Welfare

A wealthy nation should provide a safety net for those who, through misfortune or genuine incapacity, cannot provide for themselves. That is not in question. What is in question is what the safety net is for. Its purpose is to catch people when they fall and to help them stand again. It is not a permanent income, and it is not a lifestyle. When it becomes either of those things, it fails the people it supports, and it corrodes the trust of those who pay for it.

Universal Credit, introduced after 2012, is the right architecture for delivering this. It collapses six overlapping benefits into one, removes the worst of the cliff edges between welfare and work, and is structured to make every additional hour of work pay something. The vision behind it — that the system should always pull toward employment, never away — is sound and should be preserved.

The reform proposed here is therefore not the abolition of UC, nor a return to what came before. It is the reform of what has been loaded onto UC and around it: the conditions, the duration, the treatment of contributors, the assessment of incapacity, and the architecture for those who cannot work at all. These are the matters that determine whether welfare remains a bridge or becomes a dwelling.

Today's working-age welfare is almost entirely needs-based. Universal Credit pays the same standard allowance whether the claimant has worked for thirty years or never. The old Beveridge contribution-based benefits — contribution-based JSA, contribution-based ESA — still exist in name, but pay broadly the same as their means-tested counterparts and run out after six months. In practical effect the contributory principle is dead at the working-age level.

This is a failure of design. Consider the case of someone who has spent twenty years in employment, paid substantial tax and national insurance, and then loses their job. Under the present system they receive the same standard allowance as someone who has never worked, and they are subjected to the same regime of mandatory check-ins and proof of job search. The first part is unjust; the second is wasteful. A long-term contributor is highly likely to find new work without prompting. The state's machinery of conditionality is being expended on people who do not need it, while the signal of recognition for past contribution has been stripped out entirely.

A serious contributory element should be restored on all three dimensions. Rate: contributors receive a higher initial benefit, related in some defined way to their previous earnings, for a defined period. Duration: contributors enjoy a longer initial window before conditionality fully bites and before standard time pressures apply. Conditions: contributors face lighter check-in requirements during that initial window, on the principle that a track record of work is itself evidence that they will return to work without state nagging.

After the initial contributory window expires, all claimants converge onto the standard rate with full conditionality. The contributory element is a bridge from the working life back into work, not a permanent enhancement.

Hard time limits, of the kind used in some American states under which benefits simply stop after a fixed number of years, are unattractive. They produce real destitution, they push people into the informal economy or into reliance on family that may not exist, and they ignore the variation in how long it actually takes someone to find sustainable work. The British safety net should not have cliff edges of this kind.

But open-ended welfare without rising expectations is the failure mode the present system has drifted into. The answer is conditionality that ramps over time. In the first months of a claim, support is light-touch and the expectation is straightforward: look for work, accept reasonable offers. After a defined period — perhaps six months — the conditions tighten. Mandatory training in identified skill gaps. After a year, mandatory work experience or community work activity in exchange for the standard allowance. The principle is that welfare is paid in exchange for engagement, and the price of engagement rises the longer the claim continues.

Sanctions exist for non-engagement, not for duration. A claimant who turns up, takes the training, does the work activity and is still looking for paid employment retains support indefinitely. A claimant who refuses to engage loses it. The choice is offered honestly and the consequences follow.

The growth in incapacity-related claims over the last decade has not been driven by physical illness. It has been driven, overwhelmingly, by mental health diagnoses — anxiety, depression, stress-related conditions. Some of this growth is real and reflects better recognition of conditions previously hidden. Much of it is not. It reflects the steady medicalisation of ordinary human difficulty: hardship, disappointment, the wear of life. These things are real, but they are not the same as incapacity, and the welfare system has lost the ability to tell them apart.

The architecture should distinguish more sharply between conditions that genuinely prevent work and conditions that complicate it. Severe mental illness — psychosis, severe and enduring bipolar disorder, conditions requiring sustained clinical intervention — is incapacity and should be treated as such, with unconditional support and infrequent review. Common mental health conditions — anxiety, mild to moderate depression, stress-related presentations — are most often managed alongside some form of work, and the welfare system should structure its support to assume that this is so.

Two practical reforms follow. First, the independent assessment for incapacity benefit should be tougher, with the standard of evidence raised, and conducted by clinicians who do not work for the welfare department. Second, for common mental health conditions, support should be conditional on engagement with treatment — therapy, structured exercise, graduated work activity — rather than offered as an indefinite alternative to work. People who are unwell should be helped to recover, not parked.

A meaningful number of people cannot work at all and never will. Profound learning difficulty, severe physical disability from birth, conditions that progress beyond the possibility of any employment: this is the population for whom welfare is not a bridge but a permanent provision, and the architecture must say so honestly.

For these claimants, the same Universal Credit architecture applies but with no work-search conditions once assessment is complete. Review is infrequent — every five years rather than every six months — to avoid forcing people who plainly cannot work to repeatedly prove what is obvious. Sanctions do not apply. The claim is unconditional in everything but the fact of the underlying condition being verified.

The proportion of young people not in education, employment or training has risen to levels that should embarrass a serious country. The diagnosis matters because the policy answer depends on it, and the population is not homogeneous. There are two cohorts, and welfare reform must distinguish them.

The first cohort is not trying. Expectations at home are low; school has not prepared them; the cultural ladder from idleness to engagement has been removed. For this group the answer is full conditionality from the start of any claim — mandatory training, mandatory work experience, sanctions for non-engagement. The state's job is to set the terms clearly and apply them firmly. Welfare must not pretend that nothing is being asked.

The second cohort is trying and failing. There are willing young people who cannot find entry-level work, and no amount of sanctions will produce jobs that the economy is not creating. The structural answer here is not a welfare programme. It is the growth conditions set by tax, regulation and planning — the matters addressed in Item VII and which belong properly to economic policy rather than to a welfare paper. A country that suppresses growth cannot then ask its welfare system to absorb the unemployed it has created.

What welfare can do for this second cohort is sharpen the bridge between education and work: apprenticeships properly funded and properly respected, employer incentives for hiring the under-25s, technical qualifications that are taken seriously rather than treated as the consolation prize. These are practical, fundable measures. They do not substitute for growth, but they reduce the friction between a young person willing to work and an employer willing to hire.

The hardest case for any conditional welfare system is the household where an adult refuses to engage with conditions and there are dependent children in the home. The principle of conditionality says the support stops. The principle of child protection says it cannot, because the children did not choose their parents and should not pay for them.

The answer is to preserve a basic floor of support for households with dependent children, but to channel an increasing share of that support away from the household payment and toward targeted provision. Free school meals, breakfast clubs, school uniform support, direct social services involvement, in extreme cases a referral pathway into care: these are interventions that reach the child rather than the adult, and they cannot be diverted by a parent who chooses not to work.

For the parents themselves, conditionality still applies, but its form changes. Mandatory engagement with parenting support, mandatory training and structured work activity tailored to caring responsibilities, sanctions for non-participation that affect the parental element of the claim rather than the household floor. The aim is to require effort of the parents without punishing the children for their failure to make it.

The housing element of Universal Credit, which replaced the old Housing Benefit, is for many claimants the largest part of their support — often greater than the standard allowance itself. For private renters it is capped at the Local Housing Allowance rate; for social tenants it pays actual rent, with deductions for under-occupation. It is the part of the welfare system most tangled with the wider housing market, with planning, and with the supply of social housing.

The principle to be laid down here is straightforward. Housing support is part of the safety net and must remain so. But it should not function as a guaranteed income stream for private landlords charging high rents in undersupplied areas, and it should not crowd out the incentive to work that the rest of UC is designed to preserve. The architectural detail — how LHA caps should be set, the relationship between the housing element and the social housing waiting list, the treatment of under-occupation — may be addressed in a future remembrance on housing and planning, where it would sit alongside the supply-side questions it cannot be reformed without.

It is sometimes said that the country has a cultural problem of expectation and resilience — that too many young people have been taught to expect support without effort, to treat ordinary difficulty as incapacity, and to ask less of themselves than previous generations did. There is something in this. But it is not a problem welfare policy can solve. The state shapes culture through schools, employers, the criminal justice system and the family. Welfare touches the question only at the margin.

What welfare can do is refuse to pander. The system should not be designed around the assumption that ordinary effort is too much to ask. It should not classify normal life-difficulty as illness. It should not treat work as an unreasonable demand. It should set firm terms, apply them consistently, and let people choose whether to meet them. Those who do are supported into work. Those who refuse take the consequences. That is not cruelty. It is the only honest basis on which a contributory society can function.

The state pension, the triple lock, the retirement age and the broader architecture of saving for old age belong to their own subject. They are properly the matter of a separate remembrance on pensions, which may follow. The principles set down here apply to working-age support and end where retirement begins.