Earlier this month, I wrote here about the law firm EMW and their narrative – eagerly sucked up by a few so-called journalists – about the supposedly low take-up of statutory paternity leave. I suggested that EMW’s narrative is little more than a pile of pants. And today, in response to my Freedom of Information request, HM Revenue & Customs have confirmed that EMW’s narrative is indeed, as the never knowingly understated Independent might report it, “broken”.

In short, EMW have for years based their narrative on HMRC figures for (a) the number of claimants (new fathers, mostly) paid statutory paternity pay, and (b) the number of claimants (new mothers) paid statutory maternity pay. Using the latter as a proxy for the number of fathers who are eligible for statutory paid paternity leave is highly questionable, but in any case it turns out those HMRC figures are inflated by a lot of double-counting. As HMRC confirm:

The number of claimants is the total number of individuals in receipt during that year, irrespective of when the payment first started. Where a given spell of [statutory maternity pay, statutory paternity pay or statutory shared parental pay] extends across two years, the claimant will be included in both years’ figures.

In the case of statutory maternity pay, this inflationary effect is substantial: up to 75%, if we assume (as seems to be the case – see image below) that, on average, women take nine months of SMP (i.e. their full entitlement). But with statutory paternity pay, the inflationary effect is negligible (because claimants can only take one or two weeks of paternity leave).

Maybe this doesn’t matter very much – the few press reports that EMW secured last month with their annual press release about the same set of HMRC data haven’t sparked a national debate about reform of paternity leave. Interestingly, however, EMW are also responsible for another take-up figure that frequently does feature in policy debate.

Only today, I received a document that states “the current UK [shared] parental leave scheme has a very poor uptake, with only around 1% of fathers using the scheme”. And, like many before them, the author of this document was almost certainly thinking of this news release from the TUC in April 2019:

The TUC is today calling for an overhaul of shared parental leave.

Last year only 9,200 new parents took shared parental leave – just 1% of those eligible to do so.

The TUC believes take-up is low because the scheme is so low-paid (£145.18 per week) making it unaffordable for most fathers.

A ‘note to Editors’ at the bottom of the news release states: “the University of Birmingham found that only 9,200 new parents (just over 1% of those entitled to take it) took SPL in 2017/2018”. However, the University of Birmingham’s short September 2018 report makes clear (see Reference 3) that it simply lifted the 1% figure from a (since deleted or moved) press release from … the law firm EMW.

Furthermore, it’s clear (including from a later EMW press release that has not been deleted) that EMW got to that 1% figure (1.4%, to be precise) by dividing HMRC’s figure for the number of claimants paid statutory shared parental pay in 2017/18 (9,200) by HMRC’s figure for the number of women paid statutory maternity pay that year (662,000) – that is, in the same way they calculated the take-up rate for paternity leave.

However, we now know that figure for the number of women paid SMP is inflated (by up to 75%) by HMRC’s double-counting. And in any case EMW are wrong to assume that the number of women who went onto statutory maternity pay is the same as the number of fathers who are eligible to take shared parental leave. In 2013, in its impact assessment of the new policy (see Table 7 on p29), the Department for Business, Energy & Industrial Strategy assumed the maximum number of eligible fathers to be just 285,000 (a figure BEIS has never felt the need to revise).

Whatever, now that HMRC have confirmed how their raw annual figures are inflated by double-counting, and using that BEIS figure of 285,000 eligible fathers, we can adjust those HMRC figures, to remove the likely double-counting in respect of shared parental pay, and so arrive at new estimates for the rate of take-up of shared parental leave among eligible fathers since 2015/16. (Note that we also need to adjust the HMRC figures to allow for the fact that, as confirmed in Footnote 5 of this Answer to a Parliamentary Question by Ed Miliband in February 2021, at least 20% of HMRC’s claimants for statutory shared parental pay are mothers who have converted their maternity leave to shared parental leave, as they must do before the father can access statutory shared parental leave and pay.)

I have previously suggested that SPL take-up among eligible fathers was as ‘high’ as 3.6% in 2019/20. But I didn’t know then that HMRC’s annual figures for the number of shared parental pay claimants include the double-counting that HMRC have now confirmed. I should have known better, frankly, but for what it’s worth here are my new estimates of SPL take-up:

[NB: If you are a nerd and would like to see how I arrived at these figures, just ask and I’ll send you a spreadsheet]