we’re not quite sure where they come from),

We can understand why on one level some of those who hold concerns about the tendency to conflate allegations of violence with actual violence and the way that this can (at least in the interim) prejudice quite safe relationships between an innocent parent (usually a father) and their child, might perceive a bias here in CAFCAS’ decision to work with Women’s Aid, particularly given the lack of clarity about quite how they have worked together. Since writing her post Sue Whitcombe has asked @mycafcass for clarification about the extent to which Women’s Aid were directly involved in the research and whether or not they had access to confidential files. The answers, provided via twitter, should provide some reassurance – but they would have been better set out clearly in the report itself and the accompanying press material if CAFCASS wished to avoid creating an unecessary anxiety amongst some of its stakeholders.

One other aspect of the safeguarding process now built into the Child Arrangements Programme which doesn’t seem to be covered in this report is the utilisation and responses to the C1A form where a party (usually but not always a respondent mother) sets out a summary of allegations of domestic abuse. Anecdotally, where completed by a respondent rather than an applicant, these are sometimes not received and considered by CAFCASS or the court in advance of or at the FHDRA, and we wonder whether this is an area of potential safeguarding risk that might also warrant consideration in any further research study.

Other areas for useful future research might be around the rates of admission / proof of allegations of domestic abuse where made in family proceedings, and the differential responses to allegations proven as against those where an allegation has been rejected, including those where a positive finding of fabrication has been made. It would be useful to understand how often intentional alienation is demonstrably a feature of a case, as compared with the prevalence of allegations (given that anecdotally allegations of alienation seem almost as commonplace as allegations of domestic abuse). The Hunter & Barnett research findings were disturbing, and it is unfortunate that no one is funding an update.