Have new WCA claims exploded in the last year?

A quick post on some more frustrating DWP statistics, ahead of some longer posts on broader issues.

One of the most eye-opening figures I’ve seen lately is that the number of new incapacity claims (that is, successful initial WCAs) has more than doubled since July 2025, from about 1500 per working day to more than 4000.1

This figure was reported in the Times (paywalled), and Fraser Nelson has a longer (free) post about it. Fraser often creates clear data visualisations, so this is his take on what the rise looks like (though note the footnote above for context):

Just like in my last post, this is a fair way of presenting what the DWP have released - but even so, the figures and newspaper stories are unreliable, because the DWP have (again) made a complete mess out of communicating the issues in the data.

The problem here is that this graph is meant to show ‘initial’ WCAs (that is, assessments for people who haven’t already been through a WCA, i.e. new incapacity claims). But this data only looks at WCAs while people are on UC - it excludes WCAs while people were on ESA. So someone who was previously claiming ESA and is now having a WCA on UC is counted as an ‘initial claim’, even though it’s clear that this should be a ‘reassessment’.

I’ve not been able to find this clearly stated online anywhere (whether in the UC WCA stats release, or on Stat-Xplore, or on a wider Google search) - the closest is a vague statement on the Stat-xplore notes. But when my RF colleague Louise Murphy checked with the DWP analytical team concerned (as part of writing this), they explained:

“[Would] someone who has a WCA after migrating from ESA to UC count as having an initial or repeat assessment?

They would show as Initial i.e. the stats would show them as having an Initial UC WCA. In the Release Strategy we do have plans to include new breakdowns of these UC health statistics in the future to show claimants with previous (or concurrent) ESA spells. Updates on this will be via the: Statistical work programme - GOV.UK.”

(Just to add - I really appreciate how speedily and helpfully the relevant DWP analytical team responded to our queries. The issue here is a wider one about how DWP communicate the comparability of trends, rather than about the competence or helpfulness of individual DWP analysts, which I’ve found to be consistently great).

Now, Fraser Nelson’s post does briefly consider this, saying “One theory is that the jump comes from those being moved from the old to new sickness benefits. But they bypass assessment, so they should not show up in these DWP ‘assessment outcome’ figures. And even if it is this, then it’s a scandal: being moved on to LCWRA is to be written off. ”

This is true - people going through the managed migration from ESA to UC shouldn’t automatically receive a WCA just because they’ve moved onto UC. (And as far as I can tell from things like this, there haven’t been reported concerns around this, so it’s presumably been happening as it should).

But when they do eventually have a WCA, this is misleadingly going to be counted as an ‘initial’ assessment, rather than a ‘repeat’ assessment. So over time, this is going to push the numbers of ‘initial WCAs’ up - which raises the question: how big an impact will this make?

To answer that, we need to know two things. The first is the number people moving from ESA to UC. This started in July 2024 (just before the ‘initial WCA’ figures above start to rise in Sep 2024), and at it’s height hit nearly 4,000 claims per working day (which is larger than the initial WCA rise above) - as shown in another one of Fraser Nelson’s charts:

The second thing we need to know, though, is when these people have their first WCA on UC - and we simply don’t have any data. The earliest that this is likely to happen is a few months after the migration notice was sent, and the latest is maybe 3 years later (as this is reportedly the maximum amount of time before a WCA reassessment), or maybe much longer (as it seems like DWP hasn’t been prioritising reassessments for a while, so relatively few of them have been happening2).

In conclusion, one take here is “these comparability issues are unlikely to change the figures much because few reassessments have been happening. So the story of a seismic change in new incapacity claims is true.” 3

But at the same time, a different and equally reasonable take would be, “the huge increases happening just after the managed migration from ESA (and the parallel rise in approval rates4) are a smoking gun, suggesting that this is simply a methodological artefact”. And if so, then Fraser Nelson’s claim that this is still a ‘scandal’ (in the quote above) becomes hard to justify.

And I can’t confidently tell you which of these are two accounts are true, because of the way the figures have been created. This is just incredibly frustrating - it’s such a basic piece of information on a key political issue that affects huge numbers of people. If anyone has any good ideas for trying to change the way that DWP communicates issues in the data, then please let me know - otherwise I’ll try to explore this in both chats with DWP people, and (as the UC WCA stats release itself suggests) contacting the Office for Statistics Regulation.

And I promise - in my next couple of posts, I’ll try to focus on the issues and not the data…