Huge thanks Paul for some amazing comments! Mostly these can be read alongside my piece for added insight, but a couple of quick replies to each comment below
5) One thing I don't agree with is counting the ESA WRAG group (and UC equivalent) as 'unemployed' like JSA claimants and UC full worksearch. These are people who would have been eligible for Incapacity Benefit and likely Invalidity Benefit (and/or the Income Support groups required to have had the All Work Test). For consistency with previous programmes, these need to be in the analysis.
Whether they meet the ILO classification of unemployed is a different question - for the ILO, you need to have done something to look for work in a surveyed week. However, the 'something' could include what DWP would call a 'work focused interview' rather than applying for any job.
A thoughtful comment but I don't agree - this is like saying, "all those people that are classified as 'fit for work' in 2010-12 because the WCA was so harsh should be treated as incapacitated, because they'd have got incapacity benefits at different times." The point is that these people don't receive any more money, so it's confusing to call them 'an incapacity benefit claimant'.
And yes, all of these benefit classifications are a bit different to ILO classifications (though I think as you've pointed out to me before, people will be influenced in this by the benefits system itself)
That may be so if you're worried only about the spending, but if you want some consistency about people who would have been in scope of any of the benefits, this differs. Just seeing how much info can be quickly pulled out of the DWP internet archive at the National Archives, IB quarterly stats for Q ending Feb 2005 (around IB peak) has numbers for IB leavers by reason. For which you have to remember that IB Credits only claimants usually received Income Support on grounds of sickness, while not having the NI contributions required to actually receive IB. The (then) Personal Capability Assessment determined the IS eligibility. Highest group - closed certificate (Doctor's certificate finished). Under ESA or UC many of these would be dropouts. Failed to attend medical/failed to provide info - total 10%. Second biggest group, not incapable Personal Capability Assessment. All these would either have received IB or IS (pre PCA) and been counted in spending, until they left the benefit. For consistency with these, you would have to include all the WRAG and UC equivalent, plus the UC pre-WCA group now published in the UC Health stats (who are counted in the unemployed claimant count as UC counts people as that until assessed).
Of course, in the IB times, people managed to confuse themselves because IB distinguished between Claimants (including those getting Income Support only because they didn't have the NI contributions), and Beneficiaries (those receiving IB, who may also have received IS for dependants). Approx 1/3 of IB claimants did not get paid IB, but IS. IB Beneficiaries (recipients) fell almost continuously from IB introduction in 1996, while IB claimants rose through to 2005, so the gap increased.
4) The OBR report depends a lot on DWP-commissioned reports, and they reduced dramatically both the number and independence of such reports. This means that the OBR results are based on limited information. Using source DWP stats could be helpful.
For example, looking at the Work Programme, the ESA new claimants (prognoses less than 1 year) group achieved job outcome rates (cumulatively across the 2-year service period) peaking over 30% of participants getting over 3 months paid work. This was for the cohort starting in April 2015 (with a service period up to April 2017). For the ESA new claimants (12 month prognoses) group, the job outcome rate peaked at over 20% getting 3 or more months paid work (a few months later than the other group).
This Work Programme job outcome rate had built up from much lower figures - providers were strongly incentivised to secure the last possible job outcome (payment) and DWP contract managed providers who were concentrating only on new participants - they were in effect told to cross-subsidise support for the ESA groups where job outcome might take a long time from the more rapidly turning around unemployed groups.
DWP's research reports on the Work Programme were mostly about how it was introduced and implemented, and at the very end (one referenced by OBR) when DWP contract management had been moved onto other priorities - the last referrals in 2017 should have had provider out of work support through to 2019 and in-work support for a further 3 months, and DWP should have managed this. The stats became very occasional after referrals stopped, and there is more evidence of providers winding down and closing centres than supporting the remaining clients.
This was put in because people formed the view very early that this form of employment support was worse than useless. However, it's performance increased to a point well above the highest level that DWP had thought feasible. The report I co-authored for DWP on Greater Manchester Working Well (cited by OBR) found a small positive difference for the GM programme, serving ESA claimants who had completed the Work Programme without a job at the end. However, going through the figures I was quite surprised just how high the comparison group jobs and earnings were - similarly Work Programme ESA completers without a job at the end. Somehow (in both cases) ESA claimants were moving into work at relatively high rates, with the more extensive support from GM adding enough to be cost-effective.
3) In 2010, (and before under New Labour) it was seen that most ESA recipients would be at least subject to work-focused interviews. New IB claimants had been before, under Pathways to Work and before that, the pilot ONE Service. Therefore, this group was seen as a principal group for welfare to work support, which could (via black box) include some rehabilitation support, given that the NHS did minimal of that.
In practice, Work Programme providers did some, but not a lot. In the Work Programme - there were several payment groups for ESA claimants. Initially, the mandatory group was those who had been assessed as likely to be fit for work in three or six months. A second mandatory group was added later, of those expected to be fit for work in 12 months.
This indicates that, at the time, the Work Related Activity Group was not seen as those who had a lesser incapacity, but one with a time limit ("prognosis date") under 2 years.
One assessment contract wrinkle to that - the assessment provider was paid for assessments done. The inflow initial assessment wasn't influenceable by the provider, but repeat assessments were. They had an incentive to book a further assessment sooner, to increase the number of assessments required. Part of DWP contract management would have been to counter this.
Reassessments occurred on or after the prognosis date. Not surprisingly, they often found that people were still too sick to work, and the proportions of repeat assessments going into the Support Group were higher than initial assessments.
2) Using 2010 as a base for the Work Capability Assessment patterns is ridiculous. What seems evident from Select Committee Evidence sessions at the time is that either DWP, the contractor, or both, took a set of desired proportions getting into the different groups as a requirement - over 50% Fit for Work, 10% Support Group, the remainder Work Related Activity Group. These patterns are evident in the stats on initial WCA assessments.
What happened then is that both the series of reviews and a large set of tribunal appeals shifted the results and the guidance to decision makers. The percentages in the stats change in the IB reassessment period (latest stats don't distinguish new claims from IB reassessments).
This pattern of appeal results shifting the stringency of the assessment replicates the pattern under Incapacity Benefit - the All Work Test was initially seen as very tight, but over 10 years of guidance changes following court directions, it became more realistic, but too realistic for some inside DWP.
The general point is that assessment quality seems to be very differently seen inside DWP. It seems to be related to outcome percentages, rather than compliance with the criteria laid down in the regulations. Tribunals and higher courts only go on regulation compliance, and take no notice of how many other people have been assessed into categories.
1) The OBR doesn't understand who people who leave the benefit before assessment (drop-out) are. There are a lot of people who not eligible for Statutory Sick Pay, for whom an incapacity benefit forms a sickness benefit role. These include the self-employed, but also low paid employees. If they are not able to get paid while sick, they may make a claim for a period until they get well enough to work again.
There will be others who simply get well enough to work after a period on an incapacity benefit but before the assessment, and start work again. Naturally, more people will be in that position the longer assessments are delayed.
DWP ministers under the last governments used drop-outs to claim that people were being frightened off benefit rolls by the tough assessment. The last New Labour ministers tended the same way. This was never substantiated.
Really helpful again - dropouts are really important, and I think we understand them amazingly little (partly because they're hard to do research with - but it would make sense to do a focused research project to find out more)
Huge thanks Paul for some amazing comments! Mostly these can be read alongside my piece for added insight, but a couple of quick replies to each comment below
5) One thing I don't agree with is counting the ESA WRAG group (and UC equivalent) as 'unemployed' like JSA claimants and UC full worksearch. These are people who would have been eligible for Incapacity Benefit and likely Invalidity Benefit (and/or the Income Support groups required to have had the All Work Test). For consistency with previous programmes, these need to be in the analysis.
Whether they meet the ILO classification of unemployed is a different question - for the ILO, you need to have done something to look for work in a surveyed week. However, the 'something' could include what DWP would call a 'work focused interview' rather than applying for any job.
A thoughtful comment but I don't agree - this is like saying, "all those people that are classified as 'fit for work' in 2010-12 because the WCA was so harsh should be treated as incapacitated, because they'd have got incapacity benefits at different times." The point is that these people don't receive any more money, so it's confusing to call them 'an incapacity benefit claimant'.
And yes, all of these benefit classifications are a bit different to ILO classifications (though I think as you've pointed out to me before, people will be influenced in this by the benefits system itself)
That may be so if you're worried only about the spending, but if you want some consistency about people who would have been in scope of any of the benefits, this differs. Just seeing how much info can be quickly pulled out of the DWP internet archive at the National Archives, IB quarterly stats for Q ending Feb 2005 (around IB peak) has numbers for IB leavers by reason. For which you have to remember that IB Credits only claimants usually received Income Support on grounds of sickness, while not having the NI contributions required to actually receive IB. The (then) Personal Capability Assessment determined the IS eligibility. Highest group - closed certificate (Doctor's certificate finished). Under ESA or UC many of these would be dropouts. Failed to attend medical/failed to provide info - total 10%. Second biggest group, not incapable Personal Capability Assessment. All these would either have received IB or IS (pre PCA) and been counted in spending, until they left the benefit. For consistency with these, you would have to include all the WRAG and UC equivalent, plus the UC pre-WCA group now published in the UC Health stats (who are counted in the unemployed claimant count as UC counts people as that until assessed).
Of course, in the IB times, people managed to confuse themselves because IB distinguished between Claimants (including those getting Income Support only because they didn't have the NI contributions), and Beneficiaries (those receiving IB, who may also have received IS for dependants). Approx 1/3 of IB claimants did not get paid IB, but IS. IB Beneficiaries (recipients) fell almost continuously from IB introduction in 1996, while IB claimants rose through to 2005, so the gap increased.
4) The OBR report depends a lot on DWP-commissioned reports, and they reduced dramatically both the number and independence of such reports. This means that the OBR results are based on limited information. Using source DWP stats could be helpful.
For example, looking at the Work Programme, the ESA new claimants (prognoses less than 1 year) group achieved job outcome rates (cumulatively across the 2-year service period) peaking over 30% of participants getting over 3 months paid work. This was for the cohort starting in April 2015 (with a service period up to April 2017). For the ESA new claimants (12 month prognoses) group, the job outcome rate peaked at over 20% getting 3 or more months paid work (a few months later than the other group).
This Work Programme job outcome rate had built up from much lower figures - providers were strongly incentivised to secure the last possible job outcome (payment) and DWP contract managed providers who were concentrating only on new participants - they were in effect told to cross-subsidise support for the ESA groups where job outcome might take a long time from the more rapidly turning around unemployed groups.
DWP's research reports on the Work Programme were mostly about how it was introduced and implemented, and at the very end (one referenced by OBR) when DWP contract management had been moved onto other priorities - the last referrals in 2017 should have had provider out of work support through to 2019 and in-work support for a further 3 months, and DWP should have managed this. The stats became very occasional after referrals stopped, and there is more evidence of providers winding down and closing centres than supporting the remaining clients.
This was put in because people formed the view very early that this form of employment support was worse than useless. However, it's performance increased to a point well above the highest level that DWP had thought feasible. The report I co-authored for DWP on Greater Manchester Working Well (cited by OBR) found a small positive difference for the GM programme, serving ESA claimants who had completed the Work Programme without a job at the end. However, going through the figures I was quite surprised just how high the comparison group jobs and earnings were - similarly Work Programme ESA completers without a job at the end. Somehow (in both cases) ESA claimants were moving into work at relatively high rates, with the more extensive support from GM adding enough to be cost-effective.
3) In 2010, (and before under New Labour) it was seen that most ESA recipients would be at least subject to work-focused interviews. New IB claimants had been before, under Pathways to Work and before that, the pilot ONE Service. Therefore, this group was seen as a principal group for welfare to work support, which could (via black box) include some rehabilitation support, given that the NHS did minimal of that.
In practice, Work Programme providers did some, but not a lot. In the Work Programme - there were several payment groups for ESA claimants. Initially, the mandatory group was those who had been assessed as likely to be fit for work in three or six months. A second mandatory group was added later, of those expected to be fit for work in 12 months.
This indicates that, at the time, the Work Related Activity Group was not seen as those who had a lesser incapacity, but one with a time limit ("prognosis date") under 2 years.
One assessment contract wrinkle to that - the assessment provider was paid for assessments done. The inflow initial assessment wasn't influenceable by the provider, but repeat assessments were. They had an incentive to book a further assessment sooner, to increase the number of assessments required. Part of DWP contract management would have been to counter this.
Reassessments occurred on or after the prognosis date. Not surprisingly, they often found that people were still too sick to work, and the proportions of repeat assessments going into the Support Group were higher than initial assessments.
Really helpful re the conditionality that people faced at different times (and what the WRAG was meant to be) - thanks Paul
2) Using 2010 as a base for the Work Capability Assessment patterns is ridiculous. What seems evident from Select Committee Evidence sessions at the time is that either DWP, the contractor, or both, took a set of desired proportions getting into the different groups as a requirement - over 50% Fit for Work, 10% Support Group, the remainder Work Related Activity Group. These patterns are evident in the stats on initial WCA assessments.
What happened then is that both the series of reviews and a large set of tribunal appeals shifted the results and the guidance to decision makers. The percentages in the stats change in the IB reassessment period (latest stats don't distinguish new claims from IB reassessments).
This pattern of appeal results shifting the stringency of the assessment replicates the pattern under Incapacity Benefit - the All Work Test was initially seen as very tight, but over 10 years of guidance changes following court directions, it became more realistic, but too realistic for some inside DWP.
The general point is that assessment quality seems to be very differently seen inside DWP. It seems to be related to outcome percentages, rather than compliance with the criteria laid down in the regulations. Tribunals and higher courts only go on regulation compliance, and take no notice of how many other people have been assessed into categories.
Several comments, but I'll separate them.
1) The OBR doesn't understand who people who leave the benefit before assessment (drop-out) are. There are a lot of people who not eligible for Statutory Sick Pay, for whom an incapacity benefit forms a sickness benefit role. These include the self-employed, but also low paid employees. If they are not able to get paid while sick, they may make a claim for a period until they get well enough to work again.
There will be others who simply get well enough to work after a period on an incapacity benefit but before the assessment, and start work again. Naturally, more people will be in that position the longer assessments are delayed.
DWP ministers under the last governments used drop-outs to claim that people were being frightened off benefit rolls by the tough assessment. The last New Labour ministers tended the same way. This was never substantiated.
Really helpful again - dropouts are really important, and I think we understand them amazingly little (partly because they're hard to do research with - but it would make sense to do a focused research project to find out more)