This is a first post from Kelly Reeve, new contributor, who tweets as @s37CA89research. Kelly recently completed a socio-legal research study on the use of section 37 Children Act 1989 at UEA.  She was formerly a legal adviser at Coram Children’s Legal Centre and legal researcher at 4PB family law chambers.

At the beginning of May various news reports claimed that the number of separating couples going to court to resolve disputes about their children was 4 times higher than previously believed. See for example the Mail Online story, Custody fights blight four in ten break ups as 50,000 former couples battle in court rooms every year reporting new figures showing almost 4 out of 10 (38%) separating couples need the assistance of the court rather than 1 in 10 (10%), revealing “a large pool of hidden misery among broken families”.

The President of the Family Division, Sir Andrew McFarlane, in a keynote address on 5 April, had made brief reference to the new figure of 38%, (crediting Teresa Williams, Cafcass’ strategy lead) and to the “urban myth” that only 1 in 10 couples need to go to court to resolve disagreements about how they should care for their child post-separation. He acknowledged that the new figure indicated a “major societal problem”.

It is this speech which seems to have inspired a blog and press release from Families Need Fathers calling for a renewed look at shared parenting, which triggered the news articles.

So where have the new figures come from?

The media reports quote from figures released by Cafcass. A bit of digging reveals that the numbers come from estimates in a (January 2019) report for the Nuffield Foundation Family Justice Observatory by Cafcass Director of Strategy Teresa Williams, summarised here.

Williams bases her estimate on a 2017 report by Bryson, Purdon and Skipp for the Nuffield Foundation which estimates that 2% of households with dependent children separate each year. This report acknowledges that there are significant shortcomings in the UK evidence base on family separation. It is not entirely clear how the 2% estimate was calculated, the report itself cites one of the author’s own analyses (McKay) alongside the UK household longitudinal survey (UKHLS, p27). However, the authors themselves admit that the UKHLS is not entirely reliable and it appears that the McKay study is not publicly available to see how this figure has been calculated.

Williams then uses the figure of 6.3m households with dependent children in England and Wales to calculate that there are approximately 125,000 separations per year (2%). Again, it is not immediately clear where the 6.3m figure comes from, Williams’ report has no citation. Williams then uses Cafcass figures of 42,000 new private law applications in England alone per year to estimate that a third of separating families (33.3%) are using the courts.

A word of caution

The authors of these reports are very clear that these are only estimates in the absence of any  concrete data but this same caution is not reflected in the ensuing media reports. 

There are a couple of immediate points which show a need for caution. Is the same definition of ‘separation’ being used for all data sets? The Bryson et al report estimate of 2% relies on a definition of ‘separation’ as those situations where a child does not live with both of their parents with no assumption that they ever did so; this is a much broader definition than former couples disputing living arrangements following a relationship breakdown, as the media reports characterise it.

Similarly, the Mail article interprets private law applications as ‘custody fights’ between ‘former couples’ but the Cafcass figures on private law applications reflect a much broader spectrum. These applications are not just made by parents, they can also be made by wider family members, grandparents or siblings seeking contact with a child or to make living arrangements. Private law applications also go beyond ‘custody fights’ and include applications for specific issues, for example, disputes relating to a child’s education, medical treatment or religion.

Also, somewhat curiously, Sir Andrew McFarlane and the news reports mention different figures to those actually in Williams’ report. McFarlane says that there are approximately 8m families with dependent children in England and Wales, 130,000 couples with dependent children separating each year and 50,000 ending up in private law proceedings, making it around 38% of couples who need to go to court to resolve how children should be cared for post-separation. 

Clearly there is no reliable comprehensive data available which sets out what proportion of separating couples resort to court to resolve child arrangement disputes and estimates is all that we have. However, in the interest of transparency, and especially when such estimates are being used to advocate for changes to the private law system, there is a need for absolute clarity as to what these estimates are based on and an explanation as to why the figures quoted in Williams’ report differs from those in McFarlane’s speech.

This March 2019 press release ahead of a joint Cafcass/ ACSS international conference on shared parenting suggests that while Cafcass have not published the figures they are ‘available on request’. We asked for them but were signposted only to Williams’ report above. 

With the report of the private law working group on what to do about family separation due for publication imminently it would be helpful to have greater clarity on how the level of the problem has been estimated.

POST UPDATE (13th August 2019):

Professor Stephen McKay, one of the authors of the 2017 Bryson- Purdon report quoted in the post, helpfully got in touch with us in response to our post with some further insights. We set them out below for readers interested in the statistics. Our thanks to Professor McKay:

It would be nice to have a source where a sample of separated parents were asked about their contacts with the legal system, and indeed with CAFCASS, but to my knowledge there is no good source (maybe no source at all) for that. That would be the best kind of source.

Anyway – short version is that I think the estimate of about 125,000 couples-with-kids separating each year is reasonable, though still an estimate. Some justification of that, below. The press/calculation then divides the number of CAFCASS cases by the number of separations to generate the 38% (or similar) figure. This is where I have less knowledge. To do that division it implies that 1 CAFCASS case relates to 1 family. I don’t have the information to know if that is how their systems work or if there are often multiple cases for the same family (e.g. for multiple issues within the same family).

What do we know? ONS estimates 6.2 million couples with dependent children. This comes from the Labour Force Survey. [source:https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/families/bulletins/familiesandhouseholds/2017 ].

Also note that it is for the UK, rather than the jurisdiction of CAFCASS (England) so the number of couples with children in England is closer to 5 million. If I look at the Family Resources Survey for 2017-18 (the survey that is used for the low income or ‘poverty’ figures) then there are 5.1 million couples with dependent children (up to age 18 in education) living in England.

How many separations? Some older work I did on this is available here: http://marriagefoundation.org.uk/wp-content/uploads/2016/06/pdf-016.pdf. That also arrives at a 2% figure each year. The same survey data was used for the Bryson, Purdon, Skipp report – though up to age 18 rather than 16, but that didn’t change the 2% figure.

ONS divorce data (though only up to 2014) showed how many included children under 16.  So, in 2014, there were 111,169 divorces (E&W), of which 50,168 involved children under 16. Again, if we believe the survey estimates, we can roughly double that to arrive at total separations with families with children. Divorce data (2014) from sheet 8 ofhttps://www.ons.gov.uk/file?uri=/peoplepopulationandcommunity/birthsdeathsandmarriages/divorce/datasets/divorcesinenglandandwales/2014/divorcetables2014.xls)

Importantly, note that, the figures also show that 80,000 couples with children got divorced in 2004, but only 50,000 in 2013. Families may be getting more stable, so even a constant CAFCASS workload might imply a rising proportion of cases.

So, it seems plausible that about 125,000 couples with children separate each year. It may be a bit higher or lower. Trends are downward, though.


Image – Creative Commons: Nick Youngson CC BY-SA 3.0 Alpha Stock Images. With thanks.