Executive Summary

The Barnett formula decides how much of the money the UK Government spends on public services in England is matched by funding for the devolved governments in Edinburgh, Cardiff and Belfast. It was introduced in 1978 by Joel Barnett, then Chief Secretary to the Treasury, and was meant to be temporary.¹ It has never been written into law, and the Treasury has full discretion over how it is applied.¹ In 2024-25 it delivered £83 billion of block grant funding before adjustments for devolved taxes: £45.3 billion for Scotland, £20.2 billion for Wales and £17.6 billion for Northern Ireland, about 6.5% of all UK public spending.² The 2025 Spending Review sets the three grants at an average of £50.9 billion, £22.4 billion and £19.3 billion a year from 2026-27 to 2028-29.³

The formula does not look at need. It takes each year's change in spending on a service in England and passes on a population share of that change, so the original levels of the 1970s are carried forward and never reassessed.⁴ Over time that pulls funding per person in each nation slowly towards England's, whatever each nation's needs are doing. That is why Wales and Northern Ireland, which independent assessments found to have higher needs than Scotland, have each had to negotiate a needs-based floor: Wales at 115% of comparable English funding per person, Northern Ireland at 124%.⁴ Scotland has no such floor and no current assessment. The only published needs estimate for it, made by the Holtham Commission in 2009, put its need at about 105% of England's, while Scotland is funded at "over 20%" above equivalent UK Government spending.⁵³ The Holtham Commission said its Scottish figure would need refining, and that is as far as the evidence goes.

This pillar makes four arguments. First, the UK should measure relative need, independently and regularly, for all three devolved nations and for the English regions. Second, once it has, the same needs-based rule should apply to all three devolved governments, as it already does for two of them. Third, the rules should be put into law and the results published every year. Fourth, England's poorest regions, which have no Barnett grant and no needs floor, should be funded on the same published measure as England's new mayoral authorities take on more powers.

Key Proposals

1. How the Money Moves

Scotland, Wales and Northern Ireland each have a government that runs most of their public services: health, schools, local government, transport, housing and, to differing extents, justice and policing. Most of the money they spend comes from a block grant from the UK Government, and the size of that grant is set by the Barnett formula.⁴

The formula works on changes, not levels. When the UK Government changes its planned spending on a service in England, each devolved government receives a share of that change equal to its population as a proportion of England's, adjusted by a "comparability factor" for how far the service is devolved: 100% for health and schools, 0% for things the UK Government still runs everywhere, such as tax collection.⁴ The result, the "Barnett consequential", is added to an existing baseline. Those baselines, the Treasury's own rulebook says, "essentially reflect the original block grants and all previous Barnett allocations" and "are not therefore revisited afresh at each spending review".⁴ The levels set in the late 1970s have been carried forward, increment by increment, ever since.

Two consequences follow. The first is that nothing in the formula responds to need. If Wales's population becomes older or poorer relative to England's, its grant does not change. The second, often called the "Barnett squeeze", is that equal cash increments per person on unequal starting levels gradually pull funding per person in each nation towards England's level. In practice this has happened more slowly than expected since 1999.¹

Since 2016 the system has become more complicated. Scotland now sets its own income tax rates and bands and runs a growing share of social security; Wales sets part of its income tax. Each grant is reduced by a "block grant adjustment" to reflect the tax the UK Government no longer collects there. Scotland's adjustment was £13.7 billion in 2024-25, taking its £45.3 billion grant to a net £31.5 billion before its own tax receipts are added back.² Scotland has chosen higher income tax rates than the rest of the UK, and the Scottish Fiscal Commission forecasts that this leaves its budget about £969 million better off in 2026-27 than if it had matched UK rates.⁶ That money comes from Scottish taxpayers, not from the formula, and nothing in this pillar's proposals touches it.

2. Who Gets What

The Treasury publishes spending per person for every nation and English region each year. In 2024-25, identifiable public spending per person was £15,563 in Scotland, £15,155 in Wales, £16,116 in Northern Ireland and £13,134 in England, against a UK average of £13,504.⁷ Against England, that is 18% more per person in Scotland, 15% more in Wales and 23% more in Northern Ireland. These figures cover all public spending that can be attributed to a place, including pensions and benefits paid by the UK Government, not only the services the devolved governments run.

England's own regions vary almost as much. London received £15,217 per person, more than Wales and close to Scotland. The North East received £14,102, while the East Midlands received £12,158 and the South East £12,031.⁷ Part of London's figure reflects its higher costs and its transport network; part of the North East's reflects higher benefit spending in a poorer region, which rises automatically with need in a way the block grants do not.

The extra money in the devolved nations does not go evenly across services. On the Treasury's figures for 2024-25, spending per person on health in Scotland was 1% above the UK average, and in Wales 8% above. Education was 20% above in Scotland, housing and community services 71% above, and agriculture, fisheries and forestry 73% above.⁷ That is how devolution is meant to work: each government decides its own priorities within the grant it gets. It also means that a higher grant per person does not translate into any particular service being better funded.

Scotland's own official accounts, Government Expenditure and Revenue Scotland (GERS), put the same pattern in fiscal terms. In 2025-26, revenue raised in Scotland was £17,718 per person, almost exactly the UK average of £17,720 once a geographic share of North Sea revenue is included. Spending on behalf of people in Scotland was £22,281 per person against £19,561 for the UK as a whole, £2,720 more.⁸ Scotland's notional deficit was £25.3 billion, 10.9% of its GDP, against 4.2% for the UK.⁸

3. What We Know About Need

Need for public spending is not the same as population. An older population needs more health and social care; a poorer one more support; a sparse, rural one costs more per person to serve, because roads, schools and clinics serve fewer people each. Every serious attempt to fund the UK's nations fairly has tried to measure these differences. Only a handful have been published, and only two are in use.

The most influential is the Holtham Commission, set up by the Welsh Government. In December 2009 it built a needs formula from the indicators that UK and devolved governments already use to share out their own budgets, and found that Wales "should receive some £115 for every £100 of funding spent on comparable activities in England", against £112 it then received.⁵ The same formula gave £105 for Scotland and £121 for Northern Ireland, "although these estimates would need refinement to take account of different devolved responsibilities".⁵ The House of Lords Select Committee on the Barnett Formula reported the same year. On its initial analysis it concluded that "Scotland now has a markedly lower need than Wales and Northern Ireland in comparison with England", and recommended an independent, advisory UK Funding Commission to assess relative need about every five years.¹⁰ The UK Government replied that it had "no plans to set up such a body", and defended Barnett for "allowing the devolved administrations to determine their own assessment of needs and priorities".¹⁰

Since then, two of the three nations have had needs assessments written into their funding. Wales's 2016 fiscal framework adopted Holtham's 115% as the floor: a needs-based factor in the formula will ultimately be set at 15%, with a transitional factor of 5% while Wales's relative funding stays above 115%.⁴ The Spending Review 2025 states that Wales is now funded above that level.³ Northern Ireland's 2024 interim fiscal framework adopted an assessment by the independent Northern Ireland Fiscal Council that its need is 124% of England's. If its funding per person falls below 124% of comparable English funding, a 24% needs factor applies in the formula; above it, a 5% transitional factor.⁴ The Treasury applied the 24% factor to Northern Ireland's 2025-26 settlement, which under those rules means its funding had fallen below its assessed need.⁴

Scotland has no factor and no assessment. Its funding is set by the unadjusted formula, and the Spending Review 2025 says it receives "over 20% more funding per head than the equivalent UK Government spending".³ Set against Holtham's 105%, that would mean Scotland is funded well above need. But that comparison rests on a single 17-year-old estimate that its own authors flagged as unrefined, and Scotland has real costs the comparison may miss. In 2024 the age-standardised death rate in Scotland was 1,269 per 100,000 for men and 968 for women, against 1,094 and 797 in England and Wales: 16% and 21% higher.¹¹ Large parts of the Highlands and Islands are very sparsely populated, which raises the cost of every service. A current assessment might put Scotland's need well above 105%. It might not. The UK does not know, and has chosen not to find out.

4. The English Question

Barnett covers the three devolved nations. England's regions have no block grant, no formula and no needs floor; their funding is the sum of decisions made by UK Government departments and the local government finance system. This creates comparisons that sit uneasily with any principle of funding by need.

Household income per person is one of the clearest single indicators of need that the ONS publishes for every nation and region. In 2024, with the UK average at 100, it was 79 in the North East, 83 in Wales, 84 in Yorkshire and the Humber, 85 in the West Midlands, 86 in Northern Ireland, 87 in the North West, 88 in the East Midlands and 94 in Scotland. London was 141.¹² Set those against the Treasury's figures for spending per person in 2024-25, again with the UK at 100: the North East 104, the North West 102, the West Midlands 97, Yorkshire and the Humber 95, the East Midlands 90, Wales 112, Scotland 115 and Northern Ireland 119.⁷

England is now building its own devolution. The English Devolution and Community Empowerment Act 2026 creates "strategic authorities" across England, with powers over transport, skills, housing and strategic planning, economic development and public service reform, and the most established mayoral authorities will receive integrated funding settlements.¹³ That is a chance to avoid repeating Barnett's original mistake. If the funding of England's strategic authorities is set from the start by a published measure of need, the same measure used for the devolved nations, the UK will for the first time have a single, comparable answer to the question of who needs what.

5. What Other Countries Do

Most federal and devolved countries have a formal system for sharing money between richer and poorer parts of the country, and many of them measure need directly. Australia's is the longest-established. Its Commonwealth Grants Commission, set up in 1933 and independent of government, advises on how the revenue from the national goods and services tax is shared among the states. Its stated aim is that each state should have the fiscal capacity to provide services at the same standard, if it makes the same effort to raise revenue and runs its services equally efficiently.¹⁴ The Commission reviews its methods periodically, most recently in 2025, and publishes them in full.¹⁴

The Australian system is not a model without problems. Western Australia's share of the revenue fell to a record low as its mining income grew, and the dispute that followed led to a national inquiry by the Productivity Commission and, in 2018, a change in the law that put a floor under every state's share.¹⁵ That is the strongest practical warning about needs formulas: they make winners and losers visible, and the losers fight back. It is also evidence that a formula can survive that fight, be revised in public and keep working for ninety years, which is more than Barnett, designed as a stopgap and never legislated, can claim.

6. The Steel Man: The Case for Leaving Barnett Alone

The strongest case against this pillar is that Barnett works because it does not try to measure need. Every needs formula is a set of contestable judgements dressed up as arithmetic: which indicators count, how much weight each carries, how sparsity is costed, whether higher spending reflects need or just past choices. Holtham's own Scottish estimate came with a warning that it was not reliable, and Australia's experience shows that a formula becomes a permanent source of conflict between the governments it funds. Barnett's simplicity is a feature. Nobody has to agree on need; the devolved governments get a predictable share of every change in England and decide for themselves what their nations need, which is the point of devolution and was the UK Government's own defence of the formula in 2009.

Barnett is also already doing what this pillar wants, slowly. Equal increments per person on unequal baselines squeeze relative funding towards England's level over time. Where that squeeze bit too hard, for Wales and Northern Ireland, floors were negotiated and now work. Adding a needs factor for Scotland adds complexity to reach an outcome the formula is already heading towards, while giving the impression that Scotland is being targeted.

And the settlement is political as much as fiscal. Days before the 2014 independence referendum, the leaders of the three main UK parties promised in a joint statement, the "Vow", that Barnett would continue, and the Smith Commission that followed agreed that "the Barnett Formula should continue to determine the underlying block grant".¹⁶ Reopening it risks being read in Scotland as breaking that promise, at a real cost to the Union, for a gain that rests on an estimate of need nobody can make precise. On this view, a settlement that is stable, broadly accepted and slowly converging is worth more than a fairer one that is constantly contested.

This case deserves weight. It is why this pillar proposes extending the existing needs-based factor rather than replacing Barnett, rules out cash cuts, and asks for the assessment to be independent and published before any funding moves. Where it falls short is that it accepts the absence of evidence as a reason to avoid gathering it, and offers nothing to England's poorest regions, which have neither Barnett nor a floor.

7. Counter-Arguments

'This is an attack on Scotland'

It is a proposal to apply to Scotland the same rule that already applies to Wales and Northern Ireland, and to England's regions a rule they have never had. If an independent assessment finds that Scotland's sparsity, health and age structure justify its current funding, the needs factor would change little. If it finds Scotland funded above need, the proposal still rules out cash cuts and applies only to future increases. Scotland's own tax decisions, which add about £1 billion to its budget, are untouched.⁶

'A needs formula just moves the argument somewhere else'

It does, and that is the point. An argument about which indicators measure need, conducted in public with the method published, is more productive than an argument about whether a 1978 baseline is fair, which can never be settled. Northern Ireland's 124% figure shows how this can work: it came from an independent body's published assessment, the method is open to challenge, and it is used in the formula today.⁴

'The 2014 commitment means Barnett cannot be changed'

The commitment, as written into the Smith Commission's report, was that Barnett should continue to determine the underlying block grant.¹⁶ The proposal here keeps that: Barnett increments continue, and a needs-based factor is added to them, as has already been done for Wales and Northern Ireland without either being described as abolishing Barnett.⁴ Any change would still need to be agreed through Scotland's fiscal framework, as the Welsh and Northern Ireland factors were.

'GERS proves Scotland is subsidised, so its funding should simply be cut'

GERS shows Scotland raises about the UK average in revenue per person and receives more in spending.⁸ It does not show what Scotland needs, and the same kind of gap exists for most of England: only London and the South East raised more than was spent on them in 2022-23.⁹ A cut based on GERS alone would be the same mistake as the formula itself, setting funding without measuring need.

'England's regions should just get Scottish-style devolution and their own Barnett grant'

Giving English regions more control over their own spending is what the 2026 Act begins to do, and this pillar supports it.¹³ But copying Barnett for England would copy its central flaw: a share of spending set by population and history rather than need. The better model is the one this pillar proposes for the nations, with England's strategic authorities funded on the same published measure.

9. Proposals for Change

These are the evidence-based proposals of this pillar, put forward for public discussion and challenge.

• P1 Establish an independent advisory body, on the lines the House of Lords committee recommended in 2009, to publish each nation's and English region's relative need for public spending every Spending Review, using a small set of published indicators such as age structure, income, health and sparsity, with its method open to challenge and peer review.

• P2 Extend the needs-based factor already applied to Wales and Northern Ireland to Scotland, agreed through Scotland's fiscal framework, so that Barnett increases move each devolved government's funding toward its assessed need, applied only to future increases and never through a cash cut.

• P3 Put the Barnett formula and the needs-based factors into statute, with disputes referred to an independent body rather than settled by the Treasury alone, and publish each devolved government's funding per person relative to comparable spending in England every year.

• P4 Set the integrated funding settlements of England's mayoral strategic authorities using the same published needs assessment, so that the poorest English regions are funded on the same basis as the devolved nations.

• Publish the first assessment before any funding changes, with at least one Spending Review in which it is used only for information, so the method can be tested and challenged before money depends on it.

• Keep the devolved governments' own tax decisions, and the money they raise or forgo as a result, entirely outside the needs calculation.

Nobody designed the current settlement to be unfair, and nobody can show whether it is. A formula meant to last a few years has lasted nearly fifty, and two of the three nations it funds have had to prove they were short of money to get a floor. The rest of the UK, including its poorest regions, has never been measured at all. The Generational Reset does not claim to know what the right numbers are. It argues that the UK should find out, publish the answer, and apply the same rule everywhere.

The Generational Reset is a non-partisan, public-interest project. It is not affiliated with any political party, does not accept corporate funding, and publishes all its work under open licence for public discussion and adaptation.

For public discussion. Not affiliated with any political party. | generationalreset.org

The Generational Reset | S1_11: Devolved Governments | For public discussion. Not affiliated with any political party. | generationalreset.org