How far is PIP an ‘out-of-work benefit’?

On one level, it obviously isn't - but the full picture is more complex...

In the wake of Rishi Sunak’s widely-reported speech about PIP, there have been politicians of all sides talking about how PIP disincentivises work. Often this seems to completely misunderstand PIP, which on the face of it, is not an out-of-work benefit – leading to many disability commentators despairing about how little politicians understand about the benefits system.

These criticisms are fair – but at the same time, the relationship between PIP and work is more complex than it appears. In this blog post I dig into these figures to provide a fuller picture of the relationship between PIP and work.

The overlap between PIP and out-of-work benefits

Let’s start by looking at a consistent age range, to get rid of the effect of changing pension ages on PIP claims,¹ and focusing on the past decade.²

The first way of looking at the PIP-work relationship is to look at the overlap between PIP and out-of-work benefits. It’s possible to do this using Stat-Xplore, but it’s trickier than it should be³, which is why I don’t think these figures have been published before in this way - and also why it’s taken me a couple of weeks longer to write this post than I planned…

Obviously I’m looking at both PIP and DLA (because PIP replaced DLA over this period). And remember that there’s no way of counting ‘out-of-work benefits’ consistently over time (at least, not in the official data)!

What’s striking here is just how many people claiming PIP are also claiming out-of-work benefits – 71% in August 2023. We’ve known this for a while, because DWP regularly shows the overlap between PIP and incapacity benefits (e.g. the 2021 Green Paper T1.3 showed that 66% of PIP/DLA claimants also claimed ESA/UC Health, and the 2016 Green Paper T4c showed the earlier figure was 57%), but putting it this way makes the point particularly clear.

Finding out the overall level of employment among working-age PIP claimants is harder - there’s just not a lot of data available, and it’s not usually restricted to working-age claimants. But we’d expect the share of PIP claimants that are out-of-work to be even higher (because some people are out-of-work and don’t claim out-of-work benefits), and this is what the data shows. The 2024 PIP reform consultation shows that 13-18% of PIP claimants are employed, based on merging DWP & HMRC data. This is similar to other data, e.g. , a DWP-commissioned 2016 survey of claimants (part of this) found that 12% of PIP claimants reported being employed.

Finally, it’s difficult to look at trends, because – as I keep saying – there’s no way of counting out-of-work benefits consistently over time.

• Claims: Looking at the graph, the proportion of PIP claims that overlap with out-of-work benefits stayed remarkably similar over the last decade (at 70-72%). This is actually quite surprising, because the way that DWP counts ‘out-of-work benefits’ has changed over time, and the numbers on out-of-work benefits back in 2013 would be higher if they were counted in the same way as today (and there was also an unquestionable rise in out-of-work claims at the start of Covid). So it’s probably the case that if this was counted consistently, a greater share of PIP claims in the past overlapped with out-of-work benefits.

• Employment, if we look at the employment data from the 2024 PIP reform consultation (which is stronger for looking at trends), we see that there was a decline in the % of PIP claimants that were employed (from 18% to 13% 2015→2018), after which it was stable, and then rose again since 2021 to 16%.

My take is that over the five years pre-Covid, fewer PIP claimants were also out-of-work benefit claimants - but slightly more of them were out-of-work. This is consistent with a greater proportion of people using DLA/PIP when they weren’t able to work, but weren’t eligible for (or didn’t want to claim) incapacity benefits. Since Covid, in contrast, there has been a particularly rise in employed claimants.

Another take: looking at new claims

There’s a quirk of this data, though, which is that the share of new PIP claims that overlap with out-of-work benefits is lower, as shown in the chart below. For most of the last decade 50-60% of new claims overlapped with income-replacement benefits, lower than the share of all claims that overlap with them (70-72% out-of-work benefits, 77-81% for all benefits - the DWP chart below’s definition is somewhere between these).

This difference is even more striking when we look at employment, in the figure below. Nearly half of all PIP applicants in 2023 were employed. This has also changed over time: in 2016, 30% of PIP applications were by people who were employed - this rose to 36% on the eve of Covid, and to over 45% in 2023. However, over time, employed PIP applicants have become slightly more likely to be rejected than other claimants, so the rise in employment among people awarded PIP isn’t as strong - from under 30% 2016 → 35% on the eve of Covid → 41% today.

How can it be the case that 41% of new PIP awards go to people who are employed, but only 16% of all PIP claimants are currently working? There’s three possibilities.

Firstly, working PIP claimants may be more likely to fall out of work, vs. how likely non-working claimants are to start work. This is supported by evidence in the 2021 Green Paper, which showed that among one group of PIP claimants,⁴ 26.4% were employed in April 2019 - but nearly half of them (10.4% of all claimants) were not employed in March 2020, whereas only a small number of claimants (2.2%) had moved into employment over the year. So this is definitely part of it.

Secondly, underlying these figures, it’s possible that receiving PIP itself increases the chances of disabled people leaving work (or decreases the chances of non-employed people moving into work). I recently found the economist Joe Spearing’s work (also here), which suggests that - at most - receiving DLA/PIP only has a slight impact on work. (I haven’t gone through this study in any detail though; see also the comment below). Lots of politicians have weighed in on this, and I’ll come back to it in a wider briefing in June.

Finally, it could be that non-working people are likely to claim PIP for longer. This is also plausible, because non-working people may have more severe or long-lasting disabilities; and also may have a greater financial need to claim. But I haven’t seen any direct evidence of this.

So what does all this mean?

This is a really complex issue (which has taken me a long time to get my head around).

One the one hand, it’s clear that PIP is not an out-of-work benefit in the usual sense. It’s not limited to people in work, and nearly half of all new applications are by people who are working. And I’ve spoken to several people who said that PIP has helped them sustain work. So it is fair to call out politicians who seem to misunderstand the benefits system.

On the other hand, most people claiming PIP are also claiming an out-of-work benefit, and the overwhelming majority of PIP claimants at any one time are not working. My main takeaway message is that it is absolutely impossible to understand PIP without understanding how it will be affected by wider changes in the benefits system - I have a strong hunch that eligibility restrictions re incapacity benefits (e.g. ESA), and cuts in the generosity of benefits more widely have both contributed to rising PIP.

But more on this in future posts, as I need to take breath after submerging myself in this data…

1. The headline rise in PIP claims among working-age people is 1.1million, from a baseline of 1.9million – a rise of 59%. However, a quarter of this rise – 315,000 people – simply comes from people who would not have been counted as ‘working-age’ back in 2013. The real rise in PIP claims is a still-large 800,000 people, a 43% increase, but a smaller increase than some of the headlines.

There’s also been a growth in the number of people in the working-age population over this period, so an even more accurate figure is to say that there are 38% more people claiming PIP in August 2023 than Feb 20213 – a rise from 4.9% percent of the working-age population to 6.8%. So the 38% rise is the most accurate figure for the ten-year change.

2. Figures are for Feb 2013 to August 2023, because this is when I currently can get DWP data from Stat-Xplore on the combinations of benefits that people claim.

3. There is no single table that will allow you to look at the overlap between PIP and other benefits. I have therefore constructed this from the benefit combinations datasets on Stat-Xplore, including both (i) the overlap of PIP/DLA/AA and the various different benefits that are categorised by DWP as ‘out-of-work’, and (ii) excluding people on out-of-work benefits, looking at the overlap of PIP/DLA/AA and a series of other benefits (‘UC Working only, PIP, HB only, CA only, HB & CA, BSP/WB/BB only, IIDB only, HB & PC, HB,  UC Working, PC only’). Each of these needs constructing from three datasets, one pre-2021 (annoyingly this has now been changed to pre-2019…) and separate ones for England&Wales/Scotland since then. For all of these, I split this by age and gender (because ‘working-age’ changes over time), and then focus on a consistent group of 16-64 year olds. Just let me know if you’d like to look at the underlying working tables.

4. This is T4.12, which is a slightly odd table due to data limitations. Self-employed people are counted as ‘non-employed’. It’s not clear if this is new claimants in 2019, or the stock of claimants in 2019 (the employment rate doesn’t match PIP claimants as a whole, but the table notes that the data aren’t representative of all claimants because it only looks at this that could be matched to HMRC records).