What’s gone wrong with welfare?

Introduction

Over the last decade, Britain’s welfare system has malfunctioned.

In May 2026, there were 8.4 million people in families receiving Universal Credit (UC); 5.2 million of these were out of work (in November 2025, there were also another 0.3 million on other out-of-work benefits, but some of these will have since moved to UC).

Within this total, here has been a lot of focus on the rise in incapacity claims: since 2016-17, claims for health-related out-of-work benefits (i.e. Employment and Support Allowance or Universal Credit – Health (hereafter, UC-H)) have risen by around a million (to 3.5 million). The claimant count (i.e. those receiving out-of-work benefits and required to actively seeking work) has also risen by 900,000 (to 1.7 million) since mid-2016; most of the rise reflects a methodological change, as Universal Credit replaced Jobseeker’s Allowance, but the claimant count is up by 170,000 since 2022, when the Office for National Statistics decided that the ‘alternative claimant count’ series was no longer needed. And there has been an increase in the numbers who are caring for sick or disabled family members full time by just under 300,000. Altogether, this gives a total increase of some 1.5 to just over 2 million out-of-work claimants since 2016 (depending on how one treats the rise in the claimant count).

Within these out-of-work welfare claims, there are four groups that policy makers have most worried about, either because these groups tend to have long spells of time on welfare, or because they are particularly at risk of adverse effects on future employment chances. These are: those out of work because of health or with a disability, single parents, young people and those with long-term claims for unemployment. The exact turning points have not always been the same, but the trends over the past decade have been negative for all:

• Those out of work because of health or with a disability: up by around 1 million since 2016-17.
• Young people: the employment rate of 16-24-year-olds (excluding those in full-time education) peaked just before the pandemic, at just under 77 per cent (in 2016 it was just over 75 per cent). It has fallen since to 68 per cent, with almost all this decline occurring between June 2022 and June 2024.
• Single parents: the 23-year period from 1996 to December 2019 saw a massive increase in the employment rate of single parents from 44 to almost 70 per cent. This was the great success of policy reform (welfare and childcare) of this era. Since then, it has fallen back to 64 per cent, with the inevitable consequences for child poverty.
• Long-term unemployment: the Department for Work and Pensions (DWP) has stopped publishing regular data on long-term UC claims for unemployment. But the DWP said in January 2024 that nearly 475,000 Searching for Work claimants (almost one-in-three) had been claiming for six or more months, and 320,000 (one-in-five) for at least 12 months (the share of long-term claims always rises a year or so after the overall unemployment rate rises). Given the sustained increase in the claimant count of nearly 1 million since 2016, the share with long-term claims is highly likely to have risen.

Of course, there are important overlaps between these groups. The Milburn review into NEETs is concerned about the overlap between young people and incapacity benefits. And young adults make up a disproportionate share of the long-term unemployed, with around one-fifth of unemployed 18-24-year-olds having been unemployed for at least 12 months over the last two years (peaking at just over a quarter in mid-2025). In addition, a significant portion of rising single parent worklessness reflects rising ill-health.

Where did it all go wrong?

Some of the increase in claims of out-of-work benefits reflects the increases to the State Pension Age (since 2018, this is estimated to have added nearly 100,000 to claims for incapacity benefits), and the broader ageing of the population has also pushed up out-of-work benefit claims.

But the majority of the upward trend is, on the face of it, surprising given that these large rises in welfare claims came in a period of slightly rising employment rate. Since mid-2016, the employment of working-age people has risen by over 2 million people (on all three measures of employment we have available) or by nearly 1 percentage point of the population (although employment rates peaked before the pandemic, in December 2018).

One plausible story is that we have had a rising employment rate among groups less likely to receive out-of-work social security benefits if they were not in work, and falling employment rates among those who are more likely to. For example, between 2002 and 2021, the employment rate of women with dependent children and living with their partner rose by nearly 10 percentage points (or 800,000). This has been accompanied by a decline in early retirement (the average age of labour market exit has risen by 2.8 years since 1996 for men and 4.5 years for women), such that the employment rate of those aged 50-64 has been steadily rising since the mid-1990s, from 57 per cent in 1995 to 72 per cent in 2026 (despite more also claiming health-related benefits). In the other direction, the fall-off in entry-level jobs for the young in retail and hospitality since 2022 has had major adverse effects on the youth labour market, and I have already noted the falling employment rate among single parents.

However, the deeper conclusion is that something has gone wrong with our welfare system. This was a reversal of the preceding two decades where these same groups saw an improving picture. The huge increase in single parent employment rates has already been noted. The share of the unemployed who were out of work for more than 12 months fell from just under 39 per cent in the mid-1990s to 27 per cent by 2016, and total unemployment also fell. The employment rate of young people not in full-time education was on a rising trend through the first half of the 2010s, and claims for out-of-work health benefits fell slowly but steadily.

The period of improvement from 1996 to 2016 – spanning Labour and Conservative-led governments – saw the use of four strategies to reduce welfare reliance:

1. Conditions on welfare receipt that required search activity or preparations to do so. This started with the move to Jobseeker’s Allowance in 1996, but was extended to single parents and some disabled claimants later.
2. Strong financial incentives to work, focused on part-time work for those with significant barriers to working, such as single parents and disabled people. This was the central focus of the tax credit system introduced in 1999.
3. Case workers or work coaches supporting job search activity. This engagement was more intensive for harder-to-help groups, such as single parents and the long-term unemployed, but not extended to the disabled unless they volunteered to engage.
4. Job placements and employment support, which were often limited to harder-to-help groups and scaled up in periods of recession; the New Deal programmes and the Future Jobs Fund programmes are the main examples.

Universal Credit came into force for all new claims in 2018, and existing claims began to be migrated over shortly after that. Universal Credit integrates all of the old means-tested benefits (except Council Tax Support) into one system. Because it integrates Housing Benefit with tax credits when working, it produces a single withdrawal rate and a single benefit rather than a complex system of overlapping tapers from programmes that interacted with each other and were operated by different parts of government. It also has a smoother pattern of financial gains to working, rather than jumps at 16 and 30 hours when tax credit bonuses kicked in under the previous regime.

But the implementation of the new regime coincided with large cuts in the generosity of welfare benefits under the Conservative-led governments’ austerity-driven approach. Importantly here, this need to reduce spending meant that the general work allowance originally envisaged by Iain Duncan Smith – where earnings do not cause a cut in UC entitlement – was limited to those with children or a claim for health-related incapacity, thereby weakening financial incentives to move into work. In addition, conditions for job search when unemployed were ramped up under the Intensive Work Search Regime and the use of sanctioning exploded.

As a result, for those without children, the system became very binary. There was a powerful administrative pressure to search for 35 hours a week for full-time work, and an incapacity benefit regime with no job search rules and far greater generosity to those out of work (in 2024-25, someone out of work due to ill health received more than twice as much as someone unemployed). For those with an impairment, there was no in-between space if 35 hours a week job search or full-time work were not viable. And in general, incentives to work part time within UC are low for renters without children as there is no work allowance.

Job search support for the unemployed became largely limited to monitoring search activity and assessing the application of sanctions, not practical help. Likewise, job programmes were limited to periods when the labour market was weak: Kickstart, for example, ran through the pandemic but ended in 2022. There have been some successful programmes for those with poor mental health (for example, Individual Placement and Support, Connect to Work and WorkWell) but they have, until recently, been small scale. And in general, the UK spends much less on employment programmes than some of our European peer countries.

But this low-cost strategy of largely relying on conditionality and sanctions to get unemployed claimants into work has clearly failed, as numbers have up since 2016, and the binary nature of our welfare system pushes our ageing and increasingly impaired population to claim incapacity benefits over limited working with in-work financial top-ups.

System Reform

Despite the rapid rise in receipt of out-of-work benefits (whether for unemployment or ill-health), there has been surprisingly little reform since 2016. Caseloads were already rising between 2016 and the Covid-19 pandemic (by around 400,000) but the alarm bells should have been ringing in 2022, when a strong employment recovery still left out-of-work benefit receipt up by a million on 2016. It was also surprising that Labour in opposition developed no critique of the welfare system as a whole, nor major plans for reform (other than its manifesto wanting to “review” Universal Credit and “reform or replace” the Work Capability Assessment), despite their impending inheritance. This stands in contrast with the work done by Gordon Brown ahead of the 1997 election (which then led to New Deals and tax credit reform), and Iain Duncan Smith ahead of the 2010 election, which led to Universal Credit.

Since the election, the Labour Government has instigated two significant reforms. In its 2025 Green Paper, it proposed a step towards a less binary welfare system with the planned abolition of the incapacity benefit test (WCA) and a sole reliance on the Personal Independence Payment (PIP) test (which captures impaired functionality rather than proving incapacity to work), although the status of this is currently unclear. Combined with a new lower-tier Universal Credit rate, this would create a space to recognise limited working as an explicit goal. The Government has not, however, yet introduced any more conditionality, or funded the additional job coaches and supports that could make this a serious third welfare option.

For young people, systems of engagement with practical help, a hiring subsidy at six months of unemployment duration, and job guarantee pilots and job placements at longer durations have been (re-)introduced. This reflects the (belated) realisation that youth employment is in a crisis as bad as after the financial crisis. Further, a review by Alan Milburn is looking at greater preventative measures, especially to halt the direct transition from full-time education to receipt of incapacity or disability benefits. So, Labour appears to be moving towards a major overhaul of youth welfare and education/skills systems.

However, Labour’s record on delivering welfare reforms is not promising. Planned big cuts to PIP eligibility were torpedoed by a backbench rebellion; PIP is still under review, overseen by Stephen Timms MP, but the Government has had to concede that the review’s  recommendations do not need to cut spending relative to the existing, fast-growing, spending forecasts. The generosity of UC has been rebalanced away from UC-H and towards claimants who are not incapacitated, but it’s hard to escape the feeling that the 2025 reforms and the cuts to entitlements are primarily just trying to save money, rather than to reform the failing system design.

Other parties are beginning to outline more far-reaching proposals. I will discuss the options for change in a future blog.