£20 Billion Written Off Over Two Years For No Reason

In the year leading up to March 2025, seventeen government departments recorded 2.7 million separate losses worth £6.589bn. The year before, the same departments lost £13.9 billion. Twenty and a half billion pounds in twenty-four months, conceded in writing, by the people who lost it.

Averaged across the calendar, a British government department wrote off public money once every twelve seconds. Overnight. At weekends. Throughout the summer recess, when nobody was even at their desk to authorise it.

A loss, in the Treasury's own definition, is expenditure from which a public body did not receive the benefit it expected. Not an overspend. Not a prudent provision against future risk. Not a cost overrun on something still being built. The money went out, and nothing came back.

Source: National Audit Office, Audit Insights, January 2026

Thirty-two billion pounds. From seventeen departments. Formally admitted. Audited. Published. And answered for by precisely nobody.

A penny on the basic rate of income tax raises roughly what Whitehall managed to lose last year alone. Every working person in this country handed over a portion of their wages to fund a programme of self-inflicted destruction, and the state's response has been to write it all down neatly and move on to the next financial year.

Fruitless Payments And Constructive Losses

Managing public money requires departments to disclose individual losses above £300,000, sorted into cash and stores losses, "fruitless payments," "constructive losses," and write-offs.

Savour "fruitless payment" for a moment.

The civil service has developed a dedicated technical term for money spent to no purpose whatsoever, complete with disclosure thresholds and a reporting format.

It has clearly encountered the phenomenon often enough to require standardised stationery.

The British state has professionalised failure to the point of giving it a filing system.

Britain Has Built A Thriving Industry In Not Building Things

This country has spent over £2.6bn arriving at the conclusion it would rather not build things. £472m of it went on eight roads which do not exist, including the tunnel under Stonehenge which has been promised, designed, costed, consulted upon, litigated over, and abandoned so many times it should be classified as a national ritual. The £1.7bn scheme was killed in the July 2024 spending audit alongside the A27 Arundel bypass, having previously survived a decade of design work, a planning inquiry, two ministerial approvals, and a trip to the Court of Appeal.

Rwanda is the same story. The Home Office booked £290m against cancellation of an agreement which the Home Secretary told the Commons had already consumed £700m, and whose full payment schedule committed Britain to £150,874 per relocated person over five years, plus £120m the moment the three hundredth arrival landed.

Asked by the Public Accounts Committee whether these losses were acceptable, the Treasury explained the defence write-offs were largely due to a change in government policy. Roads and Rwanda got the same explanation: shifts in priority.

The Treasury appears to consider this a defence. It is a confession.

Nobody sensible objects to a government abandoning a bad scheme. A state incapable of stopping is worse than one which stops too often.

The scandal lies in the sequence.

Our tedious political class:

1. Commits billions to projects whose political foundations have not been settled
2. Watches an election arrive with all the surprise of a scheduled train, and then:
3. Pays a second enormous bill to undo the first.

Changing your mind is free. Changing your mind after signing the contracts costs £2.6bn and no career.

Any private board which incinerated capital at this rate through indecision would be facing shareholder revolt, forced resignations, and the sort of press coverage which follows a man to his grave.

Whitehall got a hearing and a recommendation to think harder next time.

How To Lose £1.9bn Without Fighting Anybody

Defence deserves its own gallery. The Ministry of Defence reported £1.90bn of losses in 2024-25 against £663m the previous year, a near-trebling achieved in peacetime.

Ministers have defended the November 2024 decisions with great energy. Both landing platform docks, they insist, were mothballed by the previous lot, were never returning to sea before their out-of-service dates in the 2030s, and were costing £9m a year to keep notionally alive. Scrapping them saves £150m over two years and up to £500m over five.

Sadly....

Between 2022 and 2024, the taxpayer spent £72.1m refitting HMS Bulwark. She was then decommissioned and sold to Brazil, reportedly for a small fraction of the refit cost, though the sale figure has never been confirmed officially.

The Royal Navy has, in effect, run a lavishly funded restoration project on behalf of the Brazilian taxpayer, and thrown in a discount for the inconvenience.

Meanwhile the Army retired Watchkeeper, a programme which consumed the better part of two decades. It produced a write-off of £734m; the single largest item in the entire defence schedule. Ministers announced its cancellation with the tone of men clearing out a garage.

The schedule of individual defence losses above £300,000 runs to hundreds of entries, and mixes categories without embarrassment. It contains the loss of a Merlin helicopter in September 2024, in which a Royal Navy pilot died; a targeting pod which fell off a Typhoon over East Yorkshire; and the deliberate scrapping of a serviceable amphibious fleet. All priced in the same column at the same level of explanation.

A fatal accident and a ministerial decision are accorded identical dignity by the accounting rules, which suits the ministerial decision very well indeed.

Every one of these choices was made by an identifiable person, on a datable day, on documented advice. Not one of those people appears anywhere in the published accounts.

The vessels get named. The officials never do.

A Billion Pounds Of Medicine Left To Go Off

The second family of losses covers money nobody bothered to chase and things left to become worthless.

Source: National Audit Office, Audit Insights, January 2026

1.2 billion pounds of medicine went out of date in a warehouse.

Every packet of every drug ever manufactured carries an expiry date printed on the side. A department staffed with pharmacists, clinicians, procurement specialists, and qualified accountants bought a perishable product in industrial quantities and then failed to look at a calendar for long enough to lose more than the annual budget of a mid-sized hospital trust.

No war caused this. No market shock, no pandemic surprise, no unforeseeable act of God.

Somebody bought too much of something which goes off, and then everybody stopped thinking about it.

Nor is this a novel talent in Victoria Street.

The 2020-21 accounts recorded £8.7bn of the £12.1bn spent on pandemic PPE written off, including £673m of equipment found to be unusable and £750m of stock which expired before anyone opened the boxes. Five years on, the same building is still losing over a billion pounds to the passage of time.

The £572m of abandoned PPE claims is worse in its way, because it involves a decision to stop. Government believed suppliers owed it money. Government pursued them. Government then gave up, and the surrender is recorded without a syllable of explanation.

Half a billion pounds of potential recovery quietly filed under losses, and the accounts do not trouble the reader with who decided, or why, or whether anyone fought it.

Then fraud, where the disclosed figures acquire a comic modesty: £28m on Covid loan guarantee settlements, £25.5m from a single legal aid provider, and £11.6m across benefits and other sources. Sixty-five million pounds in total, which would make the British state the least defrauded institution of its size anywhere on earth.

It is not.

The NAO has separately estimated total fraud against the taxpayer at £21bn across the two pandemic years, against £5.5bn in the two years before.

The losses note does not measure fraud.

It measures the fraction of fraud which has been spotted, admitted, quantified, and formally given up on, which is a considerably more flattering number and roughly as useful as counting burglaries by asking burglars.

Six Billion Pounds Of Assets Nobody Can Verify

Underneath the losses lies something worse than waste. The Comptroller and Auditor General could not obtain sufficient evidence for £6.13bn of defence assets under construction, because the Ministry of Defence does not keep adequate accounting records.

Sitting nearby is £1.5bn of legacy project cost at the Atomic Weapons Establishment, carried at an unchanged figure for years, which the department could not justify continuing to recognise.

Add a £2.56bn shortfall in provisions for personal injury claims and Afghan relocation commitments.

The department responsible for Britain's nuclear deterrent cannot produce paperwork for six billion pounds of things it says it is building.

Across government, fourteen sets of 2024-25 accounts had been qualified by the end of December 2025.

• Six because auditors could not obtain enough evidence to form any opinion at all.
• Six because of material fraud or error.
• The rest because departments spent money outside the purposes or limits Parliament approved, which in earlier centuries would have been treated as a constitutional outrage rather than a footnote.

And during the audits, departments made £21.2bn of corrections to the stated worth of assets and liabilities, plus £11.3bn of corrections to income and expenditure.

Thirty-two billion pounds of arithmetic repaired by outsiders, because the departments could not get their own books right unsupervised.

A Hundred Billion Pounds Of Apology

Six major compensation schemes have paid over £29bn since 2005 and carry £73.4bn in future provisions. Estimated lifetime cost: £102.8bn. Clinical negligence alone accounts for £82.9bn, with the provision quadrupling since 2006-07.

Britain now spends more compensating citizens for harm inflicted by British public bodies than it does on several departments of state combined, and cannot calculate the bill.

The Infected Blood Compensation Authority could not produce credible evidence for £6.987bn of its £9.582bn provision.

The Department for Business & Trade failed for the second consecutive year to produce a reliable estimate for the Horizon Shortfall Scheme, and conceded its £743m figure might have been materially different had it analysed the replies to its own mailout to former postmasters. Its explanation was the analysis might have distracted from paying claimants.

An organisation which cannot simultaneously pay compensation and add up how much compensation it is paying has no business running anything larger than a village fête.

Everything's Perfect, Nothing Happened

The machinery of accountability functioned exactly as designed, which is the most damning fact available. Permanent secretaries are Accounting Officers. Managing Public Money sets the rules. The National Audit Office audits around four hundred bodies covering £1.1 trillion of spending. The accounts reach Parliament. The Public Accounts Committee summons officials and questions them.

All of it worked.

£6.589bn was identified, categorised, disclosed, audited, and laid before the House of Commons in good order.

The Committee then delivered judgement, describing losses from cancelled projects as a particularly egregious example of poor value for taxpayers, and recommended the Treasury analyse the root causes; identify what lessons can be learned; and report back by the end of the year.

Analyse. Identify. Report back.

There is no recovery mechanism. No named individual. No obligation on the Accounting Officer who bought the antivirals, or the senior responsible owner of Watchkeeper, or whoever authorised a £72m refit on a warship already destined for disposal, to sit in a chair and explain himself.

The Committee, by long convention, does not judge policy merits either.

The one body which examines the money declines to examine the decision. The ministers who take the decision are never examined on the money.

Thirty-two billion pounds has fallen through the gap between those two positions, and the gap is not an accident. It is the point.

Sanctions Which Would Actually Hurt

The remedy needs no legislation, only nerve. Attach names. Every loss above £50m should carry the Accounting Officer and senior responsible owner in post when the commitment was made, together with a published statement within six months setting out what was decided, when, on whose advice, and over whose objections.

Publish the trail. Cancellation should automatically release the business case, the Treasury approval, the gateway reviews, and the risk assessments produced before the money went out. Let the country read what the officials knew.

Force ministers to the despatch box. Any single loss above £100m should require a Commons statement in government time, with questions taken.

Report quarterly. Departmental loss totals published as they accumulate, rather than surfacing eighteen months later in a document nobody reads.

And give the Public Accounts Committee explicit authority to declare an avoidable loss avoidable, and to name the people who made it happen.

None of this recovers a penny. All of it would change the arithmetic in the mind of the official contemplating a purchase which cannot survive an election, or a stockpile which cannot survive its own expiry date.

Britain produces the finest documentation of institutional failure in Europe.

Every catastrophe counted, cross-referenced, indexed, audited, and laid before Parliament. The National Audit Office report chronicling the whole £6.589bn is printed on responsibly sourced paper bearing Forest Stewardship Council certification and retails at £10.00.

It is an exquisite piece of work. Nothing else about the affair is exquisite, and the paper is the only responsible thing in it.