Will employment support make a substantial difference to sick and disabled people who are treated as unfit for work by the WCA?

The current Labour government is seeking to make substantial cuts to social security for sick and disabled people. A major plank of Labour’s argument in favour of this seems to be that sick and disabled people will mitigate the loss of social security income by moving into paid work. Sick and disabled people will be able to do this because of the employment support that Labour is going to provide.

In order to assess whether employment support proves to be an adequate answer to social security cuts, we need to know:

How many people assessed as unfit for work by the WCA would have to move into work, and at what pay rate, under the new system, for Labour to conclude that their approach is successful?
Conversely, how many people assessed as unfit for work by the WCA would have to remain out-of-work under the new system, for Labour to conclude that their approach is failing?
And how many people assessed as unfit for work by the WCA would have to remain out-of-work under the new system, for Labour to conclude that their approach is not just failing (but could be improved) but is fundamentally flawed, because the underlying assumption – that people assessed as unfit for work by the WCA can work if given employment support – is in fact incorrect?

Evidence from employment support

When considering outcomes for employment support, it is worth noting that 1-1.5% of WRAG/LCW claimants move into work each month (up to of 18-24%/year), and 1% of SG/LCWRA claimants (up to 12%/year). An employment support programme would need to result in more people than this moving into work, in order to have had an effect above that of the base-rate returns to work.

It is also noteworthy that when a person is reassessed under the WCA, only 4% of outcomes are Fit For Work. People are already moving themselves off sickness benefit at a substantially faster rate than occurs by reassessment. This suggests that people do not need to be assessed as fit for work in order to recognise themselves as fit for work, but can and do self-assess their capacity for work and return to work of their own initiative when they are able to do so. There is no need for social security cuts to incentivise people to return to work or to remove ‘perverse incentives’.

New Deal for Disabled People

The previous Labour government had a New Deal for Disabled People. This programme changed over time and did not carry out the Randomised Controlled Trials or Matched Controls analysis that has characterised Conservative employment support programmes. It is therefore difficult to determine how many people who got work after participating in NDDP would have done so anyway.

The NDDP was in place under the previous assessment and sickness benefit system, the Personal Capability Assessment (PCA) and Incapacity Benefit (IB). The PCA was less strict than the WCA (for example, a person who had lost one arm would qualify as unfit for work under the PCA, but not under the WCA). Participants in the NDDP were voluntary and were healthier than the typical Incapacity Benefit recipient; and the typical IB recipient was healthier than the typical ESA/UC sickness benefit recipient. This cohort of people is therefore not applicable to current sickness benefit recipients, because current sickness benefit recipients are more sick/disabled and further from work than NDDP participants.

The NDDP saw 25% of participants move into sustained paid work. This is at the top end of the current rate at which WRAG/LCW recipients move into work, which is to be expected given that NDDP participants were healthier than WRAG/LCW recipients. Report authors note that it is likely that the majority of these people would have moved into work anyway(1). A later report gave a 10-11% point increase in employment, but that refers just to job entries and not to sustained work.

The impact of NDDP is likely to have been over-estimated. Staff in the NDDP would work with IB recipients prior to registering them on the NDDP, and then register people who were close to getting work. This is likely to have artificially inflated the employment rate of NDDP compared to non-NDDP participants, by excluding from the NDDP those people who were least likely to get work despite employment support. The matched controls analysis did not include the matched variables in a regression analysis, but other research indicates that the matched variables do still have an impact that needs to be controlled for after the matching (2). Stafford et al also report that, “the inability to control for individual characteristics and circumstances not included in the administrative data, may have caused a bias away from zero accounting for about one-third of the measured effect in the estimates presented here (Orr et al., 2007:16 and 20-21). For this reason, the estimates presented here should be viewed with some caution, although the conclusion remains that NDDP did have a significant net impact.”

Employment support for disabled people who are fit for work

A number of employment support programmes have been aimed at sick and disabled people, whilst in practice predominantly serving people who are treated as fit for work. This occurs because a person can be deemed disabled under the Equality Act 2010 whilst not meeting the WCA criteria for being found unfit for work. These people are often neglected from policy consideration and treated as though they have no health problems or disabilities, despite often having substantial difficulties in getting or obtaining work, such as only being able to work part-time or needed substantial adjustments.

Despite being aimed at people who are ostensibly fit for work, employment support programmes for this group have limited success. Increased employment rates range from -5 to +12 percentage points. This demonstrates the level of incapacity for work experienced by people already excluded from the ‘unfit for work’ categories in our social security system, and the inadequacy of employment support for overcoming health-related difficulties with work even for people well enough to be treated as fit for work.

The Group Work/Jobs II programme(3) had no impact on employment and, whilst most people did have improved mental health, 20-34% of participants experienced worsened mental health.
Group Work was a one-week course of four hours per day. It was aimed at people who were struggling with their job search and/or feeling low or anxious and lacking in confidence in relation to their job search.
Employment Advisers in IAPT(4) resulted in negative work outcomes for those who were in-work (88% vs 93% of matched controls in work at the time of survey), no impact on those who were off work sick, and a positive impact on those who were looking for work (26% vs 21% of matched controls moved into work).
EA in IAPT offered employment support to people receiving low-level counselling. Only 14% of those who attended two or more IAPT sessions also took up EA support.
The report authors noted that moving into work is not necessarily a positive outcome, because work can be bad for health and wellbeing.
Health-Led Employment Trials(5) had a positive work outcome in the West Midlands Combined Authority (22% vs 18%), but no impact in the Sheffield City Region (25% vs 27%). SCR controls performed the best, and outperformed those who received employment support at WMCA. The report authors suggested that the stronger labour market in SCR may have caused this result.
HLET was based on the Individual Placement and Support model, but was time limited, had higher caseloads, made less use of job brokership, and had less integration with healthcare. It was aimed at people with low to moderate mental or physical health problems. Improvement in health was a key contextual factor in returns to work. Conversely, people with more severe (despite still being moderate) illness could not be supported towards work, and work could have negative impacts on these people’s health.(6)
Work Choice(7) saw 23% of participants achieve long-term sustained work of 26 weeks or more, but this was not compared to matched controls, and was lower than the 35% who achieved 13 weeks in paid work. When taking static point-in-time employment rates, 37% of participants compared to 25% of controls were in paid work, and these figures were sustained at eight and eleven years. However, Work Choice analysis used matched control rather than randomised controlled trials, which can lead to over-estimates of impact.
Work Choice cost £4,590 per participant and predominantly helped jobseekers. This was because Work Choice was aimed at people who were expected to be able to work at least 16 hours/week after participating in Work Choice, which typically ruled out people assessed as unfit for work.
The Solent Jobs Programme saw 17.5% of participants sustain work for 26 weeks or more. More than half of jobs were less than 25 hours/week. Looking at job entries, 28% moved into work, and it is estimated that that is 7-8.7% points higher than would have occurred in the absence of SJP. However, it is unknown what the impact on sustained work was.
SJP provided twelve months of support to sick and disabled people, and provided a range of high-quality support including job brokership, work tasters and subsidised work, and support for mental health and wellbeing. SJP cost £2000 per participant, including the cost of subsidised work. Around half of participants were jobseekers.
The Work and Health Programme(8) saw 19% of participants compared to 16% of controls enter paid work in 18-24 months. Over half of these people were only working part-time. 8% of participants had started work but had subsequently left work.
The W&HP was aimed at people defined as disabled under the Equality Act 2010 and who were expected to be able to find work within a year of starting the programme. Despite this expectation, 81% still hadn’t moved into work 6-12 months after being expected to have done so, and only 3% could be said to have obtained work because of the help provided by the W&HP.
80% of W&HP staff said that participants’ needs were a little (26%) or a lot (54%) more complex than they had expected.

Employment support for people who are assessed as unfit for work

For people assessed as unfit for work, outcomes from employment support programmes are in the range of 0-4 percentage points. Outcomes even with employment support remain below average, which is 1.5-2%/month (up to 18-24%/year) for people assessed as WRAG/LCW, and 1%/ month (up to 12%/year) for people assessed as SG/LCWRA.

Greater Manchester Working Well had no impact on job entry rates. Only 4.7% of participants entered work at all over the two-year programme, and only 1.6% of participants sustained work for 26 weeks.
GM WW was for people with WRAG/LCW status who had been through two years on the Work Programme without finding work.
Central London Working Capital had limited impact on job entry rates, with an estimate of 1.7% points. Job entry rates were around 11-13%, but job sustainment was only 7.5%. Central London had a stronger labour market than the rest of London and the UK, which may have contributed to the job entry rates compared to programmes in other parts of the country.
CL WC was for people with WRAG/LCW status who had been through two years on the Work Programme without finding work.
CL WC was a one-year programme, which was reported to be too short for some participants, despite the fact that they had just received two years of support via the Work Programme.
The Work-Related Activity Pilots(9) had negligible outcomes. Despite being aimed at people expected to be able to find work within 18-24 months, 28-30% of participants were reassessed as SG/LCWRA in that timeframe; i.e., they moved away from rather than into work.
The Jobcentre Pilot provided nine hours of support over two years. 8% of participants compared to 4% of controls moved into paid work according to Anderson et al(10), but Dawson and Smithers report the employment difference as only 1% point and not sustained over time(11). The difference in employment outcomes had been eroded by year seven, and amounted to only two extra months in work over seven years(12). There was no impact on benefit receipt.
For the Work Programme Pilot and Healthcare Pilot, there was no impact on work outcomes.
People were far more likely to move away from than into work.
Pilot staff reported that some of the participants sent to them were too ill for the programme to be helpful.
Additional Work Coach Support(13) for people assessed as SG/LCWRA saw 11% of participants compared to 8% of the comparison group moving into paid work.
AWS provided 30 minutes of appointment time with a Work Coach each month.
Looking at the broader group of people waiting for a WCA, assessed as WRAG/LCW, or assessed as SG/LCWRA:
- A consistent supportive person was helpful, but the actual knowledge base of work coaches, both in terms of employment support and knowledge of health conditions, was seen as too generic;
- Some people experienced worsened mental health because of the pressure to attend;
- There was no impact on the ability to manage physical health conditions.
- Work coaches could help financially by supporting people to move onto SG/LCWRA and/or claim PIP; and by reassuring people that “they were worthy of receiving financial support”;
- The majority neither moved into work nor felt meaningfully closer to work;
- When people did move into work, it was typically at rates low enough to not affect benefit receipt.

Conclusion

People assessed as unfit for work have severe static disabilities or severe, long-term physical or mental illness that means these people are unable to sustain themselves viably through paid work. Paid work is a risk to health and detracts from important human activities such as family, community, and religion. Employment support is unable to help these people because the fundamental barrier is a lack of health or the presence of a profound disability that cannot be overcome with aids, adaptations, or support. Cutting social security for these people is profoundly dangerous and immoral, due to their inability to mitigate an inadequate income through paid work.

Employment support for people who are unfit for work has an impact in the range of 0-4% points. Assuming that 4-600,000 people receive employment support(15), that sees a maximum of 24,000 people move into work, but a more realistic outcome is just 10,000 people (2% of 500,000). This is not enough to justify cuts to social security.

1. Evaluation of the New Deal for Disabled People Personal Adviser service pilot (2001) DSS RR144

2. Melville et al (2018) Working Well Greater Manchester

3. An evaluation of the Group Work (Jobs II) trial: Evaluation synthesis report (2021) DWP RR991

4. Employment Advisers in Improving Access to Psychological Therapies Evaluation (2022) DWP RR1014

5. Health-led Employment Trials Evaluation 12-month outcomes evidence synthesis (2022) DWP RR1025

6. Health-led Employment Trial Evaluation. 12-month outcomes report: Theory-based evaluation (2022) DWP RR1037

7. Work Choice Impact Evaluation; A voluntary employment programme for people with disability barriers to employment (2025) DWP ad hoc RR111

8. Work and Health Programme Evaluation (2023) DWP RR1044

9. Employment & Support Allowance: Evaluation of pilots to support Work-Related Activity Group customers with an 18 to 24 month re-referral period (2019) DWP RR965

Dawson and Smithers (2019) Employment & Support Allowance: Evaluation of pilots to support Work-Related Activity Group customers with an 18 to 24 month re-referral period. A synthesis of evidence.

10. Employment & Support Allowance: Evaluation of pilots to support Work-Related Activity Group customers with an 18 to 24 month re-referral period (2019) DWP RR965

11. Dawson and Smithers (2019) Employment & Support Allowance: Evaluation of pilots to support Work-Related Activity Group customers with an 18 to 24 month re-referral period. A synthesis of evidence.

12. Work-Related Activity Group Pilots: 7-year Impact Evaluation

13. The impact of additional Jobcentre Plus support on the employment outcomes of disabled people (2025)

14. The experience of Additional Work Coach Support: Findings from qualitative interviews with customers. (2025) DWP RR 1093

15. Learning and Work Institute (2025) Estimating the impacts of extra employment support for disabled people. Briefing Note.