Headline evidence · Public finances · United Kingdom · 1995–96 to 2025–26 · latest provisional

Debt interest takes 8.8p of every £1 received—compared with 9.2p 30 years ago.

In 2025–26, central-government debt interest was equivalent to 8.8% of public-sector current receipts, compared with 9.2% in 1995–96. Receipts include taxes and other public income. The interest measure is net of the Bank of England’s asset purchase scheme. Latest-year figures are provisional; this comparison uses the September 2026 data edition.

Proposed outcome: Find room in the budget for priorities while protecting useful services and preparing for emergencies. What could change →

Geography
United Kingdom
Evidence period
1995–96 to 2025–26; latest provisional
Evidence status
Site calculation from published data, provisional
Editorial review
Reviewed 22 October 2026

The short version

How much of public income is equivalent to the interest bill?

8.8p per £1 received, compared with 9.2p 30 years earlier

Central government debt interest, net of the Bank of England’s asset purchase scheme, divided by whole-public-sector current receipts. Receipts include taxes and other income. This is an equivalent share, not money earmarked for interest. Solid = outturn from the September 2026 OBR Databank; dashed = OBR forecast from the March 2026 Economic and fiscal outlook, which predates the latest outturn.

United Kingdom · 1990–91 to 2025–26 outturn (latest provisional); 2026–27 to 2030–31 OBR forecast, dashed

1990–91 to 2025–26 · observed full financial years; latest provisional. Dashed: OBR forecast, 2026–27 to 2030–31. Axis years show the financial-year end.

Pence per £1 received0p5p10p15p19952005201520258.8pPeak 10.8p, 2022–23Low 3.2p, 2020–21
The share fell to 3.2p in 2020–21, then rose to 10.8p in 2022–23. It is now slightly below its level 30 years earlier—not a steady 30-year increase. Interest rates, inflation-linked debt, the amount owed and receipts all affect the result. These are observed full financial years, not forecasts. The share peaked at 10.8p in 2022–23, fell as low as 3.2p in 2020–21 and, before that, 4.7p in 2019–20. The OBR forecast is 8.4p in 2026–27, 8.5p in 2027–28, 8.8p in 2028–29, 9.0p in 2029–30, 8.8p in 2030–31: a forecast, not an outturn.

United Kingdom · 1990–91 to 2025–26 outturn (latest provisional); 2026–27 to 2030–31 OBR forecast, dashed · Calculated from published outturn; latest year provisional; forecast years are projections that depend on gilt yields, inflation and receipts

Provisional outturn may be revised. One September 2026 workbook supplies the entire history; newer monthly ONS figures are not spliced into it.

Show the workings — 8.8p per £1 received, compared with 9.2p 30 years earlier
Geography:
United Kingdom — Central-government interest / whole-public-sector current receipts
Measure:
Pence per £1 of receipts
Period:
1990–91 to 2025–26 outturn (latest provisional); 2026–27 to 2030–31 OBR forecast, dashed
Status:
Calculated from published outturn; latest year provisional; forecast years are projections that depend on gilt yields, inflation and receipts
Calculation:
Central government debt interest net of APF (£bn) / public sector current receipts (£bn) * 100
Compare with:
1995–96: £26.531bn / £287.101bn × 100
Uncertainty:
Provisional outturn may be revised. One September 2026 workbook supplies the entire history; newer monthly ONS figures are not spliced into it.
Not included / breaks:
Interest is central government, while receipts cover the whole public sector. Includes non-tax receipts; excludes debt principal repayments. A ratio, not inflation-adjusted cash spending or a count of creditors.

Source: OBR Public finances databank, September 2026, Aggregates (£bn): matched debt-interest and current-receipts columns (released 24 September 2026).

View the data as a table
Financial yearPence per £1 receivedStatus
1990–918.6Outturn
1991–927.5Outturn
1992–938.0Outturn
1993–948.4Outturn
1994–958.8Outturn
1995–969.2Outturn
1996–979.4Outturn
1997–988.9Outturn
1998–998.3Outturn
1999–006.8Outturn
2000–016.6Outturn
2001–025.6Outturn
2002–035.2Outturn
2003–045.1Outturn
2004–055.4Outturn
2005–065.3Outturn
2006–075.6Outturn
2007–085.8Outturn
2008–095.9Outturn
2009–105.0Outturn
2010–116.8Outturn
2011–127.0Outturn
2012–136.1Outturn
2013–145.7Outturn
2014–154.9Outturn
2015–164.8Outturn
2016–174.9Outturn
2017–185.5Outturn
2018–194.8Outturn
2019–204.7Outturn
2020–213.2Outturn
2021–226.0Outturn
2022–2310.8Outturn
2023–249.8Outturn
2024–259.3Outturn
2025–268.8Provisional outturn
2026–278.4OBR forecast
2027–288.5OBR forecast
2028–298.8OBR forecast
2029–309.0OBR forecast
2030–318.8OBR forecast

Three key measures

Who feels this most?

National totals hide where pressure lands. The NAO found that English councils’ real funding rose 4% from 2015–16 to 2023–24, but did not keep pace with population, service demand or costs. For 2026–27, 36 councils were given in-principle exceptional financial support: permission to borrow or use capital receipts for day-to-day costs, rather than new grant. The older funding comparison below explains how the preceding decade affected poorer areas. NAO review · MHCLG support list.

Budget 2026 · 28 October 2026

What the Budget did to the nation’s finances

£14.5bn

Safety margin against the main fiscal rule

about half the average since 2010 (£27bn)

£7.6bn

Net extra borrowing from decisions, 2026–27

Largest: Household Energy Payment, Defence Investment Plan (DIP), Support for Energy Bills

£5.2bn

Net raised by decisions, 2029–30

Largest: HMRC, Writing-down Allowances, Distributions

£2.8bn

Over the welfare cap

OBR forecast for 2029–30

The charts below still use the OBR’s March 2026 forecast years until the OBR publishes its October 2026 Databank.

Measures aimed at unpaid tax and lost money

The Budget in 2 minutes →Is it being delivered? →

Treasury costings certified by the OBR, not measured results.

The evidence

Watch first: the national debt in a minute and a half

Britain's national debt has passed £3 trillion. What does that mean for you? · 1:31 · captioned

The film opens with ONS's newest estimate: £3.005 trillion, 94.0% of GDP at the end of September 2026 (provisional). The charts below run to the end of March 2026 (93.3%). Its interest figures are the OBR's October 2026 forecast, £123.9bn in 2026–27 rising to £160bn in 2031–32; the forecast line on this page still uses the March 2026 forecast (£109.4bn for 2026–27).

TranscriptCaptions (SRT)Sources and checksFilm page

Money in, money out

Start with one year. Receipts and spending are both flows — money in and out over twelve months — so they can sit on one scale, and the gap between them is what the state borrowed. The debt is different: it is everything still owed, a stock built up by past borrowing. Never add it to these figures; compare it with them only as a ratio.

Where does the annual public budget come from — and where does it go?

In 2025–26 the state spent £1,366bn, raised £1,231bn and borrowed the £134bn gap

United Kingdom · Financial year 2025–26 (April to March)

The government spent more than it raised and borrowed the difference. Interest payments now take about 8p of each pound spent. These totals are provisional.

United Kingdom · Financial year 2025–26 (April to March) · Provisional outturn: ONS revises the latest year for several months

Show the workings — In 2025–26 the state spent £1,366bn, raised £1,231bn and borrowed the £134bn gap

Public sector current receipts by source and total managed expenditure by type, 2025–26, £bn cash, drawn on one scale. Borrowing is the difference.

Four taxes — income tax, National Insurance, VAT and corporation tax — raise about 69% of the money. On the other side, interest is now about 8p in every £1 the state spends, more than net investment in roads, schools and hospitals combined. Borrowing is only the gap between two very large totals, so a small percentage change in either moves it a lot. These are provisional figures.

Geography:
United Kingdom — Whole public sector: central government, local government and public corporations
Measure:
£ billion, cash (nominal)
Period:
Financial year 2025–26 (April to March)
Status:
Provisional outturn: ONS revises the latest year for several months
Calculation:
Receipt categories are UK Facts groupings of the Databank’s tax columns and sum to the published total. Day-to-day spending excluding debt interest = public sector current expenditure minus central government debt interest net of APF.
Compare with:
Previous year’s borrowing: £149.8bn
Not included / breaks:
Interest paid by local government and public corporations, and the difference between gross and net-of-APF interest, sit in the day-to-day line. The debt stock is not part of this picture: it is a separate tile and chart. VAT includes VAT refunded to public bodies, which the national accounts count as receipts.

Source: OBR, Public finances databank (September 2026); outturn consistent with ONS/HM Treasury Public sector finances, 22 September 2026, sheets "Aggregates (£bn)", "Aggregates (per cent of GDP)" and "Receipts (£bn)" (released September 2026).

View the data as a table
Item2025–26, £bn
Receipts: Income tax329.3
Receipts: National Insurance204.4
Receipts: VAT210.8
Receipts: Corporation tax99.6
Receipts: Council tax50.4
Receipts: Other taxes and duties213.1
Receipts: Interest, dividends and operating surplus123.6
Receipts: total1231.3
Spending: Day-to-day spending on services, benefits and pay1097.5
Spending: Debt interest (central government, net of APF)108.9
Spending: Depreciation of public assets72.3
Spending: Net investment86.9
Spending: total managed expenditure1365.6
Public sector net borrowing (spending minus receipts)134.3

What goes unpaid or lost

HMRC estimates £59.2bn of tax went unpaid in 2024-25, 6.4% of the tax due (provisional). That is not a pot of money waiting to be collected: the costed compliance package expects about £10bn a year by 2029–30, and that is already counted in the official forecasts. Separately, the Covid Counter Fraud Commissioner estimates about £10.9bn was lost to fraud and error in Covid support and £1.8bn has been recovered so far. This section replaces the former Follow the money page; its old links arrive here.

Unpaid tax by type of tax, 2024-25

  • Corporation tax£21.0bn
  • Income tax, National Insurance and capital gains tax£20.9bn
  • VAT£12.1bn
  • Excise duties£2.9bn
  • Other taxes£2.3bn

Covid support lost to fraud and error, and recovered

  • Estimated lost to fraud and error£10.9bn
  • Recovered so far£1.8bn

Different years and bases: the tax gap is an annual estimate; the Covid figures are one-off losses over the pandemic. They are not added together or netted against spending. Tax options shows what changes to tax rates could raise.

Where the spending goes

Where does the spending go?

Benefits, pensions, social services and health take 49% of all public spending

United Kingdom · Financial year 2025–26 (outturn)
  • Social protectionPensions, benefits, social care£407.3bn
  • Health£257.5bn
  • Public sector debt interestPESA gross measure£130.3bn
  • Education£125.7bn
  • Economic affairsIncluding transport£94.0bn
  • DefenceUN COFOG definition, not NATO£65.4bn
  • Public order and safety£55.7bn
  • Other general public services£35.2bn
  • Housing and community amenities£22.5bn
  • Environment protection£19.2bn
  • Recreation, culture and religion£15.1bn
  • Accounting adjustmentsNot a service: national-accounts items that reconcile spending on services to the total£132.5bn

Health, pensions and benefits account for about half of public spending. Interest also takes a large share, leaving less room for other priorities.

United Kingdom · Financial year 2025–26 (outturn) · Accredited official statistics; the latest year is provisional outturn

Show the workings — Benefits, pensions, social services and health take 49% of all public spending

Public sector expenditure by function (UN COFOG), 2025–26, £bn cash, with debt interest shown separately from other general public services. Total managed expenditure in this publication: £1,360.1bn.

Spending is concentrated: pensions and benefits (social protection) and health together are about 49% of the total. Debt interest, on PESA’s gross measure, is larger than spending on education or defence. These categories describe what money buys, not which department spends it, and they overlap with no other figure on this page except the total.

Geography:
United Kingdom
Measure:
£ billion, cash (nominal)
Period:
Financial year 2025–26 (outturn)
Status:
Accredited official statistics; the latest year is provisional outturn
Calculation:
“Other general public services” = general public services minus public sector debt interest. EU transactions (−£0.2bn) not shown.
Compare with:
Total managed expenditure £1,360.1bn (PESA, July 2026)
Not included / breaks:
PESA was published in July; FISCAL-01 uses the September public sector finances, which revise spending to £1,365.6bn. PESA’s debt interest (£130.3bn) is gross public sector interest, not the net-of-APF measure used in the headline. COFOG defence excludes military pensions, so it is lower than the NATO measure.

Source: HM Treasury, Public Expenditure Statistical Analyses 2026, Table 4.2 (released 16 July 2026).

View the data as a table
Function2025–26, £bn
Social protection407.3
Health257.5
Public sector debt interest130.3
Education125.7
Economic affairs94.0
Defence65.4
Public order and safety55.7
Other general public services35.2
Housing and community amenities22.5
Environment protection19.2
Recreation, culture and religion15.1
Accounting adjustments132.5

Who receives social security spending, and why it grows, is on the benefits and pensions pages; health spending on the NHS funding page.

Why interest costs change

How has the interest bill changed, and how much of the budget does it take?

Interest has been above £100bn for four years and is forecast to keep rising in cash terms

United Kingdom · 2000–01 to 2025–26 outturn; 2026–27 to 2030–31 forecast

£ billion, cash

1996–2031 · dashed lines show forecasts. Shading marks years without figures: this series starts in 2001.

£bn, cash£0bn£50bn£100bn£150bn2000201020202030£137bn2025–26 provisional

Share of public sector current receipts

1996–2031 · dashed lines show forecasts. Shading marks years without figures: this series starts in 2001.

% of receipts0%5%10%15%20002010202020308.8%

Debt interest rose sharply as prices and interest rates increased. It is expected to rise further in pounds, though not much as a share of government income.

United Kingdom · 2000–01 to 2025–26 outturn; 2026–27 to 2030–31 forecast · Outturn to 2025–26 (latest year provisional); forecasts from the OBR’s March 2026 Economic and fiscal outlook, which predates the latest outturn

Show the workings — Interest has been above £100bn for four years and is forecast to keep rising in cash terms

Central government debt interest net of the Asset Purchase Facility: £bn cash (top) and as a share of public sector current receipts (bottom), financial years ending 2001 to 2031. Solid = outturn; dashed = OBR forecast.

Interest fell to £25.2bn in 2020–21, when rates were near zero and inflation low, then rose more than four times over as inflation pushed up payments on index-linked gilts and Bank Rate rose. As a share of receipts it peaked at 10.8% in 2022–23 and was 8.8% in 2025–26 — high by the standards of the last 25 years and similar to the mid-1990s. The OBR forecast rises in cash, but stays near 9% of receipts.

Geography:
United Kingdom
Measure:
£ billion, cash; and % of public sector current receipts
Period:
2000–01 to 2025–26 outturn; 2026–27 to 2030–31 forecast
Status:
Outturn to 2025–26 (latest year provisional); forecasts from the OBR’s March 2026 Economic and fiscal outlook, which predates the latest outturn
Calculation:
As published. The share of receipts is debt interest divided by public sector current receipts, both from the Databank.
Compare with:
2020–21 low (£25.2bn); 2022–23 peak (£111.9bn)
Not included / breaks:
Cash figures are not adjusted for inflation; the lower panel removes that problem by expressing interest as a share of receipts. Forecasts depend on market interest rates and inflation at the time they were made.

Source: OBR, Public finances databank (September 2026); outturn consistent with ONS/HM Treasury Public sector finances, 22 September 2026, sheets "Aggregates (£bn)", "Aggregates (per cent of GDP)" and "Receipts (£bn)" (released September 2026).

View the data as a table
Financial yearDebt interest £bn% of receiptsStatus
2000–0126.96.6Outturn
2001–0223.05.6Outturn
2002–0321.75.2Outturn
2003–0423.25.1Outturn
2004–0525.95.4Outturn
2005–0627.85.3Outturn
2006–0730.85.6Outturn
2007–0833.75.8Outturn
2008–0933.55.9Outturn
2009–1028.15.0Outturn
2010–1141.06.8Outturn
2011–1243.57.0Outturn
2012–1338.66.1Outturn
2013–1437.85.7Outturn
2014–1534.14.9Outturn
2015–1634.54.8Outturn
2016–1736.74.9Outturn
2017–1842.85.5Outturn
2018–1939.24.8Outturn
2019–2039.34.7Outturn
2020–2125.23.2Outturn
2021–2255.56.0Outturn
2022–23111.910.8Outturn
2023–24107.49.8Outturn
2024–25106.19.3Outturn
2025–26108.98.8Provisional outturn
2026–27109.48.4OBR forecast
2027–28116.78.5OBR forecast
2028–29125.98.8OBR forecast
2029–30134.79.0OBR forecast
2030–31137.18.8OBR forecast

Scroll the table sideways to see every column.

Interest fell from 9.2p to 3.2p in every £1 over 25 years, then rose after the pandemic

Debt interest took 9.2p of every £1 received in 1995–96. By 2019–20, the year before the pandemic, it was 4.7p, with net debt at 84.5% of GDP. Debt had risen from 35.9% over the same years, so the fall came from lower interest rates, not a smaller debt.

In 2020–21 the state borrowed £310.8bn (14.7% of GDP) and debt reached 95.5% of GDP. Interest still fell to 3.2p, the lowest since this record begins in 1946–47, with Bank Rate cut to 0.1% from April 2020.

When inflation and interest rates rose, the bill climbed to 10.8p in 2022–23. It was 8.8p in 2025–26 (provisional). Interest is the cost of carrying the debt. It repays none of it: the pandemic borrowing is still owed, and is now carried at higher rates.

Higher inflation and interest rates push up the debt-interest bill. Some costs rise quickly; others rise when government borrowing is renewed. The older estimates below show how sensitive the bill can be.

How interest costs change

Three mechanisms drive the bill, and the OBR has estimated how sensitive it is to each (on its March 2025 forecast basis). About a quarter of gilts are index-linked, so their payments rise with RPI inflation: one percentage point more inflation adds roughly £7bn in a year. The Bank of England pays Bank Rate on the reserves created by quantitative easing, so one point higher Bank Rate adds about £7.4bn. Higher gilt yields act slowly, because only new and maturing debt is refinanced: about £1.5bn in the first year, £2.7bn the next.

These sensitivities explain why interest jumped when inflation and rates rose together in 2022. They are not a split of today’s bill into causes, and a lower bill would not automatically buy a fixed number of hospitals: what any saving funds is a separate decision.

The longer record, and the choices people favour

Annual data: Annual receipts, spending and borrowing

The year is the financial year END: 2026 means April 2025 to March 2026. The fixed 1996–2026 window contains published outturns, with the latest provisional. OBR forecast years are excluded here and remain explicitly dashed in the main charts. Ratios use GDP in the same publication, not inflation-adjusted pounds. A dash means no comparable observation is supplied; it is not zero. UK public sector; these are annual flows. Receipts and spending must not be added to the debt stock.

1996–2026; % of GDP. No change is calculated across definition or assessment breaks.

Annual receipts, spending and borrowing; % of GDP
YearReceiptsSpendingNet borrowing
199633.137.24.1
199732.335.43.1
199834.435.51.1
199935350
200035.734.6-1.1
200136.434.9-1.5
200235.636.10.5
200334.537.42.9
200435.338.73.4
200535.939.73.8
200636.539.73.2
200736.939.72.7
200837.2402.9
200935.7437.3
201035.84610.2
201136.845.58.7
201237.144.47.3
201336.743.87.1
201436.642.25.7
201536.541.75.2
201636.740.94.1
201737.340.12.8
201836.939.82.9
201937.239.22
202036.639.22.6
202137.652.314.7
202238.443.34.8
202339.3444.7
202439.344.24.8
202538.9445.1
202640.144.54.4
Annual data: Debt stock and the interest burden

Years identify financial-year ends, 1996–2026. Debt is the stock at end-March, with GDP centred on that date; interest is a flow over the financial year. No change is calculated between the columns. The existing interest chart retains its specified 2000–01 start; this disclosure provides the longer context. A dash means no comparable observation is supplied; it is not zero. UK public-sector net debt excluding public-sector banks; central-government debt interest net of the Asset Purchase Facility. These measures have different denominators.

1996–2026; % of GDP for debt; % of receipts for interest. No change is calculated across definition or assessment breaks.

Debt stock and the interest burden; % of GDP for debt; % of receipts for interest
YearDebt, % GDPInterest, % receipts
199635.99.2
199736.49.4
199836.58.9
199935.18.3
200032.46.8
200128.26.6
2002285.6
200329.75.2
200430.85.1
200533.45.4
200634.25.3
200734.95.6
200835.55.8
200950.25.9
201064.45
201170.66.8
201273.97
201377.26.1
201478.85.7
201581.14.9
201680.64.8
201782.74.9
201881.75.5
201979.74.8
202084.54.7
202195.53.2
202294.36
202393.110.8
202494.19.8
202593.49.3
202693.38.8

OBR September 2026 databank and its original workbooks

Keeping tax and spending unchanged was the most common answer

British Social Attitudes 2025, released February 2026. Respondents chose between reducing, maintaining or increasing taxes and spending on health, education and social benefits. These are broad preferences, not a vote on a costed budget or evidence that a particular tax would raise enough money.

Published share of respondents, %
  • Keep taxes and spending the same41%
  • Increase taxes and spending36%
  • Reduce taxes and spending19%
Survey sample, question and uncertainty

NatCen’s Appendix Table 1 gives an unweighted question base of 4,727. The 2025 survey sampled UK private-household residents aged 16 and over through randomly selected postal addresses, inviting up to two people per household to respond online or by phone. Fieldwork ran from 26 August to 5 October 2025. Institutions were excluded; people within households were not selected by a strict individual randomisation procedure. Weighting addresses selection and response differences.

The technical report estimates a 16.6% adjusted household response rate and an 11.7% adjusted individual response rate. The first-look table does not publish question-specific confidence intervals or identify a restricted GB-only weight, so none is invented here. The methods also supply a GB 18+ comparison weight; the current sample and the older population are not treated as interchangeable.

The published whole percentages are shown unchanged: the three choices total 96%; “don’t know” and “prefer not to answer” are each 1%, giving 98% across the displayed source categories. They are not rebalanced to 100%. This does not establish a majority for either higher spending or cuts. The survey moved from face-to-face to online/phone in 2020 and expanded from GB 18+ to UK 16+ in 2024.

NatCen question and Appendix Table 1 · 2025 technical methods

Annual data: Preference for higher tax and spending

Thirty published observations, 1996–2025, inside the fixed 1996–2026 window. No 2026 wave is available. Columns separate the 2020 mode and 2024 population changes. No percentage-point movement across either break or statistical significance is inferred. Other answers are shown for the latest wave above; this is not an exhaustive opinion distribution. A dash means no comparable observation is supplied; it is not zero. Separate survey waves and samples, not the same people followed over time. Unweighted question bases vary by year.

1996–2026; published whole percentage. No change is calculated across definition or assessment breaks.

Preference for higher tax and spending; published whole percentage
YearGB 18+, face-to-faceGB 18+, online/phoneExpanded UK 16+ sample
199659——
199762——
199863——
199958——
200050——
200159——
200263——
200351——
200449——
200546——
200646——
200742——
200839——
200934——
201031——
201136——
201234——
201336——
201437——
201545——
201648——
201760——
201857——
201953——
2020—50—
2021—52—
2022—55—
2023—46—
2024——40
2025——36
2026No comparable observations supplied

What the totals cannot tell us

The NAO could not give an audit opinion on the 2024–25 Whole of Government Accounts, for the third successive year, because assurance over English council accounts remained inadequate. There were 280 unaudited or disclaimed council submissions. This is a gap in assurance, not a finding that all spending was wasted or that the ONS borrowing statistics are invalid. NAO’s June 2026 report.

The OBR’s July 2026 long-term scenarios show substantial fiscal pressure under most unchanged-policy assumptions. They are conditional projections, not a dated prediction of bankruptcy. Growth, ageing, health costs and interest rates change the path. Earlier policy adjustment can reduce the eventual cost, but its distribution between taxpayers and service users is a political choice. OBR scenarios and assumptions.

Why it is happening

What changes debt relative to GDP?

How has debt changed relative to the size of the economy?

Debt rose in three steps to about 93% of annual economic output and is forecast to edge higher

United Kingdom · End-March 2000 to end-March 2026 outturn; forecast to 2031

1996–2031 · dashed lines show forecasts. Shading marks years without figures: this series starts in 2000.

% of GDP0%50%100%150%200020102020203095.1%94.3%

Debt rose sharply after the financial crisis and the pandemic. The chart compares it with the size of the economy; later years are forecasts, not money already borrowed.

United Kingdom · End-March 2000 to end-March 2026 outturn; forecast to 2031 · Latest year provisional; forecast years from March 2026, before the latest outturn

Show the workings — Debt rose in three steps to about 93% of annual economic output and is forecast to edge higher

Public sector net debt at end-March as a percentage of GDP (GDP centred on end-March): excluding public sector banks, and also excluding the Bank of England. Solid = outturn; dashed = OBR March 2026 forecast.

Debt was about a third of GDP before 2008. The financial crisis took it to about 70%, the 2010s added more, and Covid pushed it above 95%. Since then the ratio has been roughly flat because high inflation raised cash GDP. Excluding the Bank of England, whose bond purchases and lending schemes affect the headline measure, debt is lower but has risen every year since 2021–22. Forecasts are not outturns.

Geography:
United Kingdom
Measure:
% of GDP
Period:
End-March 2000 to end-March 2026 outturn; forecast to 2031
Status:
Latest year provisional; forecast years from March 2026, before the latest outturn
Calculation:
As published in the Databank “Aggregates (per cent of GDP)” sheet.
Compare with:
35.5% at end-March 2008; 84.5% at end-March 2020
Not included / breaks:
The monthly ONS bulletin publishes the ratio for later months on a rolling-GDP basis, so a figure quoted for, say, August differs from this end-March figure. That is a different date, not a contradiction.

Source: OBR, Public finances databank (September 2026); outturn consistent with ONS/HM Treasury Public sector finances, 22 September 2026, sheets "Aggregates (£bn)", "Aggregates (per cent of GDP)" and "Receipts (£bn)" (released September 2026).

View the data as a table
End of financial yearPSND ex % GDPPSND ex BoE % GDPStatus
1999–0032.432.5Outturn
2000–0128.228.3Outturn
2001–0228.028.1Outturn
2002–0329.729.8Outturn
2003–0430.830.9Outturn
2004–0533.433.5Outturn
2005–0634.234.3Outturn
2006–0734.935.1Outturn
2007–0835.535.6Outturn
2008–0950.250.3Outturn
2009–1064.463.6Outturn
2010–1170.670.3Outturn
2011–1273.974.2Outturn
2012–1377.275.9Outturn
2013–1478.876.5Outturn
2014–1581.178.7Outturn
2015–1680.678.4Outturn
2016–1782.776.8Outturn
2017–1881.773.2Outturn
2018–1979.771.9Outturn
2019–2084.576.4Outturn
2020–2195.585.5Outturn
2021–2294.381.2Outturn
2022–2393.182.4Outturn
2023–2494.185.7Outturn
2024–2593.487.9Outturn
2025–2693.389.1Outturn
2026–2794.892.3OBR forecast
2027–2895.793.5OBR forecast
2028–2996.394.5OBR forecast
2029–3096.194.5OBR forecast
2030–3195.194.3OBR forecast

Scroll the table sideways to see every column.

Debt can rise in pounds while staying flat compared with the size of the economy. New borrowing and interest add to debt; economic growth and inflation increase the income it is compared with.

How debt is measured

The ratio moves with four things: new borrowing adds to the debt; nominal GDP growth — real growth plus inflation — enlarges the denominator; the interest rate on the debt compounds it; and Bank of England operations change the headline measure. Debt can therefore rise in pounds every year while the ratio is flat, as it has been since 2021, because inflation raised cash GDP. When the interest rate on the debt is above the growth rate of the economy, holding the ratio steady needs a primary surplus — receipts above spending before interest. On the OBR Databank’s measure the UK last ran one in 2001–02; the March 2026 forecast has one returning in 2028–29.

Who is affected most

Who feels this most?

Which councils absorbed the biggest cuts?

The most deprived councils lost more than twice as much funding per person as the least deprived

England (councils) · 2010–11 to 2019–20 and to 2024–25 (2024–25 as planned in June 2024)
2010–11 to 2024–252010–11 to 2019–20
  • Most deprived tenth of councils-26%
  • All English councils-18%
  • Least deprived tenth of councils-11%

Poorer councils lost more funding per resident because they depended more on government grants. Later increases have only partly restored the loss. Protecting care services squeezed other local services.

England (councils) · 2010–11 to 2019–20 and to 2024–25 (2024–25 as planned in June 2024) · Independent analysis of official data

Show the workings — The most deprived councils lost more than twice as much funding per person as the least deprived

Change in English councils’ core funding per person, real terms: 2010–11 to 2024–25 (filled dot) and 2010–11 to 2019–20 (open dot).

Central government funding for councils fell by 46% in real terms in the 2010s. Council tax replaced only part of it, and councils serving poorer areas — which relied most on grants — lost most per resident. Increases since 2019–20 have restored only some of the gap. Councils protected statutory duties such as children’s social care; spending on culture, housing, planning and transport per person fell by over 40%.

Geography:
England (councils)
Measure:
% change per person, real terms (GDP deflator)
Period:
2010–11 to 2019–20 and to 2024–25 (2024–25 as planned in June 2024)
Status:
Independent analysis of official data
Calculation:
As published in the IFS report’s key findings; core funding = central government grants, council tax and retained business rates on a like-for-like basis.
Compare with:
Average for all English councils
Not included / breaks:
England only: local government is devolved. Excludes income from fees, charges and investments. 2024–25 was a planned figure when published.

Source: IFS, How have English councils’ funding and spending changed? 2010 to 2024 (Report R318) (released June 2024).

View the data as a table
GroupTo 2024–25To 2019–20
Most deprived tenth of councils-26%-35%
All English councils-18%-26%
Least deprived tenth of councils-11%-15%

Local figures for Hampshire’s councils: see the Hampshire pages.

Council funding over 30 years, and where every English council’s money comes from and goes.

What could change

Proposals to consider. The bodies named below would need to agree to deliver them; responsibility is not a recorded commitment. Follow the links for costs and evidence.

  • Turn investment promises into a deliverable pipeline

    Publish which investment projects are funded, when they should finish and what has actually been built.

    Who could deliver it: Treasury, infrastructure bodies, departments and devolved governments.

    Main trade-off and delivery needs

    Double-counting housing/grid benefits; shared workforce bottlenecks.

    Distinguish funded projects from aspirations; sequence resources; name owners and independent assurance; protect maintenance.

    Cost: Only part of the cost has been estimated.

    Costs, drawbacks and evidence →

  • Collect the tax already due

    Improve collection of tax that is already owed.

    Who could deliver it: the tax authority and Treasury.

    Main trade-off and delivery needs

    Due-process burdens; yield lags and double counting existing forecasts.

    Reconcile existing forecast baseline; define cases, staff, legal powers and collection horizon; audit net additional receipts.

    Cost: The extra cost is not yet known.

    Costs, drawbacks and evidence →

  • Close capital-tax reliefs rather than raise rates

    Examine removing tax breaks on capital instead of raising tax rates.

    Who could deliver it: Treasury, the tax authority and Parliament.

    Main trade-off and delivery needs

    Forestalling, avoidance and investment effects; progressivity must be demonstrated.

    Name reliefs, thresholds, interactions and transition; obtain distributional/behavioural costing and administrative assessment.

    Cost: The extra cost is not yet known.

    Costs, drawbacks and evidence →

  • Property and land tax reform

    Examine changes to taxes on property and land.

    Who could deliver it: Treasury, valuation bodies and relevant governments.

    Main trade-off and delivery needs

    Concentrated losers and valuation disputes; revenue-neutral change funds no extra spending.

    Separate council-tax revaluation, land/property tax and transaction-tax changes; model losers and deferral options.

    Cost: The extra cost is not yet known.

    Costs, drawbacks and evidence →

See all options and what is still unknown →

Some problems have no assessed solution yet: see coverage and gaps.

Domain SWOT

Recorded

Show the findings
Strengths · 3 findings

  • Independent fiscal statistics are published quickly and in detail: the OBR databank already carries 2025–26 outturn for receipts, spending, borrowing and debt interest, consistent with ONS and HM Treasury public sector finances.

    United Kingdom · 2025–26 (published September 2026)

  • Appraisal rules already require departments to separate social value from affordability and to test additionality, distribution and optimism bias.

    UK central government · February 2026

  • The yearly gap narrowed: the state borrowed £134.3bn in 2025–26, down from £149.8bn the year before, or 4.4% of GDP. This is the yearly shortfall, not the debt: any borrowing still adds to what is owed.

    United Kingdom · 2025–26 and the year before (provisional)

Weaknesses · 6 findings

  • The NAO disclaimed its opinion on the Whole of Government Accounts for a third successive year because assurance over English council data remained inadequate. This limits scrutiny of the consolidated accounts; it is not proof that all spending was wasted or that ONS borrowing statistics are invalid.

    UK consolidated accounts; English council data · 2024–25 accounts; June 2026 audit

  • Debt interest was £108.9bn in 2025–26, 8.8% of public-sector receipts, and the state borrowed £134.3bn; debt stood at 93.3% of GDP.

    United Kingdom · 2025–26

  • Those totals say how tight the budget is, not which spending is least valuable: the NAO finds spending decisions short-term and reactive, with evaluation too rarely used.

    our synthesis

  • Pressure has reached councils’ day-to-day budgets: 36 English councils were given in-principle exceptional financial support for 2026–27 — permission to borrow or sell assets to balance their budgets, not new grant.

    England · 2026–27

  • More went on debt interest (£108.9bn) than on net investment in public assets such as roads, schools and hospitals (£86.9bn) in 2025–26. Interest is central government’s, net of the Bank of England’s asset purchase scheme; investment is the whole public sector’s, after wear and tear.

    United Kingdom · 2025–26 (provisional)

  • Net debt was 93.3% of GDP at the end of 2025–26, up 57.8 percentage points since 2007–08 (35.5%) and 8.8 percentage points since 2019–20 (84.5%). A larger debt means any rise in interest rates or inflation costs more.

    United Kingdom · 2007–08 to 2025–26

Opportunities · 2 findings

  • Pricing the cost of current service standards, and showing tax packages as alternatives with their losers, would make trade-offs explicit before money is committed.

    Depends on: The Treasury publishes consistent baselines and packages on one price year, and receipts are counted once.

    our synthesis

  • Planning evaluation into large programmes from the start — a baseline, measures and a comparison — would show which spending achieves its aims. The totals on this page cannot show that.

    Depends on: Departments fund and publish evaluations, and Spending Reviews use them when deciding what to continue or stop.

    our synthesis

Threats · 6 findings

  • Most of the OBR’s unchanged-policy scenarios imply mounting long-term fiscal pressure. The paths depend on growth, ageing, health costs and interest rates; they are conditional projections, not a prediction of bankruptcy on a fixed date.

    United Kingdom · July 2026 long-term scenarios · risk

  • Interest costs can jump without any policy change: on OBR sensitivities, 1 point higher RPI inflation adds about £7.0bn in a year and 1 point higher Bank Rate about £7.4bn.

    risk

  • On the OBR’s March 2026 forecast, debt interest rises to £137.1bn by 2030–31 — a projection that weaker growth or higher rates would make worse.

    March 2026 forecast · risk

  • After Budget 2026 the OBR forecasts a £14.5bn margin against the current-budget rule in 2029–30, down from £21.7bn in November 2025 and below the £27.2bn average of earlier forecasts; it puts the chance of meeting the rule at 56%.

    United Kingdom · October 2026 forecast for 2029–30 · risk

  • The OBR’s October 2026 forecast puts debt interest at £123.9bn in 2026–27, up from £109.4bn in its March forecast, and £160.0bn by 2031–32. It is a forecast that moves with gilt yields, Bank Rate and inflation, not a measured outcome.

    United Kingdom · October 2026 forecast · risk

  • The OBR forecasts spending inside the welfare cap at £207.0bn in 2029–30, £2.8bn above the cap, against a £1.9bn margin in November 2025.

    United Kingdom · October 2026 forecast for 2029–30 · risk

See the whole-UK SWOT →

Domain action plan

Proposed

Possible steps for government and the bodies responsible. These describe the order of work, not promised dates: starting an action and seeing results take different amounts of time.

  1. Short term · prepare

    Separate existing commitments, possible savings and uncertain revenue.

  2. Medium term · deliver

    Assess delivery costs and trade-offs before relying on new receipts or savings.

  3. Long term · judge results

    Track public services, debt costs and who bears the changes.

Delivery steps and owners (4)
  1. Prerequisite: Publish service baselines and the cost of current standards · HM Treasury and departments · proposed
  2. Decision: Choose between specified tax and spending packages on one baseline · HM Treasury; Parliament (Finance Act and Estimates) · proposed
  3. Delivery: Legislate, collect and allocate; record each line once in the ledger · HMRC, departments; UK Facts for the site ledger · proposed
  4. Result: Outturn compared with costings and service measures · OBR, ONS, NAO · proposed

Sources, method and changes

Shareable visual summary

Visual summary

Debt interest takes 8.8p of every £1 received—compared with 9.2p 30 years ago.

In 2025–26, central-government debt interest was equivalent to 8.8% of public-sector current receipts, compared with 9.2% in 1995–96. Receipts include taxes and other public income. The interest measure is net of the Bank of England’s asset purchase scheme. Latest-year figures are provisional; this comparison uses the September 2026 data edition. The share is of public-sector current receipts, not of tax or of spending.

  • £108.9bndebt interest, 2025–26
  • 8.8%of public sector receipts, 2025–26
  • £134.3bnborrowed, 2025–26

Three flows for the same financial year. Net debt (a stock) is a different measure and is not added to them.

Text and data version

Debt interest takes 8.8p of every £1 received—compared with 9.2p 30 years ago. In 2025–26, central-government debt interest was equivalent to 8.8% of public-sector current receipts, compared with 9.2% in 1995–96. Receipts include taxes and other public income. The interest measure is net of the Bank of England’s asset purchase scheme. Latest-year figures are provisional; this comparison uses the September 2026 data edition. The share is of public-sector current receipts, not of tax or of spending.

Records behind this visual summary
MeasureValuePeriodGeographyTypeSource
Central government debt interest, net of APF£108.9bn2025-26United Kingdomoutturn estimateOffice for Budget Responsibility, Public Finances Databank
Debt interest as a share of public sector current receiptsCalculation: Central government debt interest net of APF (% of GDP) ÷ Public sector current receipts (% of GDP) × 1008.8%2025-26United Kingdomderived calculationOffice for Budget Responsibility, Public Finances Databank
Public sector net borrowing£134.3bn2025-26United Kingdomoutturn estimateOffice for Budget Responsibility, Public Finances Databank
Public sector net debt excluding public sector banks, % of GDP93.3%2025-26United Kingdomoutturn estimateOffice for Budget Responsibility, Public Finances Databank

Available images and their source records are listed on Shareable visuals.

Sources and methods (6)

Method notes

  • Financial years run April to March. The latest outturn year is provisional and ONS revises it for several months; this page moves to each new Databank edition automatically.
  • Debt interest is central government interest net of the Asset Purchase Facility, the OBR’s preferred measure. It excludes interest paid by local government and public corporations.
  • Public sector net debt excludes public sector banks (the headline ONS measure). The version excluding the Bank of England removes the effect of the Bank’s balance sheet, including its gilt holdings and lending schemes.
  • All £ figures are cash. No chart here compares a cash series with an inflation-adjusted one; long-run comparisons use shares of receipts or GDP instead.

Corrections and changes

  • 2026-10-11: Corrected visual evidence labels: these are provisional fiscal outturn estimates, not survey estimates. Clarified that the debt-interest ratio uses public-sector current receipts.