NextFin News - Can Britain reform its welfare state without turning employment support into a benefit cut? Alan Milburn’s review of young people who are not in education, employment or training is designed to answer that question, but its importance extends beyond the under-25s. Official data put the number of 16- to 24-year-olds in this position at 1.012 million in January to March 2026, or 13.5% of the age group. The expected remedy combines tougher conditions for some young claimants with billions of pounds of employment support and a higher-rate benefit for people with severe disabilities. That trade-off is the test. If ministers fund the investment and accept that savings may take almost a decade, the review could mark a structural reset. If they use it mainly to reduce near-term spending, it will be retrenchment under the language of reform.
The Number Is Large, but the Policy Problem Is Larger
The official ONS estimate is a warning about the pipeline into the labour market, not a measure of laziness. The 1.012 million young people classified as NEET rose by 89,000 from a year earlier and by 55,000 from the previous quarter. The rate rose 1 percentage point year on year to 13.5%. The series is an official statistic in development, so the direction matters more than false precision: the deterioration is broad enough to demand policy attention, but the category is too wide to justify treating every person in it as the same.
That distinction is central to Milburn’s mandate. The review is examining the causes of rising NEET rates among young people who claim health and disability benefits, including childhood experience, health, skills, welfare design and the barriers that prevent people from learning or earning. Its terms of reference call for a holistic view of welfare, health, skills and employment rather than a narrow benefits audit. The review is therefore aimed at a transmission mechanism: poor health or weak qualifications can reduce access to education; prolonged inactivity can erode skills and confidence; a fragmented system can then make a first job harder to obtain and a return to work harder to sustain.
The government has already placed money behind that diagnosis. It announced £1.5 billion over the Spending Review for the Youth Guarantee and apprenticeships for up to 50,000 young people. But the scale of the problem exposes the limit of a single programme. A training place is not a job; a job is not necessarily sustainable; and an adviser cannot compensate for untreated mental illness, poor transport or an employer unwilling to accommodate a health condition.
The recommendations described ahead of publication sharpen the political choice. Some under-25 claimants could receive less support at the beginning of a claim if they do not engage with work-related help, while people with more severe disabilities unable to work could receive a new benefit at a higher rate. The package is also expected to include work placements, job subsidies and personal advisers. Milburn has warned that the new system would require significant investment and could take almost a decade to produce monetary savings.
“A deal isn't a deal unless there are potential consequences. Entitlements and responsibilities are two sides of the same coin,” Alan Milburn said.
That is a conditional contract, not an austerity slogan. Its credibility will depend on whether the state delivers the service before it withdraws support.
The Short-Term Fiscal Pressure Is Cyclical; the Labour-Market Barrier Is Structural
Britain faces two different problems, and confusing them would produce bad policy. The pressure to reduce welfare spending is cyclical and political: governments facing tight budgets look for savings that can be booked faster than productivity reforms mature. The underlying rise in health-related inactivity is more structural. It reflects a changed relationship between health, education, employers and the labour market, and it will not reverse automatically when the economic cycle turns.
The scale of the health channel is visible in the government’s Keep Britain Working work. Nearly 3 million working-age people are out of work because of ill health, while nearly 33 million working-age people are in employment. The review says retaining just 1% more of the working population—330,000 people—would add economic capacity comparable to a large city. The estimated annual cost of ill-health inactivity is £212 billion through lost productivity, higher welfare spending and pressure on the National Health Service. These figures are not a forecast of savings from Milburn’s proposals; they show why the potential economic prize is larger than the immediate benefit bill.
The administrative record also shows why shocks and structural problems need to be separated. The official welfare trends report says fraud and error rose from 3.1% of spending in 2019-20 to 4.3% in 2020-21 and 2021-22. That comparison demonstrates how a major shock and changes in administration can affect welfare costs, but it does not prove that health-related inactivity is temporary. The persistence of millions outside work points to a problem in health support, skills and employer retention that a change in the economic cycle may not solve.
Milburn’s approach is structurally more ambitious than a benefits audit because it tries to join benefits to skills, work placement, health support and the demand side of the labour market. It is also slower and harder to measure. An assessment can be counted this quarter. A young person who completes treatment, gains a qualification, enters a job and remains in it for two years cannot.
The mechanism runs through expectations. If a claimant believes that refusing support carries no consequence, conditionality may have little effect. If a claimant believes that accepting support risks losing income before a suitable job exists, conditionality can become a poverty trap. The same rule can therefore raise employment or increase hardship depending on the quality and timing of the alternative.
That is why the fiscal temptation is dangerous. If ministers cut first, they may show lower expenditure before showing higher employment. The apparent saving could then be a transfer to local government, charities, the NHS or households. A policy that lowers the welfare line while increasing other pressures would not necessarily be a fiscal reform; it could simply move the cost to a less visible ledger.
The Second-Order Test Is Whether the State Can Build Capacity Before Claimants Lose It
The conventional reading of the review is that stricter conditions will increase labour supply and reduce benefit spending. The more important second-order question is whether Britain has enough institutional capacity to turn conditionality into productive employment. That requires the review to change the supply of support, not only the behaviour of claimants.
The announced Youth Guarantee investment is a start, but the comparison with the NEET total matters. Up to 50,000 apprenticeship places is meaningful for those who receive them, yet it is a fraction of the 1.012 million 16- to 24-year-olds classified as NEET. The relevant measures are completion, transition into paid work and retention after 6 or 12 months. A programme that counts enrolments can look successful while failing to alter long-term inactivity.
The employment-support system also has to distinguish among groups that currently sit in the same statistical category. A 19-year-old with untreated anxiety, a 23-year-old with a chronic physical condition, a school leaver without a qualification and a young parent facing unaffordable childcare may all be NEET, but they do not respond to the same incentive. The terms of reference’s attention to childhood experience and root causes is therefore economically important. Better targeting can raise the return on spending; blunt conditionality can lower it.
Employers are the missing transmission channel. The state can pay for an adviser, but the final decision to hire or retain belongs to an employer. The Keep Britain Working work refers to the importance of a “receiving environment”, recognising that getting people into work is not enough. If workplaces do not adapt to mental-health, musculoskeletal or cancer-related barriers, claimants could cycle between short employment and benefits. That would make the welfare state appear ineffective when part of the failure lies in job design and retention.
There is also a macroeconomic angle. A credible employment reform could improve the medium-term fiscal outlook by lifting labour supply, earnings and tax receipts. But the first phase requires spending when budgets are constrained. The government expects anti-fraud measures to save £1.4 billion between 2027 and 2032, with total forecast savings from fraud and error reaching £20 billion. Those measures may improve value for money, but they do not fund the same thing as a multi-year employment-support system. Treating administrative savings as proof that deep reform can pay for itself immediately would confuse two different channels.
The expectation gap is political as much as financial. Voters may support the principle of reducing long-term dependency but oppose cuts when they see that disabled people unable to work are affected. Evidence compiled by the Institute for Government shows that 74% of Britons opposed cuts to payments for disabled people unable to work in a 2025 survey. A 2026 poll found 11% believed the benefits system offered too much support to disabled people, against 41% who believed it offered too little. The public may support a clearer deal, but not necessarily a lower safety net for people judged unable to work.
That creates an asymmetry. The political benefit of a successful employment reform arrives later and is diffuse: more people working, healthier communities and stronger tax receipts. The political cost of an early benefit reduction arrives immediately and is concentrated: a household sees its income fall. Unless ministers can show that support is available before entitlement tightens, opponents will define the policy as a cut regardless of its stated architecture.
The Strongest Counter-Thesis: Conditionality Can Become a Cut by Another Name
The strongest challenge is not that welfare should never change. It is that the state may overestimate the speed with which employment support can compensate for lower income. The New Economics Foundation argues that cutting benefits before the returns from employment support are known could deepen poverty and that the evidence base for which interventions produce high-quality, lasting jobs remains limited. This attacks the review’s foundation: if the service is not ready, the “responsibility” side of the deal becomes an enforceable penalty while the “entitlement” side of the deal is withdrawn.
That counter-thesis is strengthened by the timing problem. Milburn’s reported assessment is that savings could take almost a decade. A government with a short parliamentary horizon may therefore have an incentive to claim the investment while prioritising an early reduction in caseload or payment. Even if the long-term design is sound, implementation can turn it into a short-term fiscal exercise.
The answer cannot be to reject conditionality altogether. A system with no expectation of engagement can leave young people disconnected from work and training. But conditionality should be sequenced. First, the state must provide a credible menu: mental-health treatment, skills, apprenticeships, work placements, personal advice and jobs with a realistic route to retention. Second, it must measure outcomes rather than attendance. Third, sanctions should be linked to refusal of a suitable, available intervention, not to the mere fact that a person remains out of work.
The falsifying signal for this article’s judgment is specific: if, within the first two years of implementation, the government increases sanctions or reduces initial payments while employment-support capacity, completed placements and six-month job retention fail to rise, the structural-reform claim is wrong. In that case, the policy would be functioning as retrenchment, not reform. A second warning would be a measurable rise in hardship indicators among affected young claimants without a corresponding fall in long-term inactivity.
There is a counter-risk on the other side. If ministers spend heavily without improving employment outcomes, public support for investment could collapse. The government would then face a backlash from taxpayers and from claimants placed into programmes that do not lead to work. The reform must therefore publish a transparent scorecard: participation, completion, job entry, job quality, retention, health outcomes and fiscal cost per sustained employment outcome.
What Would Prove That Britain Really Wants Reform?
The review will be judged less by its language than by the government’s implementation choices. In the short term, ministers need to prevent a confidence shock among claimants and employers. That means clear transitional rules, accessible advisers and enough places for the conditions to be meaningful. The near-term fiscal effect is unlikely to be determined by the headline announcement alone; it will depend on whether the welfare bill is managed through investment and higher employment or through eligibility and lower payments.
Over the medium term, the key test is sustained employment. The government should report the number of young people moving from benefits into work or training, the share remaining in work after 6 and 12 months, and the proportion receiving health or skills support before any reduction in payment. The 330,000-worker illustration from Keep Britain Working is useful as a scale marker, but it is not a target attributable to Milburn. The real test is whether the system changes outcomes for the people it touches rather than merely reducing the number recorded as claimants.
Over the long term, structural reform would mean that health, education, welfare and employers operate as one employment system. That would require data-sharing with safeguards, local flexibility, stable funding and political patience across more than one budget cycle. It would also require ministers to accept that a higher benefit for people with severe disabilities may be part of reform rather than evidence against it. A system can be more conditional for those who can work and more protective for those who cannot without being either permissive or punitive.
The base case is a partial reform: better employment support and sharper distinctions between groups, but uneven local delivery and political pressure for faster savings. The upside case is a genuine prevention model in which early health and skills support lowers long-term inactivity and the government protects the investment through the next fiscal squeeze. The downside case is a two-speed system in which young claimants face lower initial support before advisers, placements and employers are ready. The trigger separating those cases is not the rhetoric of the review; it is the order in which money, services and sanctions arrive.
Britain’s public debate contains a contradiction. Voters are wary of welfare spending in the abstract but are more protective of disabled people who cannot work. They may accept a deal that makes work more attainable and benefit rules more intelligible. They are unlikely to accept a deal in which the state calls a withdrawal of income an investment whose benefits will appear years later.
“Candidly, it could mean that some people will get less financial support at the beginning of their benefit claim than under the current system,” Milburn said.
That candour makes the test unavoidable. A government that really wants to reform the welfare state must be willing to spend before it saves, distinguish incapacity from disengagement, and publish evidence when the promised employment gains do not arrive. If it is unwilling to do those things, the review will show not that Britain has chosen reform, but that it has chosen the faster politics of a cut.
Explore more exclusive insights at nextfin.ai.
