Every Pound Accounted For
A manifesto without numbers is a wish list. A manifesto with numbers but no independent verification is propaganda. Forge commits to both: full costing of every commitment and submission of all fiscal assumptions to the Office for Budget Responsibility for independent verification before any general election. Where OBR analysis disagrees with our internal estimates, we publish both alongside each other and explain the difference. Where we are wrong about a forecast, we say so publicly and adjust.
The honest year-by-year position
This manifesto costs money before it saves money. The savings are real but they are back-loaded: welfare reform takes 7 years, NHS transformation several years, digital integration several years. The costs are front-loaded: the employer National Insurance cut is immediate, defence spending ramps fast, and VAT compensation is paid from day one. The fiscal position is negative in Years 1 to 3, approximately break-even in Year 4, and a modest surplus in Year 5. This is the normal trajectory for a transformative programme. The critical commitment is that the trajectory is the right one and the OBR verifies it before we take office.
The numbers below are presented in two ways: our internal estimates, and conservative OBR-style estimates that apply standard behavioural response discounts of 30 to 40% to revenue measures and 50% implementation delivery assumptions to savings. Both are shown. The honest range sits between them.
Phased fiscal position: Year by Year
| Fiscal element | Year 1 | Year 2 | Year 3 | Year 4 | Year 5 |
|---|---|---|---|---|---|
| Revenue (tax reforms) | £8.8bn | £11.6bn | £14.0bn | £15.4bn | £16.1bn |
| Savings (structural reforms) | £10.4bn | £20.7bn | £32.1bn | £42.5bn | £48.7bn |
| Total inflows | £19.2bn | £32.3bn | £46.1bn | £57.9bn | £64.8bn |
| Gross costs (all commitments) | £30.5bn | £33.4bn | £36.2bn | £39.5bn | £40.7bn |
| Defence Bonds (off balance sheet) | £3.0bn | £5.0bn | £8.0bn | £12.0bn | £12.0bn |
| Net on-balance-sheet costs | £27.5bn | £28.4bn | £28.2bn | £27.5bn | £28.7bn |
| Net fiscal position (conservative) | -£11.3bn | -£1.1bn | +£12.8bn | +£24.8bn | +£30.6bn |
| Year 1 to 3 deficit bridged by existing government deficit trajectory. Year 4 and 5 surplus reduces national debt-to-GDP ratio. | |||||
The phased investment bar: where the money goes each year
This section is not a separate policy. It is the accountability framework for every other section.
A. Revenue and Savings: The Sources
| Source | Annual gain | Basis |
|---|---|---|
| Universal 15% VAT (net of income tax cut) | 15 to 25bn | Broader base, fewer reliefs. Net of compensating income tax rate reduction and employer National Insurance cut. OBR-scoreable from consumption data. |
| Welfare reform (30 benefits to 5, Dutch employer model, MH treatment-first, digital ID verification) | 10 to 15bn | Phased over 7 years (see welfare implementation timeline). Year 1: £2bn from digital ID fraud reduction. Years 2 to 4: benefits consolidation and employer sick-pay model bed in. Years 5 to 7: full PIP transition savings as social worker assessment replaces cash PIP. Conservative estimate: Netherlands saved 0.5 to 1% of GDP over a decade. Mental health is removed from PIP entirely and treated by the NHS via the app: £8.5bn cash removed, £1.5bn NHS expansion cost, net £5.3bn directed to defence. |
| NHS efficiency (AI triage, virtual wards, step-down centres) | 14bn | Based on published NHS virtual ward trial data, blocked bed costs at 2,000 per night, and Nuffield Trust modelling of AI triage deployment. |
| Water nationalisation (no dividends, lower borrowing costs, single regulator) | 3 to 5bn | Dividend extraction currently 2 to 3bn; refinancing at gilts rate saves 1 to 2bn; Ofwat abolition saves running costs. |
| Digital admin savings (fraud reduction, shared platforms, automated returns) | 7 to 9bn | Benefit fraud and error reduction (4 to 6bn from digital verification). Admin headcount reduction. Shared council platforms. |
| Profit shifting enforcement and platform VAT | 2 to 4bn | Minimum 10% margin on UK revenue for multinationals. Platform VAT on overseas goods. Both are enforcement-only, not new taxes. |
| Accelerated 2-year degrees | 4 to 7bn | Lower student loan outlay, faster repayment, students entering tax-paying employment one year earlier. |
| Civil service attrition through digital integration | 4 to 5bn | 100,000 fewer administrative roles at average salary plus on-costs. No compulsory redundancy; natural attrition over 5 years. |
| Cannabis duty | 1.5 to 2bn | 30% specific duty plus VAT on regulated market. Transform Drug Policy Foundation estimate, independently verified by Institute for Fiscal Studies analysis of cannabis reform scenarios. |
| Prison reform (community service replacing short sentences) | 1bn+ | 30,000 short-sentence prisoners at 48,000 per year replaced with community supervision at approximately 8,000 per year. 50% reduction in short-sentence population. |
| Crown Estate Sovereign Grant saving | 0.08bn growing | 132m grant to 50m fixed: immediate saving 82m, growing as seabed revenues grow with offshore wind expansion. |
| State pension triple lock to double lock | 5bn by Year 3, growing | Triple lock costs approximately £5bn more annually than a double lock. 2.5% floor has no economic justification. Saving grows with each year the floor would otherwise have triggered. |
| Means-tested universal pensioner benefits via HMRC | 3.5bn by Year 2 | Winter Fuel Payment, TV licence, bus pass, NHS prescriptions means-tested automatically via HMRC household income data. No new form or assessment. Attendance Allowance absorbed into Health and Disability Support direct provision. Thresholds: free below £35k household income, withdrawn above £50k. 7 million poorest pensioners unaffected. |
| Welfare reciprocity requirement (attendance-verified) | 1 to 2bn | Continued benefit after initial period conditional on verified attendance (training, supported placement, or community contribution), not gameable job-search paperwork. Saving comes mainly from the attachment effect (undeclared workers disengaging) and fraud deterrence, after placement running costs. Conservative and modest by design. |
| Business subsidy zero-basing | 3.8 to 4.8bn by Year 3 | Every business grant programme reviewed against additionality evidence within Year 1. Programmes that cannot demonstrate they caused activity that would not have happened anyway are discontinued. Conservative estimate: 40% of £9.5bn portfolio fails evaluation. Regional growth grants absorbed into BSA geographic weighting and property tax redistribution. |
| Progressive property tax (Growth Infrastructure Fund) | £10bn ring-fenced | The progressive property tax raises approximately £10bn more than a flat rate, paid by the upper value bands (properties concentrated in London and the South East). This £10bn is ring-fenced entirely for the Growth Infrastructure Fund: transport, energy grid, and digital connectivity. It is not part of the general surplus; it is hypothecated to capital investment that raises long-term productivity. |
| Government IT project portfolio review and cancellation | 0.5 to 2bn (one-off and ongoing) | The UK government has written off £37bn in failed IT projects since 2000. An independent review with power to cancel projects below a defined value-for-money threshold. Ongoing saving from stopping projects before they fail rather than after. |
| North Sea and shale gas revenues (ring-fenced) | 3 to 6bn | New permits and council-led shale extraction. Ring-fenced to Energy Sovereign Fund, not general spending. |
| TOTAL NET ANNUAL GAIN | 43 to 67bn | After income tax cut funded from broader VAT base. Welfare savings phased over 5 years; year 1 saves approximately 2bn, growing annually. |
B. Spending Commitments
| Commitment | Annual cost | Notes |
|---|---|---|
| Defence increase to 3% of GDP (phased) | 15 to 20bn | 2.3% to 2.5% in Year 1, 2.75% Year 2, 3% Year 3. Capital front-loaded via Defence Bonds (off-balance-sheet). |
| Employer National Insurance cut (threshold to £14k, EA to £15k) | 9.5bn | Replaces the business rates cut. Money reaches employment not landlords. Funded from the same envelope. Business rates rebalancing is revenue neutral. |
| British Strategic Accelerator | 2bn per year | 10bn over parliament, taking equity stakes. Returns reinvested into fund. Long-term self-sustaining. |
| National Resurfacing Programme | 2.5 to 4bn | 5-year ring-fenced programme. Current government commitment is 7.3bn over 4 years; this is roughly 4bn additional, addressing 18.6bn backlog. |
| Solar and marine energy CfDs and storage | 2 to 3bn | Capital investment partly self-financing through energy sales and CfD windfall returns. |
| NHS mental health care plans (1bn Talking Therapies) | 1 to 2bn | Expanded Talking Therapies for welfare claimants. 4-week waiting time guarantee. Offset partially by reduced long-term benefit costs. |
| Step-down care centres (50 centres) | 1 to 2bn | Offset by freed hospital beds at 2,000 per night. 3,250 freed beds saves 2.4bn annually. Net saving after centre costs. |
| Fresh Start communities and drug courts | 0.5 to 1bn | Offset by reduced prison costs at 48,000 per custodial place per year versus 8,000 for community alternatives. |
| Retraining guarantee (Danish model) | 2 to 3bn | Active labour market spend from 0.3% to 1% of GDP. Funded partly from welfare reform savings. |
| PIP transition: social worker assessment workforce | 0.5 to 1bn (Years 2 to 5 only) | Approximately 8,000 to 12,000 additional qualified social workers required to deliver 18-month cycle assessments for 3.7 million claimants. Phased recruitment over 4 years alongside social work training expansion. Transitional cost: front-loaded in Years 2 to 5, then self-funding as social worker pay replaces Atos/Capita contractor costs. Net cost after contractor saving: approximately £200 to 400m annually during transition. |
| PIP cash transitional protection (phase-in period) | 1 to 2bn (Years 1 to 3 only) | Existing PIP claimants receive cash payments until their first social worker assessment under the new system. Nobody has their payment stopped before an assessment takes place. Transitional cost reduces to zero as the assessment cycle completes. |
| School activity coaches (18,000 phased) | 0.2 to 0.7bn | Year 1 legislation only. Years 2 to 4 hire 6,000 annually at NHS Band 3 to 4. Partly offset from 700m DfE central admin reallocation. |
| Channel Patrol Fleet and border enforcement | 0.5bn | Fast patrol boats, bilateral returns infrastructure, Coastal Maritime ISR Service. |
| MP salary increases (clean wage) | 0.05bn | 650 MPs at an average increase of approximately 60,000 each. Offset by abolishing second jobs, consultancies, and outside income. |
| TOTAL ANNUAL COST | 40 to 55bn | PIP transition costs (social worker workforce and cash protection) add £1.5 to 3bn in Years 2 to 5 only, then taper to zero. School coaches phase to full cost from Year 4. Defence phased to 3% by Year 3. Employer NI cut funded from broader VAT base, same envelope as the business rates cut it replaces. |
C. The Balance
Internal estimate: revenue and savings of £43 to 67bn annually at full implementation versus costs of £38 to 52bn, producing a surplus of £5 to 15bn. Conservative OBR-style estimate (applying standard 30 to 40% behavioural response discounts to revenue and 50% implementation delivery assumptions to savings): approximately £39bn in annual inflows versus £35bn in net on-balance-sheet costs, producing a conservative surplus of approximately £4bn by Year 5. The honest range is £4 to 10bn annual surplus by Year 5. The fiscal position is negative in Years 1 to 3, approximately neutral in Year 4, and modestly positive in Year 5. This is the normal and expected trajectory for a front-loaded investment programme with back-loaded structural savings.
Two Additional Structural Savings
State Pension: Double Lock Replaces Triple Lock
The pension triple lock guarantees the state pension rises by the highest of inflation, average earnings growth, or 2.5% per year. It was introduced in 2010 when pensioner poverty was a genuine problem. It has achieved its purpose: pensioner incomes have risen substantially relative to working-age incomes, and the poverty rate among pensioners is now lower than among working-age households in some measures. The 2.5% minimum floor has no economic justification. It was a political commitment, not an economic one. When inflation is 2% and earnings growth is 1.5%, the floor costs approximately £2bn above what either the inflation or earnings measure would require. That money flows to pensioners as a group regardless of need.
Forge replaces the triple lock with a double lock: the state pension rises by the higher of inflation or average earnings growth. This protects pensioners against both inflation eroding their living standards and falling behind workers. It ends the arbitrary 2.5% minimum that costs £5bn annually compared to the double lock. The saving grows with each year the floor would otherwise have triggered.
The political argument for this reform is not that pensioners do not deserve support. It is that the 2.5% floor provides additional support regardless of whether pensioners need it, paid for by working-age taxpayers who are on average less wealthy than the pensioners receiving it. A targeted pensioner poverty top-up for the genuinely poor elderly, funded from part of the £5bn saving, would do more for pensioner wellbeing than the universal floor does. The honest framing: the triple lock was right for 2010. The double lock is right for 2026.
Estimated annual saving: £5bn by Year 3, growing. The saving varies by year depending on whether the floor would have triggered. In years of low inflation and low earnings growth it is larger. In years of high earnings growth it is smaller or zero. The OBR models this as approximately £5bn annually on average across the parliament.
Government IT: Stop Failing Projects Before They Fail
The UK government has written off approximately £37bn in failed IT projects since 2000. This includes the NHS National Programme for IT (£10bn written off), Universal Credit early phases, Disclosure and Barring Service, the Student Loans Company modernisation, and dozens of smaller departmental failures. The pattern is consistent: large contracts awarded to large suppliers, requirements that change during delivery, governance too weak to cancel projects that are clearly failing until hundreds of millions have been spent.
The digital integration programme in Section XVII (UK X-Road) prevents this for new systems by mandating open standards and interoperability. But the existing portfolio of at-risk projects contains approximately £20bn in committed but undelivered IT spending. A systematic review by the Infrastructure and Projects Authority with the power to recommend immediate cancellation of projects below a defined value-for-money threshold, within Year 1 of taking office, would identify and stop projects that are currently consuming money while heading toward write-off.
- Immediate portfolio review in Year 1 covering all central government IT projects above £20m. Red-rated projects (IPA traffic light system) are presumed for cancellation unless the sponsoring department can demonstrate a credible recovery plan within 90 days.
- No further large contracts with incumbent failed suppliers (Capita, Fujitsu, IBM government division) without a competed procurement. The instinct to re-award to the incumbent because switching is costly has historically produced worse outcomes at higher cost.
- In-house IT capability investment (covered by the consultancy cap saving in Section VII): 500 additional government digital specialists, employed directly, replacing the most expensive external contractor roles.
- Estimated saving: £500m to £2bn, partly one-off (contracts cancelled before further spend) and partly ongoing (prevention of the next generation of failing projects).
The timing matters. Welfare reform savings are phased over 7 years because the PIP transition requires building a social worker assessment workforce that does not currently exist at scale; Year 1 saves around 2 billion from digital ID verification alone, growing as the new system beds in. NHS efficiency savings build as virtual wards scale and step-down centres open. The employer National Insurance cut costs approximately 9.5 billion, funded from the broader VAT base. Unlike a business rates cut, this money reaches employment and wages directly rather than being captured by commercial landlords. The business rates system is separately rebalanced on a revenue-neutral basis. Defence spending increases in Year 1 and 2 before the full savings materialise, creating a front-loaded cost that the Defence Bonds mechanism bridges off-balance-sheet.
D. What Is Off-Balance-Sheet
- Defence Bonds (10 to 15bn): retail debt raised through NS&I. Counted as public sector net debt but not against the spending review limits in the same way as gilts issuance. Interest paid to retail savers at 4 to 4.5%.
- Water infrastructure (120bn over 25 years): funded from customer bills. Water England is a public corporation. Its infrastructure investment appears on the public sector balance sheet but is serviced entirely from revenues. No general taxation required.
- Toll-funded infrastructure: private capital, repaid through tolls. Zero Treasury exposure. Projects do not appear on the public sector balance sheet.
- British Strategic Accelerator equity stakes: investments, not spending. The BSA takes equity in companies. The value of those stakes is a public asset, not a cost, if the investments are well-made.
E. The Debt and Deficit: An Honest Account
The UK carries approximately 2.9 trillion pounds in national debt, around 93% of GDP. The annual deficit was 138 billion pounds in 2024 to 2025. Debt interest alone costs 106 billion pounds per year, more than the entire defence budget. This manifesto does not promise to eliminate the deficit in one parliament. That would require either enormous tax increases or devastating spending cuts that are neither proposed nor desirable.
What Forge commits to:
- Every new spending commitment is funded by a specific saving or revenue source, in this document, costed
- A target of reducing the annual deficit by 20 to 30 billion pounds within the parliament, funded by the reforms in this manifesto
- A full Budget published within 60 days of taking office, scored independently by the OBR, before any reform takes effect. No policy is implemented without fiscal verification
- The reforms are phased over 5 years precisely so that revenue materialises before spending is committed
- Debt-to-GDP stabilises and begins to fall in Year 4 or 5, as the structural savings compound and the deficit reduction target is approached
F. The Lesson of 2022
In September 2022, the previous government announced 45 billion pounds of unfunded tax cuts. Markets responded by raising UK borrowing costs by 1.5 percentage points in 48 hours, forcing a reversal. The cost to taxpayers of those few days was estimated at over 30 billion pounds in higher debt service costs. Forge draws the obvious lesson: credibility with markets requires specificity, sequencing, and independent verification. Not promises. Not aspirations. Scoreable, published, verified fiscal commitments.
We will not promise what we cannot pay for. We will not pay for what does not work.
The OBR commitment: before any general election, Forge submits the complete manifesto costing to the Office for Budget Responsibility. We publish the OBR's full assessment alongside our own estimates. Where they disagree, we explain why. Where we accept their revision, we update our figures publicly. This is not standard practice in British political manifestos. It should be. Forge makes it so.