Earlier this year I argued that Britain could save around £57bn a year from welfare: roughly £22bn from sickness, disability and incapacity benefits, £15bn from pension reform and £20bn elsewhere. Reform’s Making Welfare Work paper now puts its own number at £52.6bn by 2030/31.
That is against a welfare bill already running at roughly £334bn a year. £50bn is a large saving, obviously, but we are starting with a very large bill.
On sickness and disability, Reform and I have arrived at almost exactly the same place, and independently come up with the same saving, £22bn.
I argued that people with severe and permanent conditions should be properly protected, but that clear medical and functional criteria should distinguish between those who genuinely cannot work and those who retain some capacity to do so. Reform proposes a similar approach through a single Disability Needs Assessment, with protection for severe cases and Disability Support Accounts covering verified costs such as equipment, adaptations, transport and personal assistance.
That is where critics of the £50bn headline should start. If savings of around £22bn from sickness and disability are impossible, explain why. I do not think they are.
We have created a system which too often treats a diagnosis as a permanent exit from the labour market. Someone who genuinely cannot work should receive proper support. Someone who can work, perhaps with treatment, rehabilitation or reasonable adjustment, should be helped back towards it. Parking people on benefits for years is not a successful welfare policy, however compassionate the intention.
I also support Reform’s proposal for 20 hours of community work for Universal Credit claimants who remain in the Seeking Work regime after 12 months. Councils must not be allowed to use claimants as free replacements for jobs that would otherwise be paid, but the principle is sound. Reform only scores £636m of savings from this measure anyway. Its purpose is largely to keep people connected to the labour market and restore some reciprocity to welfare.
Fraud needs tougher treatment as well. Universal Credit overpayments are running at 8.5% of expenditure, with fraud itself at 6.8%. More case reviews and enforcement make sense where the savings comfortably exceed the cost. But fraud is not the main welfare problem. The bigger issue is a system whose incentives too often work in the wrong direction.
That is also why I am wary of Reform’s proposed Return to Work Cover. Employers, normally through insurance, would take greater responsibility for workers who remain sick after Statutory Sick Pay ends, offset by a 0.2 percentage-point cut in employer National Insurance.
My concern is simple: insurance prices risk. If an employee with a particular medical history becomes more expensive to insure, some employers will think twice before hiring them. A policy intended to bring people with health problems back into employment could end up making them more expensive to employ. Britain already makes hiring costly enough. I would want to see much more detail before putting this into law.
The other figure I would test very hard is the £20.6bn Reform expects to save by restricting most benefits to British citizens.
I agree with the principle behind it. Welfare cannot become an open-ended entitlement with no meaningful connection to contribution. But citizenship on its own is too blunt a test.
Someone who has lived legally in Britain for 20 years, worked throughout and paid substantial income tax and National Insurance has contributed to the system. A British citizen who has never worked has not. I would restore the contributory principle properly. Citizenship matters. Lawful residence matters. Contribution should matter too.
There are also legal and treaty complications, particularly around some EU citizens protected by the Withdrawal Agreement. When almost 40% of the total saving rests on one reform, the legal route, timetable and assumptions need to be nailed down before the full £20.6bn is banked.
Reform also needs to go further on Universal Credit itself. GBTT has previously modelled a case where splitting 40 hours of minimum-wage employment between two 16-hour workers cut the employer’s annual cost by £8,378 while increasing the Universal Credit bill by £23,801.
That is mad economics. The employer saves money, the taxpayer spends far more, and the system has effectively subsidised the fragmentation of a full-time job.
Welfare reform has to deal with those incentives as well as the headline caseload.
There is one area where Reform and I clearly diverge. My £57bn included around £15bn of pension reform. Reform largely leaves pensioners alone, even though roughly 55% of social-security spending goes to pensioners.
Poorer pensioners should be protected. But if we are serious about reforming welfare for a generation, pension spending cannot remain permanently outside the discussion.
So yes, I think £50bn is plausible. I would challenge parts of Reform’s costing, particularly the citizenship number. I would want more work on the employer-insurance proposal, go further on Universal Credit and include pensions.
But Reform has got the scale of the problem right.
If Labour thinks £50bn is “fantasy economics”, or Mel Stride thinks it is “simply not plausible”, show us which numbers are wrong, and by how much.