Every month the British government takes in around £100 billion and spends around £113 billion. The difference is borrowed. The people who will repay it are, for the most part, too young to vote — or not yet born. This is the story of how a country learned to spend tomorrow’s money today, told entirely through the government’s own figures.

Most arguments about public spending are arguments from ideology. One side wants a bigger state, the other a smaller one, and the numbers are conscripted to serve the conclusion each side reached before it started. This article does the opposite. It argues from the evidence — from the Office for Budget Responsibility, the Office for National Statistics, HM Treasury, HM Revenue and Customs, the Department for Work and Pensions and the Bank of England — and it lets those numbers, all of them official, tell a story that neither tribe finds comfortable.

Part One: Where We Are Now

In the 2025-26 financial year, according to the Office for National Statistics, total public sector current receipts — every pound the state collected from every source — came to £1,232 billion. Income tax was the largest single source at around £330 billion, followed by National Insurance contributions at around £205 billion and VAT at around £180 billion. Add corporation tax at around £100 billion and those four taxes together account for roughly two-thirds of everything the government raises.

This is not a small state starved of funds. At around 40% of GDP, the tax burden is now the highest it has been since 1982-83, according to the House of Commons Library. If the British state were a restaurant, the bill would be eye-watering, the portions generous, and the management would be putting the shortfall on a credit card in a child’s name. Whatever else is true, it is not true that the problem is a government that refuses to tax.

Figure 1

Where the Money Comes From: UK Government Revenue 2025-26

Four taxes account for roughly two-thirds of everything the government raises

Source: ONS, Public sector current receipts: Appendix D

In the same year, Total Managed Expenditure — everything the state spent — was £1,360 billion, around £113 billion a month, 44% of GDP. The largest area was social protection — the state pension and working-age welfare — at around £333 billion, of which the state pension alone accounted for £146 billion. Health came next, at around £250 billion. Then came debt interest: close to £100 billion, with the ONS putting central government interest at £97.6 billion.

Figure 2

Where the Money Goes: UK Government Spending 2025-26

Debt interest — in red — now costs more than defence

Source: HM Treasury (PESA), ONS

Set the two side by side and the arithmetic is unforgiving. The gap between receipts and spending — public sector net borrowing — came to around £130 billion, more than 4% of GDP. In the opening months of 2026-27 it has run harder still: the ONS recorded borrowing of £24.3 billion in April 2026 and £23.3 billion in May, the second-highest May on record.

All of that borrowing accumulates. Public sector net debt reached £2,989.9 billion at the end of June 2026 — for all practical purposes, £3 trillion — equal to 94.9% of GDP. The ONS notes, month after month, that this is the highest level relative to the economy since the early 1960s.

Figure 3

The Gap That Has to Be Borrowed: UK Annual Deficit 1990-2026

After each crisis the debt ratchets up and never comes back down

Source: ONS, HM Treasury

Britain now spends more servicing its debts than it spends on defence. The interest bill alone is larger than the entire schools budget in England. It buys nothing. It builds no hospital, trains no soldier, teaches no child. It is the rent on decisions already made.

Part Two: Who Actually Pays

There is a comfortable story in British politics that the country’s problems could be solved if only “the rich paid their fair share.” The government’s own figures tell the opposite story. The rich are already paying most of the bill — and the base of people who pay more than they take is far narrower, and far more mobile, than almost anyone admits.

Start with income tax. According to HMRC’s Income Tax liabilities statistics, in 2023-24 the top 1% of taxpayers paid 27% of all income tax. The top 10% paid 59%. The bottom half paid under 10% between them.

This concentration has been rising. In 1999-2000 the top 10% paid around 50% of income tax; today it is nearly 60%. Thomas Piketty documented the broader pattern in Capital in the Twenty-First Century: when returns on capital outpace economic growth, wealth and income concentrate at the top. The British tax base has followed that logic precisely. The state has come to lean ever more heavily on an ever-smaller number of shoulders.

Capital gains tax tells the same story in a more extreme form. In 2023-24, all of CGT was paid by around 378,000 people — roughly 1% of income taxpayers. Within that, just 2,000 individuals with gains above £5 million paid about 40% of the entire individual CGT take. A group smaller than the crowd at a local amateur football match provides four in every ten pounds of a national tax.

Figure 4

The Load Is Carried by a Few: Income Tax by Percentile

One in ten taxpayers provides nearly two-thirds of income tax revenue

Source: HMRC, Income Tax liabilities statistics 2023-24

Turn the telescope around. When the ONS accounts for the whole system — not just the taxes people pay, but the cash benefits, the NHS, the schooling and everything else they receive — 53% of individuals live in households that receive more from the state than they pay into it. Only the top four income deciles are net contributors. The top tenth of households pays, on average, more than £65,000 a year in taxes and receives around £14,000 back — a net contribution of over £50,000 each, every year. The entire edifice of the modern British state rests on the net contributions of roughly the top third of the population, and disproportionately on the top few per cent within that.

It is important to be fair about what “net recipient” means. It includes the pensioner drawing the state pension she paid National Insurance for across a forty-year working life. It includes families using the NHS and state schools they will pay for later. This is not a moral indictment; it is the arithmetic of a redistributive system working as designed. But the arithmetic has a consequence: a state this large can only be financed by a narrow band of high earners, and that band is not captive.

High earners are the most mobile people in the economy. When Britain abolished the “non-dom” tax status in April 2025, the OBR modelled significant departures — roughly a quarter of non-doms with trust structures and around 10% of those without — and attached “very high uncertainty” to the revenue forecast. Early independent assessments suggest departures have run at least in line with, and possibly ahead of, the OBR’s assumptions.

When a person who pays £250,000 a year in tax departs, it takes several dozen average taxpayers to replace them. The burden does not vanish. It rolls downhill, onto the broad middle class, who cannot so easily relocate their lives. “Tax the rich” is not wrong because the rich are sacred. It is wrong because there are not enough of them, they already pay most of the bill, and they can leave.

Part Three: How We Got Here

A century ago — 1925

A hundred years ago, the British state was a different creature. According to the Bank of England’s long-run dataset, A Millennium of Macroeconomic Data, total government spending in the mid-1920s was around 30% of GDP — roughly two-thirds of today’s share. There was no National Health Service and no comprehensive welfare state. The old-age pension, introduced in 1908, was small and means-tested and did not begin until the age of 70, at a time when most people did not live to collect it for long.

What did the state spend on? Overwhelmingly, two things: defence, and the interest on the debts of the First World War. Public sector net debt stood at roughly 175% of GDP. The OBR’s own history records that during the interwar years debt interest swallowed around a quarter of all government revenue. Winston Churchill, as Chancellor, returned Britain to the gold standard in 1925 at the pre-war parity — a decision Keynes called “the economic consequences of Mr Churchill” — precisely because the government believed that only a hard currency could make the war debt manageable. It crushed British exports, threw miners out of work, and triggered the General Strike of 1926. That is what it looks like when a nation tries to honour a debt it can barely carry.

As Niall Ferguson argues in The Cash Nexus, the history of sovereign finance is a history of states borrowing their way out of the consequences of previous borrowing — a treadmill that accelerates until either the runner finds a new source of energy or collapses. (For the full pattern — from Philip II’s serial defaults to Rome’s debasement — see The Debasement.)

But note the crucial difference. The 1925 debt had been incurred to win a war of national survival — a debt a generation could look its children in the eye and defend.

Forty years ago — 1985

By the mid-1980s the state had transformed. The welfare state built after the Beveridge Report — the NHS launched in 1948, National Insurance, council housing, universal education — was fully mature. Government spending had roughly doubled as a share of the economy, to around 41% of GDP.

And yet public sector net debt was only around 39% of GDP, and falling. A far bigger, more generous state was running on a debt burden less than a quarter the size of 1925’s. The trick was sustained economic growth and a young, expanding workforce. The economy grew faster than the debt, and so the debt shrank in relative terms even as the state expanded. That combination — growth outpacing borrowing — is the quiet assumption on which the entire modern welfare state was built.

Twenty years ago — 2005

On the eve of the financial crisis, spending was around 38.7% of GDP and debt about 36%. The composition of the state had inverted since 1925: where once it existed mainly to defend the realm, it now existed mainly to provide health, pensions and welfare.

Then came two shocks. The financial crisis of 2008 added some 28 percentage points of GDP to the national debt. COVID-19 added a further 14. Debt as a share of GDP more than doubled, from 36% to the 95% of today. After each crisis, the debt never came back down. The ratchet only ever turns one way.

Part Four: Where We Are Going

Figure 5

A Century of Debt — and Where It Leads

UK public sector net debt as % of GDP, with OBR 50-year projection

Source: ONS, Bank of England, OBR Fiscal Risks & Sustainability 2026

Every two years the Office for Budget Responsibility publishes a Fiscal Risks and Sustainability report that projects the public finances over fifty years. Its July 2026 edition deserves quoting plainly.

On current policy, the OBR projects public sector net debt rising from around 95% of GDP today to roughly 300% by the mid-2070s. Net interest spending alone climbs toward 12% of GDP — more than the entire NHS costs today. The downside scenario, incorporating recurring shocks and the vicious feedback loop of debt begetting interest begetting more debt, produces paths several times higher.

The debt explodes not because of waste or fraud or foreign aid — rounding errors in a £1.3 trillion budget — but because of demography colliding with three specific promises. Health spending is projected to rise from around 8% of GDP to 13% as an ageing population consumes more care. The state pension is projected to rise from around 5% to around 9% — and the OBR is unusually blunt about the cause. A large part of that increase is the triple lock, the guarantee that the state pension rises each year by the highest of inflation, earnings, or 2.5%. The OBR estimates the triple lock will cost around £15.5 billion a year by 2029-30, roughly three times the figure predicted when it was introduced in 2010, and that over the long run it costs some 2 percentage points of GDP more than simply linking pensions to earnings. That is a policy choice, not a law of nature, and it is one made by today’s voters at the expense of tomorrow’s taxpayers.

Stack health, care and pensions together and the OBR has them rising by close to 10 percentage points of GDP over fifty years. There is no plausible tax increase that covers a gap of that size without crushing the economy that generates the taxes.

Part Five: The Growth Assumption That No Longer Holds

Figure 6

The Engine That Stalled: UK Fertility Rate 1960-2024

Britain has been below replacement fertility continuously since 1973

Source: ONS, Births in England and Wales; OBR

Everything described so far rests on a single hope: that the economy will keep growing fast enough to outrun the debt, exactly as it did between 1945 and 2005. The demographic foundation beneath that hope has quietly collapsed.

The total fertility rate in England and Wales fell to 1.41 children per woman in 2024, according to the ONS — the lowest figure ever recorded. The replacement rate is 2.1. Britain has been below it continuously since 1973. (For the full demographic picture, see The Empty Cradle Bargain.) From 2026, the ONS projects deaths outnumbering births in the United Kingdom. Natural change turns negative. From this point on, net migration is the only source of population growth.

The old-age dependency ratio — pensioners relative to working-age people — has held roughly steady at around 30% since the 1970s. The OBR projects it rising above 40% by 2075: from roughly three-and-a-half working-age adults for every pensioner toward two-and-a-half, and then fewer.

Between 1945 and 2005, a growing, youthful workforce meant the economy expanded faster than the debt and the burden per worker fell. Reverse the demography and the machine runs backwards. The assumption that “we’ll grow our way out of it,” true for two generations, becomes a comforting fiction for a third.

Part Six: The Immigration Answer That Isn’t

For thirty years, the standard answer to Britain’s demographic problem has been migration. If the natives will not have enough children, import working-age taxpayers. It is worth taking seriously, because the half that is wrong is the half that matters.

The OBR, in its 2024 fiscal risks analysis, modelled the lifetime fiscal contribution of representative migrants by earnings. A high-wage migrant arriving at 25 is a substantial net contributor — on the order of a million pounds over a lifetime. But a low-wage migrant is, the OBR concluded, “probably not a net contributor at any age.” The Migration Advisory Committee sharpens the picture: skilled workers outside health and care contribute nearly £690,000 over a lifetime, while care workers specifically are modelled at a net lifetime cost of around £36,000. And net migration surged to a revised peak of around 944,000 in the year to March 2023 — overwhelmingly through the lower-wage and dependant-heavy routes that the official modelling identifies as fiscal costs. The composition was wrong.

As the contributors to Borderless Welfare State have documented across multiple countries, the fiscal arithmetic of migration is inseparable from the generosity of the welfare state that receives the migrants — a generous system with open borders is a contradiction that resolves itself through fiscal pressure.

The decisive point comes from the OBR itself. Even in its optimistic, high-migration, high-earnings scenarios, immigration changes the level of debt in a given year but “does not fundamentally change the long-run debt dynamics.” A migrant who arrives at 25 to pay for today’s pensions is themselves 25 years closer to drawing a pension. You cannot permanently outrun an ageing problem by importing people who also age.

There is, however, one variable that neither the OBR’s projections nor the immigration debate adequately accounts for: productivity growth driven by artificial intelligence and robotics. If the demographic problem is too few workers supporting too many dependants, the question is not only how many workers there are but how productive each one can be. Milton Friedman argued in Capitalism and Freedom that the free market’s capacity for innovation is the only reliable engine of broad-based prosperity — and AI may prove the most powerful such engine since electrification. A care worker assisted by AI diagnostics and robotic lifting can do the work that today requires two or three. The OBR’s projection assumes productivity growth of roughly 1.5% per year. If AI raises that to 2.5% or 3%, the debt-to-GDP path flattens and the dependency ratio bites less hard. None of this is guaranteed, and the history of productivity forecasts is a history of disappointment. But it is the one credible pathway by which Britain could maintain its commitments to the old without crushing the young — not through importing people who also age, but through making each worker dramatically more productive. (For the full case, see The Robot Bargain.)

Part Seven: The Price of Money

One more piece of the machine, and it is the one least understood.

Since 2009 the Bank of England has bought and sold hundreds of billions of pounds of government bonds through quantitative easing and its reverse, quantitative tightening. During the years of near-zero rates, QE earned the Treasury a cumulative profit of roughly £120 billion. But once the Bank raised rates to fight inflation, the machine reversed: the Bank now pays out more on the reserves backing its bond holdings than those bonds earn, and — because the Treasury indemnifies the whole scheme — the losses land on the public finances. The OBR estimates these losses at around £17 billion in a single year (2023-24) and more than £100 billion across the forecast period. On top of that, the Bank is actively selling gilts back into the market at a loss, at the very moment the government is trying to sell more of its own.

The case for central bank independence is a serious one: hand politicians direct control of interest rates and money creation, and they face a permanent temptation to print their way out of trouble — a temptation that has produced catastrophic inflation everywhere it has been indulged, from Weimar Germany to modern Argentina. But the post-1997 settlement — total operational independence, an ever-expanding balance sheet, its costs socialised automatically onto the Exchequer — was designed for a world of modest government debt. That world is gone. The government’s cost of borrowing is now too large a share of national life to sit entirely beyond democratic scrutiny.

And that leads to a principle that ought to govern the whole of this. There is an old idea in British fiscal policy, the “golden rule”: that a government should borrow only to invest, and pay for its day-to-day spending out of taxation. Borrowing to build something lasting — a railway, a power station, a port — is defensible, because the asset serves the future generation that will help repay it. Borrowing to fund this year’s pensions, this year’s salaries, this year’s running costs, is not investment. It is consumption charged to a credit card that our children will inherit. Britain today does the opposite. The great bulk of its borrowing funds not the assets of the future but the consumption of the present.

Part Eight: The Difficult Decisions

If the trajectory is unsustainable — and the OBR, an official body with no axe to grind, says in plain words that it is — then it will not be sustained. The only choice is whether to change course deliberately, while we still can, or to have the change imposed by a bond market that one day decides Britain’s promises are no longer credible.

Deliberate change means confronting things that are, at present, close to unsayable. It means asking whether the triple lock can survive in a country where there are ever fewer workers per pensioner. Whether the state retirement age must rise faster as longevity rises. Whether a health service designed for the demography of 1948 can be financed unreformed through the demography of 2075. It means a migration policy built around fiscal contribution rather than raw numbers, since the government’s own models show that who comes matters far more than how many. It means protecting the narrow base of high earners who already fund most of the state, rather than squeezing them until they emigrate. And it means borrowing, if at all, to build assets that earn their keep — not to pay today’s bills.

Above all it means an honest conversation about the fact that the current level of spending is not affordable at any politically plausible level of tax.

None of this is comfortable. All of it is easier now than it will be later, because the arithmetic only worsens with delay. Every year the deficit is not closed, the debt grows, the interest bill grows, and the number of future workers who must service it shrinks.

Conclusion

In 1776 the American colonists coined a phrase for a system in which people are taxed by a body they had no say in electing: taxation without representation. Britain has built its mirror image. The people who benefit from today’s spending — the pensioner drawing a triple-locked pension, the voter enjoying public services not fully paid for by current taxation — are, disproportionately, today’s electorate. The people who will repay the £3 trillion, and the interest on it, are tomorrow’s workers, who had no vote on any of it.

A debt incurred to build something lasting is an investment. A debt incurred to avoid difficult choices today is a transfer — from the powerless future to the comfortable present.

History does not offer many examples of nations that borrowed their way out of a demographic decline. It offers a great many examples of nations that borrowed until the borrowing stopped being their choice. Which of those we become is still, for a few more years, up to us. It will not be up to us for long.

The figures in this article are drawn from the Office for National Statistics, the Office for Budget Responsibility, HM Treasury, HM Revenue and Customs, the Department for Work and Pensions, the Migration Advisory Committee, the Bank of England and the House of Commons Library. Where estimates are contested — as with the fiscal impact of migration, or the scale of high-earner departures — the range and the uncertainty have been stated rather than hidden. The argument is not that any one number is beyond dispute. It is that the direction of every number points the same way.

Sources