.png)
The £4.7 billion the Budget has to find: Funding the Defence Investment Plan at Budget 2026
On 30 June the government published the Defence Investment Plan, committing £298 billion to defence over the four years to 2029/30. Buried in the accompanying funding document is an admission that has attracted almost no attention: £4.7 billion of it has not been funded, and the money will have to be found at the Budget on 28 October.
That sentence is the government's own. It is not an external estimate or a critic's inference. The plan sets out how £10.3 billion of its additional £15 billion will be paid for, and states plainly that the balance will be confirmed at Budget 2026.
Our first policy paper, published today, asks a single question: where does that money come from?
The shortfall is not where you would expect
Most unfunded commitments are pushed into the distance, where a future Spending Review might absorb them. This one does the opposite. The £4.7 billion falls as £1.8 billion in 2026/27, then £1.1 billion, £1.0 billion and £0.9 billion.
The largest single unfunded amount is in the financial year already more than half over. That changes what the Budget has to do. It is not setting a path for future years. It is closing a hole in the current one, with most of the year's spending decisions already taken.
It also narrows the options sharply. HMRC's own estimates of tax measures assume implementation in April, and changes to benefit entitlements take effect at uprating. Neither route can deliver meaningfully before next April. Whatever funds the £1.8 billion will almost certainly be another reallocation.
None of the identified funding is new money
The composition of the £10.3 billion the government has explained matters more than the total.
£6.8 billion comes from other departments: £4.0 billion from a one per cent cut to every department's capital budget, £2.0 billion from the energy department, £0.8 billion from transport. A further £1.1 billion comes from asset sales. The final £2.4 billion is the Treasury taking on costs the Ministry of Defence would otherwise have borne, including Ukraine security guarantees in the event of a ceasefire.
There is no tax measure in the package and no additional borrowing. Not one pound of the identified funding is new revenue. The government says as much itself, in a way that is easy to miss: the £15 billion includes £11.6 billion of additional cash. The remaining £3.4 billion is spending power released by removing costs from the defence budget rather than money added to it.
This is not an accusation of bad faith. Reallocating toward your stated top priority is what prioritisation means. But a transfer changes which department spends the money, not how much the state has to spend, and it does not recur unless it is repeated. The one per cent capital levy already recurs in each of the four years, so returning to it in October would mean raising the rate on departments already absorbing an annual charge.
Four commitments, four different kinds of promise
Part of the reason this has gone largely unexamined is that the public conversation treats four quite different statements as though they were one.
The cash settlement to 2029/30 is firm, though incompletely funded. The 2.7 per cent of GDP figure is not a target at all; it is what the cash works out to when divided by a growth forecast, and it moves when either moves. The 3 per cent commitment is an ambition for the next Parliament, conditional on economic circumstances, with funding deferred to the 2027 Spending Review. The 3.5 per cent NATO commitment for 2035 has no domestic funding path attached, and the plan leaves spending between 2029/30 and 2035 deliberately unallocated.
The firmest of the four is the one with the acknowledged hole in it. The softest are the ones most often quoted.
Six routes, assessed the same way
The paper assesses six funding routes against identical criteria: how much each yields, how fast it reaches the force, whether it produces new resource or moves existing money, and whether it requires legislation or has already been attempted and reversed.
The finding is starker than we expected. Three of the six produce no cash at all. Reclassifying spending into NATO's resilience category raises a reported percentage without altering the force. Private capital cannot contribute at this Budget because the strategy that would enable it has not been published. Offsets redirect where money is spent rather than creating any.
That leaves taxation, welfare reallocation and borrowing. On the government's own documents, the Budget has two real choices.
Five tests for 28 October
The paper closes with five questions, each answerable yes or no on Budget day from the published documents. They are set out now, before the event, so they cannot be adjusted to fit it.
The first is the one that matters most. A Chancellor could announce £4.7 billion for defence on 28 October and add nothing whatever to the plan published in June, because that is precisely the sum the plan already assumed would arrive. The test is whether new money is additional to the £297.7 billion, or whether an existing shortfall is being closed and presented as an increase.
We will publish a scorecard against all five in the days after the Budget.
A note on method
This paper was not commissioned, funded, or reviewed before publication by any government, party, company or other organisation. It rests on departmental publications, official statistics, Office for Budget Responsibility forecasts and parliamentary sources, and where a figure is our own calculation rather than a published aggregate, the footnote says so. Where we have relied on a think tank or other non-official source, we name it in the text so readers can weigh it accordingly.
The numbers can be traced. We would rather they were.
Read the full paper: The £4.7 Billion Question: Funding the Defence Investment Plan at Budget 2026.