No workaround

In March, the Government released its Pathways to Work Green Paper, setting out a package of welfare reforms that amounted to a net reduction in spending of £4.8 billion in 2029-30. On the Government’s own figures, 3.2 million families will lose out, 250,000 people will fall into poverty, and 700,000 families will fall further below the poverty line. These benefit cuts were accompanied by a significant increase in employment support costing a cumulative £1.9 billion between 2026-27 and 2029-30 but with over half of that not coming until the final year.

When discussing these changes, the Chancellor said that: “I am absolutely certain that our reforms, instead of pushing people into poverty, are going to get people into work.” But the reforms will most likely do both: some people will move into employment and some people will fall into poverty. We estimate that the Government’s cuts to disability and incapacity benefits will lead to between 38,000 and 57,000 more people in paid work by 2029-30, while additional employment support delivers extra employment of between 23,000 and 48,000. Under a best-case scenario 105,000 more people would be in work by the end of Parliament.

There are large uncertainties around these numbers, and we have not tried to quantify any gains from changes in conditionality regimes, or possible disemployment effects from the increase in the UC standard allowance. But it is clear that any increases in employment by the end of the decade will be far too small to fully offset the hit to family incomes.  Even if each and every one of the extra jobs were to prevent people from crossing the poverty line, the increase in poverty would not be halved. And if employment gains are evenly spread among the losers, they amount to only 3 per cent of those affected moving into work. So even after employment increases are accounted for, low-income families will suffer material income losses and the reforms will cause higher poverty rates.

Given this, the government should: accelerate the improvements in employment support; provide at least six months’ notice for those losing Personal Independence Payment (PIP); and ensure that existing recipients of the health-related element of UC do not face reassessment and lower benefits in the future if they move into work.

The backdrop: the Government’s Pathways to Work Green Paper has announced a multi-billion pound set of benefit cuts, alongside £1.9 billion of extra employment support

In March, the Government announced a set of benefit reforms with the clear intention of reducing spending and helping to ensure it met its fiscal rules in the recent Spring Statement.[i] The main confirmed changes are:

• Reforming the assessment process for Personal Independence Payment (PIP), by narrowing the eligibility criteria for the daily living part of PIP. Previously, claimants needed to score at least eight points from the 10 daily living headings to qualify for the standard rate of this element; under the new system, claimants will also need to score at least four points in any single heading. This will affect any new claim or reassessment of an existing claim from 2026-27 onwards. By 2029-30, 800,000 people are expected to lose the daily living part of PIP and 400,000 will lose PIP altogether.
• Cutting the rate of Universal Credit for people with health conditions (UC-H). For existing claims, this element will be frozen in cash terms for the rest of the Parliament; for new claims from 2026-27, its value will be halved (from £97 a week to £50 a week) and then frozen for the rest of the Parliament. This change is estimated to affect 3 million families by 2029-30 (730,000 new claims and 2.3 million existing claims).
• Increasing the value of the UC standard allowance. This will happen gradually between 2026 and 2029; by the end of the Parliament, the value of the UC standard allowance will be 5 per cent (£5 a week) higher than it would have been under the default uprating scenario. This will benefit 6.8 million families overall, but will lead to a net increase in benefit income only for the 3.8 million households expected to be receiving Universal Credit by 2029-30 who are not affected by the UC-H cut.

The ’winners and losers’ from these benefit changes are summarised in Figure 1 below.

Figure 1: By 2029-30, 7 million families will be affected by the welfare reforms announced this Spring, almost half of whom (3.2 million) will lose support

These benefit changes will happen alongside additional spending on Department for Work and Pensions (DWP) employment support programmes. The additional spend will total £1.9 billion between 2026-27 and 2029-30, but the majority of the spending (£1 billion) does not kick in until 2029-30 – well after the benefit cuts begin to bite. The shape of this new employment support has not yet been specified; the Spring Statement says that it “will provide employment and health support to anyone receiving out of work benefits with a work-limiting health condition. This investment will build on existing support from WorkWell, Connect to Work and the Get Britain Working trailblazers.” This additional spending comes on top of increased funding of £240 million in 2024-25 for employment support already announced by the Government in the Autumn Budget and Get Britan Working White Paper.

The Government hopes that the increase in poverty resulting from its benefit cuts will be minimised by more people entering employment

The Government’s own impact analysis suggests that 3.2 million families will lose out, and 250,000 extra people will fall into poverty as a direct result of its benefit reforms, with this figure rising to at least 300,000 if we do not account for the poverty-reducing impact of scrapping the previous Government’s changes to the Work Capability Assessment. Meanwhile, 700,000 families who are already in poverty will fall deeper below the poverty line.[i]

But Ministers have said that this static modelling does not paint an accurate picture of the real-world impacts of the benefit cuts. The Chancellor went so far as to say, “I am absolutely certain that our reforms, instead of pushing people into poverty, are going to get people into work. And we know that if you move from welfare into work, you are much less likely to be in poverty.”

At the time of writing, and with a vote on these reforms coming up, there is no official assessment of this statement. In their March 2025 Economic and Fiscal Outlook, the Office for Budget Responsibility (OBR) said that ‘the Government did not provide us with a comprehensive and robust analysis of these potential effects, and we were not, in the very limited time available, able to develop our own analysis of their net impact on labour supply.’ With this in mind, in this note we present our own assessment.

There are three main mechanisms through which the set of reforms in the Green Paper might lead to an increase in employment:

• The PIP cut reduces incomes for the affected group, irrespective of the amount of work done. The financial pay-off to working hasn’t changed, but this group may seek to mitigate the loss by working more (a positive ‘income effect’).
• The UC-H cut reduces out-of-work incomes for the affected group, increasing the financial pay-off to doing paid work. This strengthening of work incentives will increase labour supply (a positive ‘substitution effect’).
• Finally, the increase in funding for employment support will increase labour supply.

Our approach to calculating the impact of these four aspects of the Government’s benefit reforms on labour supply is summarised in Box 1 below.

Box 1: How we estimate the employment effects of changes to PIP and UC-H

Our starting point is to identify individuals in the Family Resources Survey who are directly affected by the benefits changes – i.e.  those living in households in receipt of PIP, and those eligible for UC-H. Next, we calculate the proportional changes in unearned incomes (for PIP) and gains to work (for UC-H) that the benefit changes create. In the latter case, for households not currently in work, we make assumptions about how much they would earn if in work.

We then calculate the average proportional changes in unearned income and gains to work and multiply them by elasticities from the literature (i.e. measures of how sensitive behaviour is to financial changes) to obtain proportional or percentage-point changes in full-time equivalent (FTE) employment. For the PIP cuts, we consider elasticities of FTE employment to unearned income between -0.12 and -0.22. For UC-H, we use a range of 0.08 to 0.12 for the semi-elasticity of participation with respect to the gains to work.  See Annex 1 for some key references in the literature from which we draw these estimates.

Finally, we gross these changes up by the size of the affected groups to obtain these changes in FTE terms. This final step is important in shaping the results: the UC-H cuts are proportionally much smaller for the affected group (the part of the UK working-age population that has a work-limiting health condition and would pass the UC asset and means test) than the PIP cuts, but affect many more people, as can be seen in Figure 2.

Contrary to the Government’s claims, the overall employment effects of this Spring’s benefit and employment support reforms are expected to be small

Using the methodology described in Box 1 above, we estimate the labour supply impacts of the Government’s benefit reforms (to PIP and UC-H). Our results suggest that these reforms could increase labour supply in 2029-30 by between 38,000 and 57,000 (in FTE terms).

We have also considered the impact of the £0.9 billion employment support that will be allocated before 2029-30. To do this, we follow the method set out in a report from the Learning and Work Institute to calculate the average ‘cost per job’ of recent UK-based employment support programmes. With that figure, we estimate that £0.9 billion of employment support is likely to help between 23,000 and 48,000 enter employment.[ii] Note that we take into account only the £0.9 billion of employment support that has been allocated between 2026-27 and 2028-29, considering that the money allocated for 2029-30 cannot plausibly have an impact on labour supply in that year due to the time lag between people between receiving employment support and having a sustained employment outcome.

Having done this, our estimate is that the total FTE increase in labour supply could lie between 60,000 and 105,000, but this must be seen as highly uncertain. These estimates are shown in Table 1.

Table 1: The overall employment effects of the Spring’s benefit and employment support reforms are much smaller than the affected population

The fact that this increase in labour supply is much smaller than the number of people affected by the cuts reflects two main points.

First, the groups affected by the cuts to PIP and UC-H are – by definition – those with health conditions and disabilities, many of whom face significant barriers to work. Only one-in-six (17 per cent) PIP claimants are currently in employment, and more than half (55 per cent) of PIP claimants are aged 50 or over, a group who tend to have low labour supply elasticities.[i] As a result, not all of these claimants will be able to respond to reductions in benefit income by entering the labour market, and this is reflected in the elasticities we use in our modelling.

Second, our modelling has accounted for the phased nature of the cuts which mean that only some of the big reductions in income are experienced by 2029-30.[ii] For example, the PIP changes will only affect existing claimants when they go through a reassessment, and existing UC-H claims only see their support frozen in cash terms (rather than halved, as is the case for new UC-H claims). The implication of this is that the labour supply response could grow through the 2030s.

Figure 2 helps us understand why the increase in labour supply is much smaller than the 3.2 million families affected by the cuts. While those affected by PIP cuts (both existing and new claimants, shown in red) will face a large reduction in income, this group is relatively small in size and includes claimants with low income elasticities. We estimate that losses for this group stand at an average of £4,200 per year for those only affected by PIP cuts, and £4,940 for those affected by cuts to both PIP and UC-H. On the other hand, while more people will be affected by the cut to UC-H (shown in orange), on average this group face a smaller change in income by 2029-30, of £960 per year. This is because three-quarters of affected families (2.3 million out of 3 million) are existing claims whose UC-H is frozen rather than halved. Finally, Figure 2 shows that the 3.8 million families who gain from the UC standard allowance boost gain by just £265 per year on average.[iii]

Figure 2: Families who lose out from the PIP reforms see much bigger changes in income than those who gain or lose from the UC reforms

Finally, it is important to reflect on the limitations of our modelling, and to consider missing factors that might lead to an increase in labour supply. We list the most important ones below:

• By reducing the value of UC-H and increasing the value of the UC standard allowance, the Government has reduced the gap between unemployment- and incapacity-related benefits. The Government hopes that fewer people will claim UC-H as a result, with these people instead ending up in the non-health-related UC caseload.[i] If this happens, we would expect an increase in labour supply, because people in receipt of non-health-related UC move into work much more often than those in receipt of UC-H. In November 2024, 5.9 per cent of those in receipt of non-health-related UC flowed into employment each month, compared to just 0.9 per cent of those in the Limited Capability for Work and Work-related Activity (LCWRA) group of UC-H and 1.9 per cent of those in the Limited Capability for Work (LCW) group of UC-H.[ii] Some of this difference reflects differences in the characteristics of these groups, but some will reflect the tougher conditionality regime that applies to non-health-related UC. If, for example, 1 per cent of the UC LCWRA group were instead in receipt of non-health-related UC, then, holding current into-work flow rates constant, we would expect the number of people entering work each year to increase by around 18,000.
• The Green Paper also included changes to conditionality for people receiving UC-H by creating “a new baseline expectation of engagement” where “most people in receipt of the health element in UC should be expected in the reformed system to, as a minimum, engage in conversations from time to time about their aspirations to work and to hear about the support available to them”. This change is being consulted on, so it is not clear if or when it will take effect, or what this new engagement will look like. However, if this change does take effect, and is distinct from the extra funding for employment support that we have accounted for, we would again expect labour supply to increase as a result of more people in the UC-H group entering employment.
• The Government has estimated that, by 2029-30, 150,000 people will lose entitlement to Carer’s Allowance or the carer’s element in UC when the person they are caring for loses entitlement to PIP. It is possible that some of these people may move into work as a result, although as full-time carers their behaviour may be relatively insensitive to their financial circumstances.

On the other hand, we have also not considered some points that might mean that the estimated employment responses are over-estimates. For example:

• There is anecdotal evidence that some people use PIP to help stay in work – such that a cut could potentially reduce labour supply for some. The aggregate labour supply impact of the PIP cuts could accordingly be closer to zero.
• We have not considered any potential disemployment effects of the increase in the UC standard allowance. This increase is small, happening gradually between 2026 and 2029, and is raising the rate of UC from a very low base. However, it has the potential to influence the labour supply of a large group of current and potential recipients of UC.

The employment effects will be too small to prevent a fall in incomes and a rise in poverty in the affected group

Given the lack of detail on some of the reforms, the limited evidence on how the labour supply of people with health conditions respond to financial incentives, and the difficulty of modelling switches between and changes to conditionality regimes, our estimates of the eventual size of the employment effects should be seen as tentative. However, it is clear that any positive employment effects will be far smaller than the 3.2 million families affected by the cuts, and certainly too small to fully offset the 250,000 crossing the poverty line or the 700,000 families pushed further below it.[iii]

Under our best-case scenario, 105,000 people enter employment – or just one-in-forty (3 per cent) of the estimated 3.8 million adults (in 3.2 million households)[iv] affected by the benefit cuts by the end of the Parliament.[v] Many of these will not earn enough additional income to fully offset their income losses. If we assume that 3 per cent of those who are affected by cuts to the extent that they are dragged below the poverty line respond by moving into work, then the increase in poverty is reduced by just 7,000. This would be a rounding error on the government’s poverty estimates (which are rounded to the nearest 50,000). Even if we took the very extreme assumption that all of the employment gains are among those who are moved into poverty and that the employment gains were sufficient to avoid the move into poverty, this would not even halve the increase in poverty (it would be reduced by 105,000, from 250,000 to 145,000).  Assuming that those who move into work are guaranteed to escape poverty is an unrealistic assumption. In 2023-24, 17 per cent of PIP claimants who work part-time are in poverty, and 6 per cent of those working full-time are in poverty.[vi]

Extra funding for employment support is welcome, but this will not protect families from benefit cuts unless brought forward to earlier in the Parliament, so the Green Paper reforms should be accompanied by transitional protections and safeguards

Our estimates show that it is extremely unlikely that the rise in employment these reforms cause will be big enough to prevent a major rise in and deepening of poverty. The Government should accordingly bring forward measures to accompany the cuts to benefits announced in the Green Paper. First, the improvements in employment support should be accelerated. Second, the changes to the assessment for the daily living part of PIP should come with a form of transitional protection, giving affected existing recipients (i.e. those who do not score at least four points on any of the activities when reassessed) at least six months’ notice before any cut comes into effect. These changes should not enter into force until the Government’s announced review of PIP has concluded to ensure the changes are being made to a system that the Government is satisfied is fit for purpose. Thirdly, the proposals in the Green Paper that moving into work does not trigger reassessment of UC-H recipients should be implemented by 2026-27: currently, the Government is only consulting on these proposals, so it is not certain if or when they will be implemented. Doing so would mitigate the disincentive to work that the prospect of rejoining the system at lower rate would create.[vii]

Annex 1:  What does academic literature tell us about the impact of health-related benefit cuts on labour supply?

• • A 2014 paper from Kostol & Mogstad, using data from Norway, finds that many disability benefit recipients do enter work when benefit policy changes. Importantly, they find that it is only younger disability benefit claimants (aged 18-49) who are responsive to changes in work incentives; they found no labour supply response among older claimants. This has important implications for understanding the likely impact of the PIP changes, since older claimants are more likely to be impacted than younger claimants.
  • A 2016 paper from Koning & Van Sonsbeek, using data from the Netherlands, calculates a labour elasticity of 0.12 among affected disability insurance claimants. They find that labour elasticities are highest among younger claimants, as well as those whose health condition is a ‘mental impairment’.
  • A 2011 paper from Marie and Vall Costello, using data from Spain, finds that disability benefit recipients who receive a more generous award reduce their labour supply. They calculate a labour elasticity of 0.22. In Spain, disability benefit receipt is not work-contingent, so the authors conclude that the labour supply effect is mainly due to an income effect.
  • A 2020 paper from Garcia Mandico et al, using data from the Netherlands, assesses how employment and earnings change in response to benefit cuts. They find that both employment and earnings rise in response to disability benefit cuts, and that younger recipients, women, and those with ‘more subjectively defined disabilities’ are most able to increase their earnings. On the other hand, those who had been claiming disability benefits for a long time were less likely to respond by increasing their earnings.
  • A 2015 paper by Frutos and Vall Costello, using data from Spain, finds that the probability of working is 5 per cent lower among disabled people who receive benefits than those who do not. However, they find that the disincentive effects of receiving benefits are only significant for those with ‘the mildest level of disability’.
  • A 2023 paper from Biro et al, using data from Hungary, finds that among those who lose eligibility for disability benefits as a result of reforms, the chances of finding employment vary sharply depending on previous employment history. The chances of finding employment were small among those who were not in work pre-reform. They also find that those who returned to employment tended to end up in lower-quality jobs than previously.
  • There is lots of wider academic literature on labour force participation elasticities. For example, a 2016 paper from Senaj et al finds that lower-skilled workers, and women, are particularly responsive to changes in income (by studying the effect of changing income taxes and transfers). A 2020 paper by Bastanie, Moberg and Selin similarly finds that labour force participation elasticities are larger (twice as large) among the group with the lowest employment level than those with the highest employment level.

[i] See Section 3.20 in: OBR, Economic and Financial Outlook, March 2025

[ii] RF analysis of DWP, Pathways to Work: Evidence pack: Chapter 1 case for change evidence, May 2025.

[iii] While the overall impact on incomes and poverty is likely to be adverse, the precise numbers and identity of the winners and losers will depend on who among the affected populations moves into work. Considering the PIP cuts as an example, and individuals with less unearned income to begin with will see a proportionally bigger cut and will be at greater risk of poverty, so their labour supply might respond more. On the other hand, those with more severe health problems might be expected to respond less than average to a change in financial circumstances.

[iv] We have estimated the number of adults affected by the UC-H cut by using the number of families affected by the UC-H cut and a ratio of adults to households in receipt of Universal Credit. Source: RF analysis of DWP, Stat-Xplore; DWP, Spring Statement 2025 health and disability benefit reforms – Impact, May 2025.

[v] This also assumes that the rise in FTE employment is equal to the rise in employment in heads. In practice, the rise in heads could be bigger if the extra work is mostly part-time, or smaller if a lot of it corresponds to extra hours worked by people already in work.

[vi] These figures refer to the proportion of working-age PIP claimants who are in poverty, measured on an after housing costs basis. Source: RF analysis of DWP, Stat-Xplore.

[vii] Data citation: Department for Work and Pensions, NatCen Social Research. (2021). Family Resources Survey. [data series]. 4th Release. UK Data Service. SN: 200017, DOI: http://doi.org/10.5255/UKDA-Series-200017.

[i] RF analysis of DWP, Pathways to Work: Evidence pack: Chapter 2 reforming the structure, May 2025; DWP, Stat-xplore.

[ii] We might expect the labour supply impacts to increase over time, when the scale of the PIP and UC-H cuts increases.

[iii] Our estimated net gain differs from that published in the DWP impact analysis (£420) since we do not include the impact of not taking forward plans to change the Work Capability Assessment as set out by the previous Government; our figure matches the DWP estimated gain from the increase in the standard allowance (£265). Our estimated net losses differ slightly from those published in the DWP impact assessment: our net loss resulting from the UC-H cuts is an average across new and existing claimants, and our net loss for those affected by PIP changes reflects our best attempt to model the change to the PIP assessment with limited data.

[i] These two estimates come from different pieces of DWP analysis, with one being done in terms of individuals, and the other in terms of families. We thank Chaminda Jayanetti for sharing Freedom of Information (FoI) request FOI2025/29715.

[ii] Another summary of the effectiveness of relevant DWP employment programmes can be found in: DWP, Pathways to Work: Evidence pack: Chapter 3 supporting people to thrive, May 2025.

[i] For a more detailed account of the benefit changes announced in the Pathways to Work Green Paper in March 2025, see: M Brewer, A Clegg & L Murphy, A dangerous road? Examining the ‘Pathways to Work’ Green Paper, Resolution Foundation, March 2025. For further analysis of the 2025 Spring Statement, see: C Aref-Adib et al., Unsung Britain bears the brunt, Putting the 2025 Spring Statement in context, Resolution Foundation, March 2025.