A further word on the 'real trend in out-of-work benefit claims'
As a share of working-age people, we're nowhere near 'record levels'
Over the past few weeks, I’ve explained why it’s misleading to use the official statistics on trends in out-of-work benefit claims, and what the real trend looks like. However, this looked at the total number of claimants - and didn’t take into account the increasing number of people in Britain over time (as Jonathan Portes has pointed out).
I thought I’d quickly add a chart on the share of the working-age population that’s been claiming out-of-work benefits over time - further details in the footnote.1 I cover two periods: the trend since 2013 (when the official figures become untrustworthy), and the trend since 1981 (for a broader historical context). The results look like this:
The real trend in out-of-work benefit claims, accounting for (some, but not all) methodological changes
Sources: ‘DWP out-of-work claims is the Fraser Nelson chart from post #1. ‘Corrected out-of-work claims’ adds the a additional claimants from the official Alternative Claimant Count statistics (because we’ve changed our way of counting people in Universal Credit), and subtracts claimants who are not aged 16-59 (women) or 16-64 (men) - see the post #1 for explanations and sources. The dashed figure for the Aug 2023 data reflects the lack of data from the Alternative Claimant Count since Aug 2022, which has therefore been estimated (see footnote to post #2), and also the lack of population data (see footnote 1 for population data sources).
The picture here is pretty similar to my last post - we cannot account for all the methodological changes, so that we know that claims in the past should really be higher (there’s no cut-off for this: the effect starts in 2015 and gets bigger over time, and will continue to get bigger until UC is fully rolled-out). But even using the things we can account for, it’s clear that claims of out-of-work benefits are lower in 2023 than in 2013.
But it’s also important to go back further, to put this in a broader context. Adding the graph above to earlier figures gives us the following picture:
The real longer-term trend in out-of-work benefit claims, accounting for (some, but not all) methodological changes
Sources: see chart above.
There’s a break in the series here, because the old way of counting misses lots of people off who would be counted as claimants under Universal Credit (as I explained here). And this isn’t a clean break, because we gradually roll out UC, the numbers artificially go up.
So to be clear - there’s no way of knowing if claims are higher today than they were just before Covid. They might be, but then again, Covid led to a large number of people moving from legacy benefits to UC, and the claim is artificially higher on UC than it is on legacy benefits. (And as I said before, at some point I’ll try to look at other data sources to see if we can say more about this).
But what is clear is that despite the old way of counting leading to lower numbers, there were still barely fewer out-of-work benefit claimants for most of the past 40 years than there are today - and sometimes there were a lot more.
Failed to render LaTeX expression — no expression found
For most of this period, the working-age population has included men aged 16-64, and women aged 16-59. As I explained in the first post in this series, the state pension age has been rising, initially for women, and then for both men and women. To make things consistent, I’ve here focused on the 16-59 (women)/16-64 (men) population.
Population data are from nomis (‘Population estimates - local authority based by five year age band’) - note that this only gives information on the 15-59/64 year old population for 1981-1991, so I’ve crudely estimated the number of 15 year olds in the earlier period (simply by dividing the 15-19 age group by five). There’s also no data for Britain for 2022 or 2023 yet; for 2022 there are data for England & Wales, so I’ve applied the 2021→2022 rise for E&W to the whole of Great Britain; and for 2023 I’ve conservatively assumed no population growth 2022→2023 (which is unlikely to be true, but won’t change the figures very much).
Also, I’m not sure I’ve clarified before - these are all figures for Great Britain unless otherwise specified (the DWP’s figures are given for Great Britain rather than the UK, as Northern Irish figures are published separately).