Starmer's £270 Billion Was Wrong. A Watchdog Wrote a Letter.
On 13 March, in front of the packed benches and the union jack flags that always seem to travel with the Prime Minister these days, Sir Keir Starmer told the House that Labour was delivering "the biggest boost to defence spending since the Cold War, £270 billion over the course of this Parliament." He said it twice, for weight. It is the kind of figure a government wants attached to it: enormous, martial, unarguable.
It was also, according to the UK Statistics Authority, wrong. The £270bn is not an increase. It is the Ministry of Defence's entire projected budget across four financial years, 2025–26 to 2028–29, restated as though it were new money on top of what already existed. Penny Young, the authority's interim chair, put it plainly in a letter to the shadow defence secretary, James Cartlidge: it is "incorrect and misleading to describe the £270bn as an increase in defence spending over the course of this Parliament, or to describe it as an extra £270bn." That is the Whitehall translation of a superlative: not a boost, a total.
What makes this more than a slip at the dispatch box is what happened next, or rather didn't. The figure did not stay in March. In June, an MoD spokesman repeated it. On 3 June, the Wales Office posted on X that "the UK government is increasing defence spending by £270bn over this Parliament to strengthen our national security" — the same claim, the same number, three months after the Prime Minister had first used it and after, one assumes, someone somewhere might have noticed. BBC Wales reports the figure originated in the MoD rather than the Wales Office itself, then run by Jo Stevens. The Wales Office deleted the post once challenged and told the authority the wording "was a mistake and was amended as soon as it was noticed." Three months is a long time not to notice your own government's headline defence number.
There is a case, and it should be made properly before it is dismissed. Multi-year budget profiles genuinely confuse the distinction between total spend and increase, and most people outside the Treasury glaze over the difference between a cash rise and a restated total the moment anyone puts a comma in a number that size. The Wales Office did delete the post. The government did answer the Statistics Authority rather than stonewall it, telling the watchdog that "we have previously acknowledged this phrasing and took immediate steps to ensure it was not replicated." Compared with how governments usually treat inconvenient corrections, this looks almost prompt.
Except the PM said it twice, deliberately, as a rhetorical flourish rather than a footnote — "the biggest boost since the Cold War" is a phrase designed to prevent nuance from happening. And the same wrong number then travelled, unprompted, into two further departments across three months: a line handed down from the MoD and repeated by anyone who needed a big figure for a press release. Downing Street, asked for comment, said it had nothing to add. No minister corrected the record in the Commons. The entire mechanism of accountability here consists of a deleted tweet and a spokesperson's careful passive voice.
It matters what the real number is, because there is one, and it is duller and more honest than £270bn. Defence spending rises from 2.3% of GDP now, around £66.3bn, to 2.5% by 2027–28, with a promised 3.5% by 2035 — a genuine, gradual, contestable increase that a government could defend on its merits, argue about in real terms, and be judged on in five years. Instead ministers reached for a bigger, vaguer figure that sounded like victory and required no defending at all, because by the time anyone checked it, the news cycle had moved to something else.
Attlee's government, bankrupt and rationing bread, still managed to make its numbers mean something. When a minister told the Commons what was being spent, it was because the Treasury had signed off actual cash that would be defended, item by item, in supply estimates that ran for days. Nobody needed a Statistics Authority to catch Attlee out, because the whole apparatus of the state was built to make sure a false claim could not survive contact with Parliament. Ours produced a body that can write a very stiff letter.
That is the accountability gap this episode maps with unusual precision. The UK Statistics Authority has no power to compel a correction, strike a claim from Hansard, or force a minister to stand up and withdraw a number. It can write to a shadow minister, note that the government "should reduce the risk" of repeating the error, and leave it there.
Nobody voted for the Statistics Authority, but then nobody voted against it either, because it was never given enough teeth to be worth objecting to. The number Starmer used in March is still in Hansard, unamended, quoted twice, true in the narrowest sense and false in every sense that mattered to the person listening. Ask who corrected it, and the honest answer is: a watchdog, eventually, to nobody who could do anything about it.