Evaluations the UK government commissioned keep finding its back-to-work plan barely moves the number

Listen to this article

Played by your own device. Nothing is recorded and nothing is sent anywhere.

By the Living Unlimited Team

The argument for cutting disability benefits usually rests on a promise: that the right support will move people with health conditions into work, so the payments they lose can be replaced by wages. In late June 2026, the UK’s Department for Work and Pensions published a run of evaluations of the schemes built to keep that promise, most of them carried out for the department by outside researchers including Ipsos UK, the Institute for Employment Studies and IFF Research. Read together, they are a caution against assuming the support does what the promise needs it to do.

We report these findings as findings. They come from evaluations the department commissioned and published, which state that their views are not necessarily those of the Department for Work and Pensions, and they matter here because the same logic, that employment support can substitute for benefit income, runs through Canadian debates about disability poverty and the value of work.

A scheme for complex needs, and a 15 per cent result

The Intensive Personalised Employment Support programme, known as IPES, ran in England and Wales from December 2019, with the last people referred onto it in December 2023. By the end of that referral period it had supported over 11,000 people, all of them with what the department calls complex support needs related to health conditions and disabilities. The program offered fifteen months of support and a further six months once someone was in work.

The evaluation’s follow-up survey found that 15 per cent of participants were in paid work after completing the program. The report does not set that figure against a comparison group who received no support, so it cannot say how many of those jobs the program itself produced. It is a headline number without a counterfactual, and the evaluators are candid that the barriers this group faces are heavy and that the pandemic hit the labour market hard during the years the scheme ran.

When the people who left do just as well

The second evaluation is harder on the theory. The Individual Placement and Support in Primary Care scheme, or IPSPC, puts employment advisers inside GP surgeries and other primary care settings. Its survey found that the share of participants in paid work rose from 28 per cent at the start to 39 per cent at the final survey, which sounds like a win.

Then comes the sentence that undoes the win. Employment rose by a similar amount among the people who disengaged from the program, from 38 per cent to 46 per cent, as among those who stayed with it, from 27 per cent to 34 per cent. The evaluators put it plainly: the changes “cannot be attributed solely to programme participation.” If the people who walked away improved as much as the people who stayed, the program is not clearly the reason anyone found work.

The same report notes that the model never fully took hold as designed. Only 30 per cent of referrals came from primary care, against a target of 90 per cent, and advisers struggled to co-locate inside GP practices or reach patient records. Getting health and employment services to sit in one room turned out to be, in the report’s words, a complex and time-intensive undertaking.

The work coach evidence cuts both ways

The third strand is about work coaches, and it is genuinely mixed, so we will not pretend otherwise. A DWP impact evaluation published in March 2025 found that Additional Work Coach Support, extra appointment time for people in the limited capability for work and work-related activity group, produced a small but statistically significant lift: 11 per cent of participants were in work twelve months on, against 8 per cent of a matched comparison group, a three-percentage-point gain.

A later evaluation of the same scheme, published by the department on 25 June 2026 and built on a longitudinal survey with a matched comparison group, found no employment effect at all. It reported no significant evidence, overall or in any subgroup, that the support changed whether people said they were in paid work: 18 per cent of both the supported and unsupported groups were in paid work at the time of the follow-up survey, a difference of zero percentage points. It also found “very little significant evidence” of an effect on wellbeing. It did find that the scheme improved how people rated their relationship with their work coach, and that it lifted job search activity and volunteering among customers in the limited capability for work group.

The two studies are not measuring the same people, and that is worth holding on to. The March 2025 impact evaluation covered only customers in the limited capability for work and work-related activity group. The June 2026 evaluation left that same group out of its impact estimates, because the researchers could not build a robust matched comparison group for them. Neither study overturns the other. What is safe to say is that the two work-coach studies, one by the department’s own analysts and one carried out for it by IFF Research, do not add up to a settled case that more coaching time reliably moves people with health conditions into jobs.

Disability News Service also reported that in the same week these evaluations landed, ministers announced they were expanding a similar voluntary support scheme from six jobcentres to thirty-three.

Why a Canadian reader should watch this

Canada is not running these exact programs, but it is having the same argument. The Canada Disability Benefit is now paying out, provincial income-assistance systems still attach work expectations to disability support, and the assumption underneath much of it is the one these UK reports test: that with enough of the right help, people with health conditions can be moved off benefits and into earnings that replace them.

The value of this batch of research is not that support is worthless. Participants in these schemes reported real gains in confidence, routine and wellbeing, and those matter. The value is that the employment numbers, measured in evaluations commissioned and published by the government that runs the schemes, are modest, sometimes uncertain, and in one case indistinguishable from what happened to the people who left. That is worth knowing before anyone treats a job as a benefit cut’s replacement.

Sources

Similar Posts