The Leaky Cask: Rethinking Welfare

The Problem

The parlous state of Britain’s public finances are well-known to any who care to look: a national debt of 95% of GDP (the total value of economic output for a year); public sector debt interest of £127bn a year, a tenth of the £1,290.6bn of total managed expenditure; and a budget deficit of £152bn (5.2% of GDP), for the financial year 2024-25.¹

The primary factor is a relentless increase in public spending, which in real terms has increased by a third since 2007-8² (despite the ‘austerity’ years of 2010-17). There are various causes of this — population dynamics, poor productivity and political largesse for popularity — focusing on what we might call ‘social spending’, such as welfare, pensions and health, which together are responsible for some 57% of the increase in real spending since 2007-8 (the remainder is mostly debt interest).³ Indeed, that whole increase on welfare, pensions and health is larger than the present budget deficit.

“The poor are always receiving and always wanting more and more, for such help is like water poured into a leaky cask.” Aristotle, Politics 1320a (VI.5)

Aristotle, the ancient Greek philosopher, writing in the 330s or 320s BC, was not making a moral judgement on ‘the poor’. Rather, he was making an observation on the tendency of social spending in democracies to spiral out of control, due to both the natural inclination of people to expect more of what they have already received and the inevitable temptation of politicians to indulge in such largesse for sake of popularity. Each handout begets the next, for it is always easier to give money away than it is to take it back and, of course, politicians wish to keep their jobs. He proposed helping people buy land or go into a trade, rather than simple handouts — as we might say, ‘getting people into work’ and ‘helping them to help themselves’.

In the UK, this ‘leaky cask’ dynamic has certainly been at play. Think of the pensions triple lock, free childcare allowances, benefits rising faster than wages, bumper pay rises for public sector workers, real health spending increasing (always a popular line) by 50% since 2008 yet performance deteriorating — just to name a few off the top of my head. This is exacerbated by rapid, mass immigration. An immigrant population which, being mostly low-skilled and with, or about to have, children or dependents, will inevitably be a net fiscal cost over their lives, even if working (any fiscal benefit is short-term). Overall, 53% of UK households are net fiscal recipients i.e. they get more from welfare and services than they pay in through tax.⁴

Any effort to restore the UK’s fiscal health requires spending control, in particular over welfare spending. Tax rises are not the solution, as they inherently work to suppress growth which has already been poor at best in the UK for the last 20-odd years (and taxes have increased substantially since 2008 already). Yet there is a marked reluctance by politicians and institutions to be drawn into this, in part because of the risk of unpopularity but also, I suspect, due to mindsets amongst our leaders forged 25 years ago when the system seemed more sustainable.

From here on, I am referring to benefits (e.g. Universal Credit, Personal Independence Payment, etc.), not pensions, when I say ‘welfare’. For all the attention pensions garner, they are less of an immediate problem than the benefits system. Abolishing the triple lock, while logically sensible and somewhat of a cause célèbre, makes little short-term difference to pension spending, and is by far the biggest policy reform on the table. Moreover, pensions do not interact with work and incentives to work, which are particular problems with welfare currently.

There are two approaches to reducing welfare spending: reducing the size (the amount of money) of claims or reducing their number. The former often seems attractive, particularly to some parts of the political Right and is certainly part of the solution, but misses the main issue, which is with the latter; namely, structural deficiencies in the welfare system that disincentivise working for lower pay and its excessive inclusivity (ease of claiming).

These are my ideas on correcting just that, focusing on the two main — and most problematic — benefits: Universal Credit (UC) and Personal Independence Payment (PIP), alongside some whole-system reforms. They are intended as realistic, pragmatic reforms, aiming at making the current system work better (i.e. what a prospective new government would be able to do relatively quickly), rather than a total reconstruction or demolition of the welfare system, which I suspect there is little public appetite for as the contributory principle is deeply embedded in the British body politic.

Principles for Welfare

Before looking at the specifics though, let us first consider what the point of the welfare system is. DWP’s stated vision is ‘to improve people’s quality of life, now and in the future’, buttressed by goals framed around caring, valuing everybody, enabling people into work and pursuing a just, equal and inclusive society.⁵

Leaving the poor grammar aside, we can see this is an inherently maximalist and expansionist mandate. There is no mention of who should be eligible for welfare and why, only trite niceties that dissolve upon inspection: ‘we care’, ‘we value everybody’, ‘enable to get into work’, ‘pursue a just, equal and inclusive society’, and so on. Welfare is effectively unbounded: ‘ensuring such-and-such for all’. It is this mindset — that the welfare system is a vehicle through which to fix society and exists to serve recipients — that begets its endless expansion. There is always another person ‘in need’, or someone who could use ‘a little extra help’. It never ends; hence the leaky cask.

Welfare was envisaged as a (temporary) safety net for those who had fallen on hard times or, through no fault of their own, were unable to work. This is perfectly reasonable and attracts little controversy and is exactly what we should re-orientate the welfare system back towards. It ought not to be a lifestyle subsidy or a replacement for work, or a ready source of income to be exploited by immigrants, which it has become. My principles for a renewed welfare system would be:

• In the first instance, only British citizens should be able to claim welfare of any kind (with some nuance).
• Income from benefits should not be treated differently to income from any other sources i.e. it should be subject to income tax (currently it is usually not, no matter how large).
• Access to benefits should be tied to inherent need, and last only so long as those needs persist, not to life decisions or inconveniences.

These articulate a different idea of welfare: restricted, treated equally with work and a safety net of last resort. It reconciles the State’s accepted role in providing a safety net to its people but not in trying to fix every problem in people’s lives or allowing them to live as comfortable a life as possible.

I emphasise these because reality is complex and when trying to reform welfare one will inevitably encounter moments of competing and contradicting objectives, or where unpleasant trade-offs have to be digested. Having clear directional principles in mind allows these conflicts to be resolved. DWP’s present vision and goals do not allow this; how can one balance competing intents or make difficult trade-offs when the vision and goals of welfare require you to be everything to everyone? This is what sank the Government’s aborted disability benefit reforms: it could not reconcile the reality of out-of-control spending with its stated vision and goals (which are the views of most of its MPs) of ‘ensuring independence and control for all disabled people’.

These changes would, by design, mean many present welfare recipients losing out. But the present system is dysfunctional and plainly out of control, having lost sight of its original purpose (as above). This is a pain that must be borne. It is not a war on welfare but rather a rediscovery of what welfare should be: a way to help people help themselves, rather than a subsidy which traps people. Where there is an inherent need, there is a role for the State; where there is not, there is not. That boundary must be carefully watched and applied with a dose of tough love.

Now for the specifics.

Benefits for citizens

It does not strike me as controversial or harsh to argue that only British citizens ought to be able to claim benefits in the first instance. Benefits are a safety net for sudden hardships and those incapable of work; it is not Britain’s job to be a charity to the world and all its people, providing succour to all and sundry.

DWP publishes some data on the immigration status of Universal Credit (the main benefit) claimants.⁶

15% of UC claimants are neither UK/Irish citizens nor have Right of Abode (a historic immigration status for certain Commonwealth citizens prior to 1983).

Some 224,000 people with Indefinite Leave to Remain (ILR) are claiming UC (and potentially other benefits), amounting to between c.27-36% of the approximately 622-820,000 persons with ILR status (excluding settled EU nationals under the EU Settlement Scheme).⁷ That is a strikingly high proportion, though some of it may be caused by DWP data lagging real-time immigration status e.g. they have since acquired citizenship, as it is only recorded when the claim for UC is made or updated.

Even if there is a lag in the data, however, there is clearly a steady — and relatively sizeable — stream of individuals with ILR claiming benefits, and then potentially going on to be citizens, presumably still on benefits. It begs the question of who these people are and why exactly they are in the UK; it is certainly not the British government’s responsibility to see them looked after. There is a ring of the ‘leaky cask’ about this. As a first step, one ought to make ILR persons ineligible to claim welfare. This is also important given the likely impending surge in ILR numbers as ‘Boriswave’ immigrants become eligible — wider immigration reforms to ILR and citizenship are also necessary.

Limited Leave to Remain (LLR) persons are normally ineligible for welfare due to No Recourse to Public Funds (NRPF) conditions. In some cases exceptions are granted and people can apply to have NRPF conditions lifted too. Given these people are not refugees or humanitarian cases, it seems strange that there are any such claimants at all. My instinct here is to disqualify all LLR persons from claiming welfare without exception; again, it is not the British government’s responsibility to see to the welfare of foreigners here by their own choice.

It is unreasonable to expect Humanitarian (e.g. Ukraine Scheme visas) and Refugee persons (the latter are individuals granted asylum) to be able to enter work immediately. Nonetheless, they should have conditions or time limits attached to their welfare grants; say, they could claim for two years after being given their status. Once again, it is not the duty of the government to provide a perpetual income to people who are capable of working simply because of their origin. Separately, of course, there needs to be wholesale reform of the asylum system to control this as a route into the UK in the first place.

The EU Settlement Scheme (EUSS) governs the immigration status of EU nationals resident in the UK prior to its withdrawal from the EU and is functionally similar to ILR, having their access to welfare guaranteed by the UK-EU Withdrawal Agreement. There is growing evidence that this group is not all that it would seem, with large numbers of non-EEA nationals managing to claim EUSS and in turn access benefits. In an absolute legal sense specific conditions could be imposed, though at risk of likely dispute with the EU. Some 57% of EUSS UC claimants are in work (defined as employees, self-employed, and both), much higher than the UK/Irish/RoA equivalent of 36%⁸ but the reality stands that these people are not citizens and the EUSS is increasingly being taken advantage of as a route to the UK. My preference would be to apply the same rules to EUSS as ILR, which is to say no benefits access, and accept the same for UK citizens living in the EU. These nationals can always return to their home country if this is a problem for them.

There is no published breakdown of UC spend, amounting to £70bn in 2024-25,⁹ by immigration status, so it is not possible to estimate the savings from these proposed reforms. Presumably though, given their proportion of the claimant count, the savings would be in the order of multiple billions per year (high single digits).

Personal Independence Payment (PIP), spending on which totalled £27.3bn in 2024-25¹⁰ (the second largest single benefit after UC), does not have data published on the nationality or immigration status of claimants. There is, however, an overlap between the Health element of UC and PIP claimants (more on this below). About 42% of UC claimants are claiming the health element (3.5m)¹¹ and thus may also be claiming PIP, so disqualifying non-citizens from benefits would certainly reduce the PIP spend too.

Benefits are income

Presently, most benefits (e.g. UC and PIP) are untaxed, not counting towards taxable income. This permits a situation where individuals can receive a relatively large income from benefits, more than the average worker earns after the worker has been taxed. It strikes me as odd that benefits income is treated so favourably compared to earned income; that disparity incentivises maximising benefits income and, where the choice exists, choosing benefits over work. That is Aristotle’s ‘leaky cask’ once more.

There is no rational, moral or economic basis for this distinction. Income is income, regardless of whether it is received from benefits or earned through work, and should be treated — taxed — accordingly. Pensions are largely taxed like salaries already, for example. Those who rail against the tax regime for the unearned income of the very wealthy should also be on board with this, since the principle is the same: income should not be treated differently by virtue of its source (the UK tax system is egregious for this tendency).

Requiring income tax to be paid on benefits income (i.e. income tax is due on all benefits income above the Personal Allowance) would end its privileged status and remove part of the existing financial disincentive to earn rather than claim at the lower end of the income distribution. That disincentive is about time as much as simple money numbers; time is money.

If you could receive £20,000 a year in benefits — a Universal Credit claim could surpass that easily enough — versus working full-time 40hrs a week at minimum wage for about £26,450 a year gross, a rational person would be hard-pressed to choose work. The latter then being taxed while the former is not eliminates any economic incentive to work: after tax (income tax and national insurance), the worker’s £26,450 earnings a year is reduced to about £22,550 net, just £45-odd a week more than being on UC. 40hrs of free time a week (before other costs e.g. travel) is more valuable than £45, no doubt. If the benefit income were subject to income tax, the worker would be £80-odd a week better off than the UC claimant. Still not great, but better and fairer.

Income is income, regardless of whether it is received from benefits or earned through work. There is no rational, moral or economic basis for the distinction.

Taxing benefits as income would also cancel out (by raising corresponding revenue) a portion of the total benefits spend. To illustrate, the average UC claim was £1,030 a month in November 2025,¹² amounting to £12,360 a year, a smidgen below the Personal Allowance. For parsimony then, let us say that half of the UC spend of £70bn in 2024-25 is to households whose total UC income is above the Personal Allowance; that would mean about £35bn of UC income would be subject to income tax. Say it were all taxed at the basic rate of 20%, like earned income, that would recoup a few billion pounds. In practice, it would likely be higher as the distribution of UC claims is probably right-skewed i.e. there are a relatively smaller number of large claims which dominate spending. Accounting for all other benefits now being taxable, there might be another couple of billion raised.

Universal Credit

Fixing welfare requires dragons to be slain, so let us turn now to the ‘big beast’: Universal Credit (UC). UC is the primary benefits vehicle, with 8.4m claimants as of January 2026 and accounting for some £70bn of spending in 2024-25. It combines and replaces several legacy benefits as ‘elements’:

• the Standard Allowance (the basic payment);
• Housing (replacing Housing Benefit);
• Child (top-up for children, per child);
• Health (top-up for being unable to work due to ill health e.g. the Limited Capability for Work and Work-Related Activity, LCWRA);
• Carer (for carers); and
• Childcare (covers a portion of the costs of childcare for working parents).

There is an immediate problem with this design: while the legacy benefits were complex through their number and different rules, they each had a clear intention. What is the intention of UC? What is the specific need it is aiming to alleviate? It is sort of an unemployment benefit, but with add-ons and top-ups, like health, caring and childcare, that are separate to unemployment (more relating to economic inactivity). Some of these elements overlap with other benefits, such as the Carer’s Allowance and PIP (for the Health element), which can be claimed alongside it. It is also complex, with the different elements having their own rules and qualifications.

UC ought to straightforwardly be unemployment benefit; the ‘inherent need’ is survival during a period of unemployment. It should not be ‘stackable’ with other benefits serving different ends. To that end, I would end the overlapping benefits. Disability or ill-health would no longer be part of UC, so the Health element would be abolished, instead being dealt with under the PIP regime (more on this below). I would also abolish the Carer element, increasing the Carer’s Allowance to compensate, and making the two benefits mutually exclusive. Put simply, people on UC due to disability or with caring responsibilities would be claiming specific benefits for those circumstances rather than UC.

Turning next to the Housing element, which replaces Housing Benefit and is fiendishly complex, with payments varying by locale, number of unoccupied bedrooms (the ‘bedroom tax’) and non-dependent adults living in the property. I would replace it with a single, uniform top-up payment for those who are privately renting. So, a claimant would by default get the Standard Allowance and then, if renting, would get the Housing top-up, which would be a flat payment regardless of where they live.

These reforms would disadvantage those with high rents and council tax, principally in certain city areas. But recalling the principles, it is not the responsibility of the government to subsidise someone to live in a particular part of a city. There is nothing unfair about this; why should a person on benefits be able to afford to live in a location while someone in work cannot? There is no reason why the government should be paying for someone to live in, say, central London.

What of the Childcare element? This speaks to the next issue with UC: its dysfunctional interaction with work. Currently, claimants are permitted to claim UC and work, with about 38% of claimants being in-work as of February 2026. Net income (after tax, national insurance and pension contributions) earned from work counts towards a ‘work allowance’, a fixed amount per month which is dependent on whether a claimant receives the Housing payment or not (lower if they do). For a household with Housing payments, the work allowance is £427 a month, amounting to about 8hrs of work a week on minimum wage. All income from work above the work allowance then reduces the UC payment received by 55p per £1 — this is the ‘taper’.

To see the effect of this, let us imagine three individuals, A, B and C, who each have a UC award of £1,200 a month, including housing, giving a monthly work allowance of £427. A and B both work as employees on minimum wage, while C does not work. A works 30hrs a week, earning £1,650 gross a month on average, while B works 15hrs a week, earning £825 a month on average. A will earn about £19,800 gross a year, so will have to pay income tax (20%) and national insurance (8%). The monthly incomes of these three, after tax and taper deductions as relevant, would be about:¹³

• For A: £2,100.
• For B: £1,810.
• For C: £1,200 (their monthly UC payment).

So, A works twice as much as B but is not seeing even a £300 a month uplift from doing so. Basic rate income tax and national insurance at 20p and 8p in the pound, and then the UC taper applied at 55p in the pound on the remainder means that for every additional £1 earned by A, they lose about 68p — an effective marginal tax rate of 68%. That is before other factors e.g. Council Tax Reduction, which tapers as well so would push the effective marginal rate even higher.

For every additional pound A earns, they keep 32p. That makes the ‘£100k tax trap’ look generous.

The UC taper strongly disincentivises working to earn more than one’s Personal Allowance i.e. when income tax and national insurance kicks in. Rationally, people want to maximise their net income while minimising the amount of effort required for it (remember, time is money). There is little incentive to work at a 68%+ tax rate. UC therefore traps people half-in, half-out of work, adding to the benefit spend while contributing minimal tax revenue. Once again, the leaky cask in action.

Tinkering with the taper rate would merely shift the income maximisation point, maintaining the incentive to keep claiming UC, and not solve the problem of it disincentivising work. I would abolish it and end the ability to claim UC while in work at all, in line with it being a safety net for people who find themselves unemployed. This would also render the Childcare element redundant. Alongside that, I would abolish Council Tax Reduction/Support (a reduction in Council Tax) for UC claimants, which is administered separately by councils, as it also tapers by earned income and so further distorts incentives to work.

Taken together, these proposals would cap UC claims. Presently, total benefit claims are capped, per the below, but this does not apply if the claimant has been judged as LCWRA, receives PIP, or is a carer (amongst other things); through these exemptions, it mainly caps UC claims of the conventional unemployed or part-time workers on UC.

Some 42% of UC claimants (3.5m) are in receipt of the Health element, meaning their claims are uncapped. Shifting these claimants onto targeted support (a reformed PIP, outlined below), or them forgoing it (depending on the nature of their conditions) would make this cap more effectively applicable to UC. In turn, that allows UC to be properly calibrated to incentivise work by ensuring that e.g. full-time work on minimum wage would be financially much better (after tax, which as above would apply to benefits income too). Indeed, the present caps should achieve this objective easily enough. Let us imagine four households, W, X, Y and Z, and have the above caps as the uniform payments I proposed above (all figures gross):

• W: single, no children, in London, on UC and receives £1,413.92 a month.
• X: single, no children, in London, works at minimum wage 40hrs a week, earning £2,200 a month.
• Y: couple, with children, in London, on UC and receives £2,110.25 a month.
• Z: couple, with children, in London. One works 40hrs a week at minimum wage, earning £2,200 a month, the other works 15hrs a week at minimum wage, earning £825 a month.¹⁴

These households represent common setups in the real world. I chose the minimum wage to create an in extremis comparator for a lower-pay worker, where the trade-off with going on benefits would be most pronounced. They would have net earnings (accounting for benefits being subject to income tax) each of:

Under the present system (benefits untaxed), W would, in addition to receiving UC, be able to work up to their monthly £427 work allowance (about 8hrs a week at minimum wage, remember) without losing a penny, which would bring their income almost level to X who is working full-time (40hrs). That plainly disincentivises work, especially if they can also claim PIP or other benefits. For Y and Z, some gap is maintained — presently, Y can reach £2,110.25 + £427 = £2,537.25 before the taper kicks in — but one of less than £200, and an extra £200 a month is probably not worth the hassle of both parents working for (40hrs plus 15hrs versus 8hrs a week).

Without this work allowance, plus the fair taxation of benefit income, there would be a clear gap between income from full-time working and claiming UC (about £540/month for W and X, about £800/month for Y and Z). Full-time work would thus clearly be a financially better choice than claiming UC, creating the incentive to get back into work, while allowing an unemployed person to live, as is the benefit’s intention. They would not be able to live a particularly comfortable lifestyle, true, but that is not the intention of the benefit; to do so would require a benefit as large as a full-time salary, at which point there is no incentive to work and we are talking about a universal basic income, not an unemployment benefit.

Those who are presently receiving UC in-work (38%) would be the most adversely affected by this change, being forced to either seek full-time work or solely receive UC (and thus likely reduce their income). This is necessary to correct the perverse incentives of the present system which traps people by incentivising taking a small amount of work to maximise income and not go full-time, or top-up their UC with other benefits e.g. PIP.

I would also explore time limitations to claiming UC through local public works programmes. These programmes would be mandatory for long-term UC claimants (say, after 18 months), and involve basic but important work: street and park cleaning, gardening, maintenance and so on. Improving the state of public spaces is valuable in and of itself, and the work would be beneficial to the long-term claimants doing it.

A final point on UC. Getting people into work requires there to be jobs for them to fill, especially at the entry-level of the labour market — jobs in pubs, shops, supermarkets and reception desks. These are the very jobs which have been most overwhelmed by immigration and suffocated by escalating business costs, such as increases to employer’s national insurance, above-inflation minimum wage rises, high energy prices and regulatory burdens. Fixing welfare to incentivise working also means fixing the dysfunctions in the labour market caused by the government.

PIP

PIP was introduced as the new primary disability benefit in 2013 and, in terms of spend, is the second largest single benefit, with £27.3bn of spending in 2024-25, after UC. The equivalent benefit is devolved in Scotland, so all figures and proposals here are exclusive of that. Spending on PIP, in real terms, increased by about 60% between 2019-20 and 2024-25¹⁵ and the number of PIP claims has risen from 2.7m in May 2021 to 4m in April 2026.¹⁶ Under ‘normal rules’, would-be claimants are assessed to two scales, Daily Living and Mobility, and can receive either standard or enhanced awards for each (and both). There is a separate process for terminally-ill persons.

There is clearly an issue with the present assessment regime for PIP, given it has allowed such an explosion of claims and high rates of Enhanced awards. The answer to this is not more assessments. The present assessment regime has manifestly failed; I am sceptical that adding more bureaucracy (with associated costs) will achieve much. Rather, a re-think of what PIP is addressing is required.

There are no easy solutions here, as this is the heart of the problem with PIP. Instinctively, Enhanced awards ought to be explicitly reserved for disabilities which make it impossible to work i.e. you are either on Enhanced PIP or UC, not both; they are mutually exclusive, reflecting that PIP is for disability and UC is for unemployment. As Enhanced awards would now have to be lived off, they would be increased (to be on par with UC).

I would end the Daily Living/Mobility split in PIP and just have a Standard award band for conditions where there are some needs which are not covered elsewhere e.g. by the NHS or provided by Local Authorities. Most acute needs are already covered through these channels. The payment would be structured to trend lower than the present Standard awards (£76.70 and £30.30 per week for Daily Living and Mobility respectively, so £107 for both), perhaps capping out at around £80 per week. This would be claimable alongside UC. In the end, this is a question of what is government’s responsibility; life is, in the end, not fair, and it is not government’s responsibility to correct — certainly not through handouts — every instance of that.

Also, per the principle of benefits being income, all PIP awards would be treated as taxable income (as with UC and other benefits).

Alongside this broader reform, I propose excluding specific conditions from being eligible for PIP. We can see that psychiatric disorders are by far the largest group of claimants and have high rates of enhanced awards. The number of awards for psychiatric disorders increased from c.980,000 in May 2021 to 1.56m in April 2026, an increase of three-fifths. 78% of these claimants fall into one of: mixed anxiety and depressive disorders, mood disorders, hyperkinetic disorder (i.e. ADHD), autism spectrum disorders and anxiety disorders. Most of these awards are short-term (up to two years), with a review over whether they will continue, though thus far three-quarters of reviews result in the award being maintained or increased.

Digging into this in more detail, we can see explosive growth in claims over May 2021 to April 2026¹⁷ for:

• Stress reactions, which doubled to 68,000 claims;
• Anxiety disorders, which more than doubled to 69,000 claims;
• Mixed anxiety and depressive disorders, which increased by 77% to 435,000 claims;
• Autistic spectrum disorders, which more than doubled to 259,000 claims; and
• Hyperkinetic disorder (ADHD), which almost tripled to 100,000 claims.

Together, the growth in these five claim grounds (out of the 20 in this group) accounts for 90% of the growth in psychiatric disorder PIP claims since May 2021. This is plainly unsustainable and, frankly, not believable. Yes, the lockdowns during Covid messed with a lot of people’s heads — mine included — but in the overwhelming majority of cases I do not believe it generated ‘disabilities’ that necessitate receipt of benefits. Part of the growth, particularly in autism spectrum disorders and hyperkinetic disorder, comes from young people moving from Disability Living Allowance onto PIP at 16, but not all. Moreover, these conditions (with the partial exception of autism) are largely self-reported and thus difficult to authoritatively diagnose; inevitably there will be an incentive to exaggerate or embellish, especially when money is on the line. In addition, these conditions do not, in the great majority of cases, prevent one from working.

Looking at the award spread:¹⁷

This is mind-boggling. How stress or anxiety qualify for enhanced Mobility awards at all, for example, I cannot fathom. Yes, other conditions are taken into account in assessments, but this data records the main disability of the claim.

Stress, anxiety, depression and ADHD (hyperkinetic disorder) should not qualify for PIP at all; returning to the principles, there is no inherent need for government support (in the form of benefits) here as these people are capable of working. They should be explicitly disqualified. Autism is a broad spectrum, with most autistic people perfectly capable of working and living ordinary lives and therefore not needing either component, while more severe cases do require support. These severe cases are medically identifiable; how severe the PIP cases are is unknown. My instinct, given that these claims have doubled in just five years, is that there is a padding of less severe cases inflating the claimant count.

We can estimate the savings from such changes. If all stress, anxiety, depression and ADHD awards were ended, then that would save about £4.7bn a year, a little over a sixth of the total PIP spend.¹⁸ Reducing autism awards, which are mostly enhanced, would save a lot too; to illustrate, reducing claims back to their May 2021 figure would save in the region of £1bn a year.

A similar process of restricting what conditions are eligible for PIP could be applied to other disability groups, such as musculoskeletal disease (general and regional), which together account for 26% of assessments since 2021, the second largest after psychiatric disorders. The objective here would likewise be to trim claimant numbers to disabilities where there is a serious need.

There will of course be losers from these changes, probably a significant number. It is, however, absurd to suggest that 1.3m people have become newly life-inhibitingly disabled in the last five years, certainly to the point where they are unable to work and require perpetual support from the government. Keeping the principles in our mind, it is not the purpose of government to support people through all of life’s challenges.

Inevitably there will also be edge and complex cases where someone narrowly falls into the Standard award. This is an unfortunate fact of life. Exceptions and special cases have a tendency to creep and expand, for there is always another edge case. Better a general system, accepting a proportionately small number of edge cases, than a system degraded, compromised and inflated by exceptions and special cases.

Uprating

Some benefits are automatically uprated by inflation by law, such as PIP, while others like UC conventionally are but are not legally required to be. There is thus a ratchet effect to benefits spending; if inflation is 3%, then benefits spending effectively increases by 3% before anything else has happened. It also immunises those on benefits from the effects of inflation. Workers have no such privilege; if prices increase, that erodes the purchasing power of their wages. They do not have the guarantee that (nominal) wages will increase to offset that.

I propose shifting the default uprating regime for all benefits to a ‘wage link’, whereby they uprate by the rate of inflation or nominal wage growth, whichever is lower. That ensures benefits income is not advantaged relative to wages and serves to maintain an incentive gap between earnings from benefits versus work. A wage link would also allow benefits to be ‘downrated’ if nominal wages fall, though that is a rare occurrence. Aside from fairness, wage-linked benefits uprating would control the ratchet effect on benefits spending, especially in times of persistent inflation as has been the case in recent years.

Concluding thoughts

These proposals would not fix the welfare system in one fell swoop. There remain some imperfections, such as precisely how to address benefits access for EUSS persons (given the legal complexities vis-à-vis the EU) and the assessment and award classification process for PIP. They should, though, resolve the most acute and pressing failures of the present system, which for a new government would be ‘enough’ and be a critical step.

It is not possible to precisely estimate an overall savings figure for these reforms as they interact with each other and involve shifting claimants between benefits (e.g. from UC to solely PIP and vice versa). That said, looking at the orders of magnitude of the prospective savings for each reform, and potential revenues raised from taxing benefits as income, I would roughly put the total savings at around £20-30bn per year. There would likely be additional savings over time from the uprating reform and the re-incentivising of work causing people to come off benefits.

Lastly, welfare is a prime example of how the structural failures besetting the UK intersect and sustain each other. Benefits, the labour market and immigration, for example, are inseparable; these welfare reform proposals would be much less effective if the suffocation of the labour market by rising costs and regulations, and the simultaneous immigration influx into entry-level jobs, are not also tackled.

A prospective government wishing to ‘fix’ the UK would have to be fleet of foot and broad of mind, not getting bogged down and not letting the perfect be the enemy of the good. Fixing, say, three-quarters of the problem with welfare in one or two years is much better than trying to fix all of it at once and coming unstuck in the process. Doing so would shift welfare onto a more sustainable trajectory and create a fairer system for workers and recipients alike.