Welfare Reform

"Three Benefits. One System. No Excuses."

30+
Separate UK benefits replaced by 3
£9.5bn
Lost to benefit fraud and error annually
37 years
NAO has refused to certify DWP accounts as accurate — fraud is so large they cannot verify the figures
3 months
Civic Reward paid to you or a charity of your choice for a verified fraud report
430,000
Overseas social care workers — addressed through domestic retraining programme

The Problem

Britain's welfare system has over 30 separate benefits, 11 separate assessment processes, and is administered across DWP, HMRC, 300 local authorities, and a separate Scottish agency. Universal Credit was supposed to simplify this. After 13 years and 2.9 billion pounds, it merged six benefits into one but left PIP, Carer's Allowance, Child Benefit, Attendance Allowance, Council Tax Reduction, and the State Pension all outside the system. The DWP's accounts have been qualified by auditors for 37 consecutive years. This is not a system anybody designed. It is what happens when you add a new benefit to solve each new problem without ever removing the old ones.

New Zealand proved in 2013 that consolidation is achievable. They compressed 11 working-age benefits into 3 core payments, administered through a single agency, and reduced future welfare liability by 1.4 billion New Zealand dollars in the first year. We will do the same.

A. Forty Benefits Reduced to Five

The full list of what can currently be claimed runs to approximately 40 distinct working-age and household benefits. Many people do not know they exist. Many overlap. Many create perverse incentives. Most are administered separately with separate forms, separate assessments, and separate fraud vectors. Here is what Forge reduces them to:

Current benefit categoryNumber todayForge equivalent
Working-age income (UC, JSA, ESA, SSP, maternity etc.)9Jobseeker Support / Carer and Parent Support
Disability and care (PIP, Attendance Allowance, DLA, Carer's Allowance)5Health and Disability Support (provision, not cash)
Housing (Housing Benefit, Council Tax Reduction, Discretionary Housing)3Within the three-benefit system, paid direct to landlord
Children and family (Child Benefit, Tax Credits, SEND vouchers)4Child Benefit retained, auto-tapered by household income via HMRC. SEND provision paid direct to school.
Bereavement (2 payments)2Absorbed into three-benefit system as time-limited transition
Industrial injury (IIDB, REA, Pneumoconiosis schemes)3Within Health and Disability Support with specialist assessment
Winter and seasonal (Winter Fuel, Cold Weather, Warm Home Discount)3Replaced by structural energy bill reform (Section XI). Decoupling electricity from gas saves £200 to £400 per household permanently.
Passported and discretionary (uniform grants, council funds, household support)~12Single Crisis and Transition Fund. One application. Provision not cash.
Total~405 named entitlements

The five that remain as distinct named entitlements: Jobseeker Support, Carer and Parent Support, Health and Disability Support, Child Benefit (universal, automatically tapered), and the Crisis and Transition Fund. Everything else is either absorbed, replaced by a structural reform, or eliminated because it was a cash payment that should have been direct provision all along.

B. Three Core Working-Age Benefits

Every working-age benefit consolidates into three payments, administered through a single digital system.

BenefitWho it coversConditionsTime limit
Jobseeker SupportAnyone unemployed and able to workRegister immediately. Apply for jobs weekly. Accept retraining offer at month 3. Refusal tapers benefit.2 years. Tapers to subsistence rate after 2 years unless in approved full-time training.
Carer and Parent SupportSole parents with children under 6; full-time carers of disabled family membersPart-time work expected once youngest child turns 6. Carers assessed annually by clinical panel.Duration of caring responsibility. Transitions to Jobseeker Support when caring ends.
Health and Disability SupportPeople with permanent physical or mental conditions preventing workAssessed by independent specialist clinical panel. Mandatory care plan for recoverable conditions. No work requirement for permanent severe disability.No limit for permanent conditions certified by specialist. Reassessed every 3 years for recoverable conditions.

Child Benefit and the State Pension continue as universal entitlements outside this system. Everything else folds in. Three benefits. One application. One assessment. One agency.

C. What Happens to an Unemployed Person Under Forge: The Danish Model

The Danish model is the clearest answer to how Forge handles unemployment. Denmark pays 90% income replacement to unemployed workers but strictly limits benefits to 2 years and mandates active retraining from day one. The result: 2.6% unemployment, the second lowest in the OECD, and half of all benefit spells resolved within 3 months. The UK pays far less but allows indefinite claims with no meaningful retraining requirement. The result is 1.4 million people who have been on out-of-work benefits for more than a year. Forge adopts the Danish principle: generous short-term support, active support to find work, hard limit on passive claiming.

Week by week: what Jobseeker Support looks like

Why this is not punitive: the community contribution scheme at month 12 is sometimes described as "workfare." It is not. The person is being paid, contributing something real and visible to their community, building recent employment history that makes them more attractive to employers, and maintaining the routine and structure that paid employment provides. The alternative, paying someone to do nothing for two years, damages their mental health, erodes their employability, and costs significantly more. The Danish evidence is clear: active labour market spending generates a 2 to 4 times return in reduced long-term benefit spend.

The Dutch model is separate and applies to people already in work

The Dutch model in Section G applies to people currently employed who go off sick, not to unemployed people. The distinction matters. The Dutch model makes employers financially responsible for the first 6 months of sick pay, incentivising them to invest in occupational health and early return-to-work support. It reduces the flow of employed people onto long-term benefits. The Danish model gets unemployed people back to work. Together they address both routes into long-term welfare dependency: losing a job and drifting there, and staying in work until sickness tips you into it.

D. The Reciprocity Requirement: Attendance, Not Paperwork

A welfare system keeps public support when people believe it is reciprocal: that those who can contribute something do. Most working people who fund the system believe this, and they are right to. Forge builds reciprocity into long-term claiming, but it does so in a way that actually works, by learning from why previous attempts failed.

Why job-search requirements do not work

Every UK government has required benefit claimants to "actively seek work," usually measured by counting job applications. This is trivially easy to game and everyone knows it. You apply for jobs you have no chance of getting. If called to interview, you underperform deliberately. You tick the box and keep the benefit. The system cannot tell the difference between a person sabotaging an interview and a person who is simply nervous or unlucky, because intent cannot be proven. The DWP has known this for years. Its own evaluations of the Mandatory Work Activity scheme and the Community Action Programme found little or no employment benefit, precisely because effort-based conditions can be faked.

Forge does not count job applications as the qualifying activity for continued benefit. Searching for work is expected and supported, but it is never the test, because it cannot be verified and it can be gamed.

What cannot be faked is showing up

The qualifying activity for continued benefit after the initial period is verified attendance, measured by a register, at one of a defined set of activities. You can fake intent. You cannot fake presence. The requirement is not "try to get a job," it is "be at this place, doing this thing, for these hours," signed off by attendance.

Real job search continues alongside any of these, with work coach support, but the benefit is paid against the attendance register, not against an application count. A person working cash-in-hand while claiming cannot be at a placement three days a week during working hours. They must choose: give up the undeclared work, or give up the claim. This is the mechanism that actually deters fraud, and it is the largest single source of saving from the reform.

The honest fiscal position

Forge is honest about what this saves and how. The saving does not come from the unpaid labour itself, and it does not come from large numbers of people walking straight into jobs. Supervised placements cost money to run: organising, supervising, attendance management. Historically £3,000 to £5,000 per participant per year. The genuine saving comes from three secondary effects: the attachment effect, where people doing undeclared work disengage from the claim rather than attend (the DWP found this effect is often larger than the employment effect); the routine and recent-work-history effect, which moves some people back toward employment over time; and the strengthening of public consent for the welfare system as a whole. After running costs, the conservative net saving is approximately £1 to 2 billion per year. This is a real but modest number, and Forge presents it as such rather than inflating it. The reciprocity requirement is justified first by fairness and work-readiness, and second by a modest net saving, in that order.

The exemptions that make it humane

The placements must be genuinely useful, must not displace paid jobs, and must have a clear route toward real employment. This is the Danish model with a sharper reciprocity element, paired with genuine support. It is not the discredited workfare model of unpaid labour with nothing wrapped around it, which the UK already tried and which its own evaluations found did not work.

D. Citizen Digital ID: Real-Time Verification

The single most significant anti-fraud reform in this manifesto is not a new inspector or a new form. It is a Citizen Digital ID linked to real-time data from HMRC, the Home Office, and the NHS. Benefits are calculated from verified facts, not self-reported claims. The current system asks claimants to declare their income, their household composition, and their residency. Then it checks, slowly and incompletely. Forge inverts this: the system knows, from verified sources, and pays accordingly.

Why it launches as voluntary but becomes mandatory: The Digital ID cannot be made compulsory on day one because the government database connections (X-Road — see Section XVII) take 12 to 18 months to build. Mandating an ID card before the system can actually use it creates resentment without the benefit. So it launches as voluntary, incentivised by genuinely useful services — faster benefit processing, pre-filled tax returns, unified NHS login. Take-up is expected to be high because it makes life easier for the honest majority.

Compulsion matters for one specific reason: if voluntary take-up leaves a 20 to 30% opt-out, those opt-outs are disproportionately the fraud cases. The honest person signs up because it helps them. The fraudster avoids it because it exposes them. Without eventual compulsion for benefits specifically, the verification system has a deliberate structural hole in it. By Year 2 to 3, once the system is proven and the majority already have one, Digital ID becomes mandatory for accessing welfare benefits. At that point the remaining opt-outs face in-person verification, which is slower and more burdensome than digital ID — a further incentive to register.

What the Digital ID verifies automatically

The privacy architecture matters. This is not a surveillance database. The Citizen Digital ID uses the Estonian X-Road model: each agency holds its own data; the ID enables a secure, audited query between them with explicit claimant consent recorded at application. Every query is logged. Claimants can see which agencies have queried their data and when. The system knows what it needs to know to administer benefits correctly. It does not hold a centralised record of all activity. The consent is specific: "I agree that HMRC and DWP may share my income data to calculate my benefit entitlement." Not a blank surveillance authorisation.

E. Benefit Fraud Reporting: The Civic Reward Scheme

The UK loses approximately 9.5 billion pounds per year to benefit fraud and error. Much of it is known about locally before it reaches a DWP investigator. Employers know someone is working cash-in-hand while claiming. Landlords know there is an undeclared occupant. Teachers know a SEND voucher is being used for the wrong child. Schools know someone has not been in the country for months.

The current reporting mechanism is a phone line offering nothing in return. Reporting rates are low, prosecutions are rare, and the visible impunity of fraud in plain sight undermines public confidence in the system.

An honest limitation: The Civic Reward Scheme works well for fraud that is demonstrable from records — undeclared employment (visible via HMRC), overseas residence (visible via border data), wrong-child voucher use (visible to the school). It is less effective for cohabitation fraud where someone gives a false address, because proving someone actually lives somewhere requires investigation rather than data-matching. The Digital ID cross-referencing of council tax records and electoral roll makes cohabitation fraud harder to sustain over time as more data points appear at the real address. But it is not a complete solution. We say this honestly.

Forge introduces the Civic Reward Scheme: a structured financial incentive for verified benefit fraud reports, modelled on the US Internal Revenue Service whistleblower programme and the UK's successful Crimestoppers reward structure.

How it works

F. PIP: Services Not Cash, Targeted to Actual Need

Personal Independence Payment was designed to cover the extra costs disabled people face: heating, equipment, transport, personal care. The principle is correct. Disability is genuinely expensive. Disability charity Scope calculates that disabled households face average additional costs of around £1,095 per month. These costs are real.

But the current implementation has a fundamental design problem. PIP is delivered as cash, paid into a bank account every four weeks, with no requirement that it reaches what it was meant for. A flat rate is paid regardless of whether your actual disability costs are £200 a month or £1,500 a month. Universal Credit replaces lost income. PIP was supposed to cover the extra cost of being disabled. Cash with no verification cannot reliably do that.

Forge restructures PIP around a clear principle: PIP is for long-term physical disability, and it is delivered as services and targeted support rather than cash. Mental health is a health condition and belongs with the NHS, not the disability benefit system. So mental health is removed from PIP entirely and addressed through a major expansion of NHS mental health services accessible directly through the NHS App. This is not a cut to mental health support. It is moving mental health from a cash payment that does nothing to treat the condition, to actual treatment that does. The money saved funds both the NHS mental health expansion and the defence rebuilding the country urgently needs.

The political point: this reform is not an attack on disabled people or people with mental health conditions. It is the opposite. PIP becomes what it was always meant to be: support for the extra costs of long-term physical disability, delivered as the services and equipment people actually need. Mental health conditions are treated where they should be treated, by the NHS, with fast access to talking therapies through the NHS App rather than a cash payment that does nothing to address the underlying condition. Someone with depression does not need a monthly cheque that they lose if they recover. They need treatment, quickly. Forge delivers that. The money no longer spent on cash payments that fail to treat anyone funds both the NHS mental health expansion and the defence of the country.

The Three Streams (Plus NHS Mental Health Pathway)

G. Mental Health: Treated by the NHS, Not Paid by PIP

This is one of the most important reforms in the welfare programme, and the one most likely to be misrepresented, so the principle must be stated plainly. Approximately 1.37 million PIP claimants (37% of the total) have a mental health condition as their primary basis for the claim. Under the current system they receive cash, on average £6,200 per year, paid into a bank account with no requirement that anything is done with it, no treatment, and no expectation of recovery. The payment often continues indefinitely.

This is not support. It is abandonment with a cheque attached. A person with depression or anxiety receives money that does nothing to treat their condition, while waiting up to 18 months for NHS talking therapies. Worse, the cash creates a financial trap: if they recover, they lose the payment, so the system gives them a reason to stay unwell. No humane system should be designed this way.

What Forge does instead

Mental health is removed from PIP entirely. It is a health condition and it is treated by the NHS, where health conditions belong. The money currently spent on cash payments funds a major expansion of NHS mental health services accessible directly through the NHS App:

Severe and enduring mental illness

People with severe, enduring psychiatric conditions, including psychosis, schizophrenia, severe bipolar disorder, and severe treatment-resistant conditions certified by a consultant psychiatrist, are treated as having a long-term disability and are assessed for Stream 3 direct provision. They receive full ongoing support and care. The removal of mental health from PIP applies to the large majority of claims, which are for anxiety and depression, not to the small number of severe psychiatric disabilities that genuinely meet the threshold for long-term disability support.

Why this is better for the person, not just the budget

A person with moderate depression currently receives roughly £6,200 a year in PIP cash and waits up to 18 months for treatment. Under Forge they receive treatment within weeks through the NHS App, with income covered by Universal Credit while they cannot work, and a clinical team committed to their recovery. The cash they currently receive treats nothing. The service they receive under Forge treats the actual condition. This is a better deal for the person with the mental health condition, not a worse one.

The fiscal effect

Removing mental health from PIP saves approximately £8.5 billion per year in cash payments. The NHS App mental health expansion costs approximately £1.5 billion per year. On a conservative basis, accounting for the minority of claimants who have physical conditions alongside their mental health condition and are assessed for Streams 1 to 3, the net saving is approximately £5.3 billion per year. This money is directed to the defence rebuilding the country urgently needs: the drone industrial base, the integrated AI battle management system, and the conventional forces that three years of war in Europe have shown to be essential. A nation that cannot defend itself cannot protect anyone, including its most vulnerable.

The defining principle: PIP is for long-term physical disability. Mental health is a health condition treated by the NHS. The current system confuses the two and serves people with mental health conditions badly by paying them to stay unwell rather than treating them. Forge separates them cleanly. The NHS treats mental health, quickly and through the app. PIP supports physical disability through services and equipment. The money that was achieving nothing as cash funds treatment that works and the defence the country needs.

Stream 1: Employer-Funded Support, No Cash Payment

For disabled people assessed as capable of work with the right support (the largest single group, approximately 25% of current PIP claimants), the PIP cash payment is replaced entirely by employer-funded support delivered through the Employer Accommodation Fund. The disabled person receives no cash from PIP. They receive paid employment with all the additional costs of their disability covered by the Fund and substantial financial incentives for the employer that make them genuinely attractive to hire.

What the disabled person receives

What the employer receives

The principle

The disabled person's PIP cash payment of £4,000 to £6,000 per year, which is paid into their bank account regardless of whether they work, is replaced by approximately £4,000 to £6,000 per year channelled through the employer in the form of payments, NIC rebates, and risk guarantees. The disabled person receives the same value, but they receive it as a job rather than as a benefit. The fiscal cost is broadly similar. The economic and social outcome is transformatively different. A person in skilled work, earning a salary, paying tax, contributing to a pension, and building a career is a fundamentally different outcome from a person at home receiving cash to compensate for the absence of work.

This stream is the single largest fiscal restructuring in the welfare programme. Approximately 930,000 current PIP claimants whose primary issue is a mild to moderate physical disability that prevents work without support are moved off PIP cash and into supported employment under this framework. Estimated saving from this stream alone: £2.5 to 3.5 billion annually once fully transitioned, with the disabled person better off because they have a job and earnings rather than a cash payment.

Stream 2: The Personal Employment Budget — tightly controlled

For those building toward work where one specific barrier is identifiable, the daily living component is replaced by a Personal Employment Budget. This stream is deliberately limited and transitional — it is not a softer form of PIP. Without strict controls it becomes exactly that.

Stream 3: Direct provision for those who genuinely cannot work

For people with permanent, severe conditions where employment is genuinely impossible — certified by a specialist clinical panel, not a form — PIP is converted from cash to direct provision assessed by a qualified social worker.

Why this is fairer, not harsher: Scope estimates disabled households face average extra costs of around £1,095 per month. The maximum PIP cash award is £10,119 per year — still thousands short of actual costs for severely disabled people. Direct provision assessed by a social worker funds what is actually needed. The genuinely severely disabled person gets better provision. The system stops funding things that are not disability costs.

G. The Dutch Model: Employer Responsibility First

In the Netherlands, employers are responsible for the first two years of sick pay and must fund a structured return-to-work programme before the state takes over. The result: Dutch employers actively invest in occupational health, early intervention, and workplace adjustment to avoid the cost. Their long-term sickness rates are significantly lower than the UK's.

I. Anti-Abuse Measures

J. Investing in Domestic Social Care Workforce: Reducing Dependency on Immigration

The UK currently employs approximately 430,000 overseas workers in adult social care, representing roughly a third of the workforce. This reliance reflects a systemic failure to train, pay, and retain domestic workers rather than a genuine shortage of British people willing to do the work. Care work is demanding, skilled, and essential. It is also paid, on average, £11 to £12 per hour, often on zero-hours contracts, with minimal career progression and no professional recognition.

Forge addresses this directly through the same welfare reform framework:

K. Consolidate Small Payments Into Real Provision

The current system pays out dozens of small discretionary cash sums that are difficult to audit, easy to misdirect, and impossible to verify at scale. We replace untraceable cash with verified provision:

L. State Pension: Double Lock Replaces Triple Lock

The pension triple lock guarantees the state pension rises by the highest of three measures: inflation, average earnings growth, or 2.5% per year. It was introduced in 2010 by the coalition government at a time when pensioner poverty was a genuine and serious problem. Pensioners in 2010 were disproportionately poor relative to working-age households. The triple lock was the right policy for that moment.

It is not the right policy for 2026. The triple lock has worked. Pensioner incomes have risen substantially relative to working-age incomes over 16 years of operation. The UK state pension, as a share of average earnings, is now higher than it has been in decades. In some age cohorts, average pensioner household income now exceeds average working-age household income when housing costs are accounted for. The poverty rate among pensioners has fallen significantly since 2010. The policy achieved its purpose.

The 2.5% minimum floor — the part that Forge removes — has no economic justification. It was a political commitment, not an economic one. It was designed to be a visible signal to pensioner voters that the state pension would always grow in real terms, regardless of economic conditions. When inflation is 2% and earnings growth is 1.5%, the floor triggers and costs approximately £2 to 3bn more than either the inflation or earnings measure would require. That additional money flows to pensioners as a group regardless of individual need, paid for by working-age taxpayers who are on average less wealthy than the pensioners receiving it.

The double lock: what it is and what it protects

The double lock retains the protection that actually matters. The state pension rises by whichever is higher: inflation or average earnings growth. This means:

  • Pensioners are always protected against inflation eroding their living standards. If prices rise 5%, the pension rises at least 5%. The triple lock's core consumer protection is fully retained.
  • Pensioners never fall behind workers. If earnings grow faster than inflation (as they do in a healthy economy), the pension tracks earnings. Pensioners share in economic prosperity.
  • The 2.5% floor is removed because it provides an arbitrary minimum that bears no relationship to economic conditions, and because the cases where it triggers (when both inflation and earnings growth are below 2.5%) are precisely the cases where the broader economy is struggling and working-age taxpayers are least able to fund a floor that exceeds economic reality.

The people who need more support get it through a different route

A portion of the £5bn annual saving from the double lock is directed to a targeted pensioner poverty top-up for genuinely poor elderly people. This means more support for the 1.6 million pensioners in genuine poverty, funded by reducing the universal floor that benefits all 12 million pensioners regardless of their income or wealth. The poorest pensioner does better under the Forge model than under the triple lock.

The honest political argument

Every party in British politics knows the triple lock is fiscally unsustainable as the pensioner population grows. Every government since 2019 has found ways to temporarily suspend or avoid the floor when it became too expensive. Nobody will say publicly that it should be reformed because pensioners vote in large numbers and any change is immediately characterised as "attacking pensioners." Forge says it plainly: the 2.5% floor is no longer justified by the economic circumstances that created it, it is regressive relative to working-age people, and removing it while fully protecting pensioners against inflation and earnings falls is the honest and fair thing to do.

Estimated annual saving: £5 billion by Year 3, growing. The saving varies annually depending on whether the floor would have triggered. The OBR models it at approximately £5bn on average across the parliament. Part of this saving funds the targeted pensioner poverty top-up. The rest goes to deficit reduction.

M. Means-Testing Universal Pensioner Benefits via HMRC

Five universal benefits are currently paid to all pensioners regardless of income or wealth: the Winter Fuel Payment, free TV licence (over-75s), free bus pass, free NHS prescriptions (from age 60), and Attendance Allowance. Together they cost approximately £10.5bn annually. The majority flows to pensioners who do not need them — people with final salary pensions, investment income, and property wealth who are among the wealthiest households in the country.

The political reason these benefits are universal is that means-testing them historically required a separate application, a separate assessment, and a separate bureaucracy. The administrative cost and the intrusion on pensioners was substantial enough that successive governments left them universal.

That argument ends when the HMRC real-time household income data becomes available through X-Road. Once household income is verified automatically at the tax year end, means-testing costs almost nothing to administer. The assessment is the HMRC data. There is no form. There is no interview. The benefit adjusts automatically based on the previous year's verified household income, exactly as Child Benefit already does through the High Income Child Benefit Charge.

What changes and what does not

BenefitCurrent positionForge positionThresholdAnnual saving
Winter Fuel Payment Labour means-tested to pension credit recipients only (2024) Extended to all pensioners below median household income. Withdrawn above. £35,000 household income. Tapered to zero by £50,000. £500m
Free TV Licence (over-75s) BBC funds for pension credit recipients only Free below £35,000 household income. 50% subsidy to £50,000. Full price above. £35,000 to £50,000 taper £350m
Free Bus Pass Universal from state pension age Universal to £50,000 household income. Withdrawn above. £50,000 household income £270m
Free NHS Prescriptions (60+) Universal from age 60 Free for household income below £35,000 and all on pension credit. Reduced £3 per item above £50,000 (vs £9.90 standard charge). £35,000 to £50,000 taper £400m
Attendance Allowance Universal, cash, no means test Already absorbed into Health and Disability Support Stream 3 (direct provision assessed by social worker). Not cash. Provision assessed on need not income £2bn (from direct provision conversion)
Total annual saving£3.5bn

Why the HMRC mechanism changes everything

Previous means-testing proposals for these benefits failed because the administrative cost was too high and the intrusion on pensioners was too great. Forge's position is different because:

  • No application. Pensioners do not apply for means-testing or fill in any form. The HMRC system verifies their household income automatically at the tax year end using data already held.
  • The same mechanism already exists. Child Benefit is currently means-tested through the High Income Child Benefit Charge — HMRC claws back Child Benefit from households above £60,000 through the self-assessment system. Forge applies exactly the same mechanism to pensioner benefits. The plumbing already exists.
  • Annual adjustment. If a pensioner's income falls below the threshold in any year (through retirement, investment losses, or reduced pension), the benefit is automatically restored. The system tracks real income, not a one-time assessment that becomes outdated.
  • Full protection for the poorest. Every pensioner on pension credit or with household income below £35,000 receives every benefit in full, exactly as now. The withdrawal affects only those above the median household income threshold.

The political argument

Universal pensioner benefits were justified when pensioners were disproportionately poor. That is no longer the case. Pensioner household incomes, after housing costs, are now on average higher than working-age household incomes in some age groups. A household with a final salary pension, investment income, and a paid-off house does not need a free bus pass or a Winter Fuel Payment. The working-age person on £28,000 funding those benefits does.

This is not an attack on pensioners. The 7 million pensioners on the lowest incomes retain every benefit in full. The 4 million pensioners with household incomes above the threshold pay for their own bus pass and TV licence. This is a reasonable ask of people who are, on average, among the wealthier households in the country.

Forge is also honest about the electoral consequence: pensioners are less likely to vote Forge regardless of this reform, and the working-age voters who benefit from the savings are more likely to. The political economy is correct as well as the policy.

What this is not: we are not stripping benefits from genuinely disabled people, cutting state pensions, or reintroducing cruel box-ticking assessments. We are spending less on cash transfers and more on treatment, retraining, and getting people back to productive lives. The measure of success is not the benefit bill falling. It is employment rising, ill-health reducing, and genuine claimants receiving better support than they currently get. Every pound saved on administration and fraud goes directly into the retraining guarantee and disability provision.

L. Projected Savings

Consolidating 30 benefits into 3, eliminating duplicate assessments, closing the 9.5 billion fraud and error gap through real-time verification, and reducing the DWP's 89,000-person department through automation generates substantial savings. Combined with Dutch-model employer responsibility and tightened mental health eligibility, the projected savings are 15 to 20 billion pounds per year by year 5. These are phased because the new system takes time to establish: year 1 saves around 2 billion, growing each year as the new processes bed in and fraud reduction compounds.

Disagree with any of this?

Tell us. The discussion hub is open. Forge Club members can formally propose amendments. 60% support gets it into the policy review.

Join the Discussion Full Manifesto (PDF)