Evidence that Labour is ignoring: Solent Jobs Programme

The Learning and Work Institute recently released a briefing which, rather optimistically to my mind, suggested that 45,000-95,000 sick and disabled people might move into work as a result of employment support offered by Labour between now and 2030. The optimism came from treating all sick and disabled people as equally sick or disabled, rather than recognising that the most effective employment support programmes were offered to people who aren’t treated as unfit for work under our current system. Employment support is substantially less effective for people who are unfit for work, as might be expected.

One of the programmes that LWI referenced was the Solent Jobs Programme. This programme was offered to a range of sick and disabled people. Participants could be: people who had done two years on the Work Programme without finding work; JSA, UC, lone parents, or WRAG/LCW who had worked less than 13 weeks in the previous two years; and SG/LCWRA who wanted to work. There was therefore a mix of severities: from people in SG/LCWRA or who had been through the Work Programme without finding work, to people who weren’t treated as unfit for work despite illness or disability. Unusually, only 48% of participants had more than one health condition.

At the start of the programme, 90% of referrals were JSA recipients. It was more difficult to make ESA referrals because Jobcentre staff did not know these people as well, which made it harder to find time to explain the programme, but also led to inappropriate referrals. To address this, there was a six-month period where only ESA recipients could be referred, and mandatory interviews were put in place for ESA recipients. This may have contributed to 48% of participants being ESA recipients in 2016, though this then dropped to 38.5% in 2018, at which time 56.4% received JSA or UC.

1153 people took part in SJP, representing 42.5% of those referred. Participants received support for 12 months, though for sicker and more disabled people this was not long enough. There was an average of 3.5 individual support sessions per person, predominantly on employment support. There were also 9.7 face-to-face monthly reviews per person, and a similar number of completed phone calls.

Overall, 28% of participants moved into open employment, and 17.5% sustained it for at least 26 weeks. More than half of jobs were less than 25 hours/week. It is possible that more people sustained work, because outcomes were not monitored after Dec 2018. Regression analysis, comparing participants to similar people who had responded to the Labour Force Survey, suggested that participation in SJP may have raised employment rates by 7-8.7% points.

Interestingly, ESA SG (30%, n=83) had higher job outcomes than ESA WRAG (25%, n=363) and JSA (24%, n=381), although the ESA SG group was much smaller, so this result may not be robust. It is, however, unusual that the JSA group did not perform better than ESA WRAG. This suggests a similarity in illness/disability levels. Given that three-quarters of participants still did not find work despite relatively intensive employment support, including healthcare and job brokerage, this may mean that many of the JSA participants ought to be treated as unfit for work. This would require changes to the WCA to make it accept people with a lower level of illness or disability, which contrasts with the government’s intention of making sickness benefit harder to access.

There were limited referrals from wider agencies, outside of the Jobcentre. The report authors note that “this prevented greater engagement from residents with health conditions and disabilities that were not engaging with Jobcentre Plus and further partnership working.” There was therefore a strong selection effect.

Staff noted early on that “they struggled to provide suitable advice for those with complex and unfamiliar conditions”. This led to Occupational Health support being introduced. Staff said that this helped participants to improve their health management and general wellbeing, though Solent NHS Trust only provided direct support to 50 participants. Staff also appreciated being able to refer participants with autism to an employment profiling tool. There was also the option to refer to a two-day nature therapy course for mental, emotional and sensory resilience; and a three-day CBT-based mental health course.

A Transitional Employment Programme was offered, which was a 25-hour subsidised job paid at NMW for up to six months. This was not available to people assessed as able to enter work within 3-6 months. Just over 20% of participants registered to a TEP, and 58% of those who completed went into employment. Despite the majority of participants not using TEP (possibly implying that the majority of participants were deemed able to work within 3-6 months), it was felt to be key to outcomes. It provided an incentive for employers to engage, and reduced the risk of work for participants. Staff would help to negotiate a TEP that was appropriate for participants. Staff encouraged participants to see the TEP as a step towards employment, so that participants would continue job search during the TEP time.

A real job broker, who spent time engaging with business and becoming familiar with the local labour market, was seen as a strength in this programme. Strong links with local employers was seen as key.

Small and medium employers were often keen to support local programmes and recruit local staff. Links with employers could open up work experience, volunteering, TEP, and job opportunities. Some employers offered multiple opportunities to participants, based on having a good experience, but it is questionable whether this repeated use of unpaid and subsidised staff – without leading to a paid job in that company – is appropriate.

Sometimes, employers found that participants put forward for vacancies could be uninterested or unable to cope with the demands of the job. They wanted better job matching. This is likely to be an increasing problem as more severely sick or disabled people are expected to engage in employment support, and risks harm to sick and disabled people when they are encouraged to take up jobs that they cannot do. Other employment support programmes have also found unrealistic expectations amongst sick and disabled people. This was matched by some participants with physical illness or disability reporting accessibility issues.

For those who were positive about action plans, this was because the plans went at their pace (e.g. prioritising health and wellbeing before work), and set realistic activities and goals. Action plans needed to have appropriate timeframes, as constantly showing people who were further from work how little progress they’d made was disheartening and disappointing.

121 participants took part in a work taster. 12 left early due to finding work, and 10 left early due to it being unsuitable. Of those who completed the taster, 46 moved into work, and 34 sustained that work for at least 26 weeks.

243 participants took part in subsidised work. 94 left early due to finding work, and 33 left early due to it not being suitable. 12 people didn’t attend. Of those who completed the programme, 39 moved into work, and 21 sustained it for at least 26 weeks. Of those who left early due to finding work, 41 sustained that work. Subsidised work was seen as an important and unique part of SJP. Overall, it led to sustained work for 25% of participants, which is higher than the sustained work outcomes for the group as a whole (17.5%).

20% of participants disengaged with the programme, often because of worsening health. For 24% of participants, the outcome was unknown.

Problems gaining work could be because of worsening health; difficulties accessing healthcare; difficulties with transport; and insufficient time in a 12-month programme to address the level of need.

The SJP lacked a randomised control group, and also lacked the opportunity to use matched controls. An attempt was made to use a Coarsened Exact matching approach, but unfortunately there were too few individuals in the Labour Force Survey to match with the SJP participants, resulting in only 7% of SJP participants having a match. Nearest Neighbour matching was used instead, but the closer the ‘nearest neighbour’ was, the fewer participants could be matched. Previous research had found that the variables on which people were matched could still have significant impact on outcomes, despite the matching process being designed to eliminate this.[1] The matching variables were therefore included in a multi-variate regression.

The SJP regression found that employment outcomes were raised by 7-8.7% points (a closer match gave a higher employment rate differential). However, this is for job starts and not for sustained work. It is likely that outcomes for sustained work were smaller. As the majority of participants were treated as fit for work, these outcomes are not directly applicable to sick and disabled people who are unfit for work.

The programme cost £2000 per participant, including the cost of subsidised work.

[1] Melville et al (2018) Working Well Greater Manchester