Cutting aid to spend more on defence: how do the numbers add up?

At Prime Minister’s Questions, Keir Starmer repeated his claim that the increase in defence spending he announced on Tuesday will amount to £13.4bn a year. He was asked by Kemi Badenoch to explain how he arrived at that figure when the independent Institute for Fiscal Studies (IFS) says the increase is £6bn a year in real terms – a figure John Healey, the defence secretary, has accepted.

As Badenoch pointed out, the IFS has said the government is playing “silly games” with numbers.

Starmer said: “If you take the numbers for this financial year and the numbers for the ’27-28 financial year, there is a £13.4bn increase.” He offered no further information, and neither did his spokesperson afterwards, so we – and the IFS – cannot be sure how he arrived at this figure.

First, by comparing a cash figure this financial year with one in three years’ time, the prime minister is not taking inflation into account. Second, Tuesday’s decision was expressed as a share of GDP, and even though the economy is not growing as fast as people would like, it is forecast to grow over three years, so by fixing this year’s cash figure as the starting point, he is pretending that defence spending would have fallen as a share of GDP by 2027 if the government had not decided to increase it. This is not what was set out in the government’s spending plans, which were that defence spending would remain constant as a share of GDP.

Finally, Starmer may have included spending on the intelligence services in the 2027 figure, but not in this year’s baseline. He said on Tuesday that spending on the intelligence and security services would be called “defence spending” in future. As Badenoch pointed out in her reply, this is not new money. Never mind “silly games”: if the prime minister has counted reclassifying existing spending as an increase in defence spending, he is simply wrong.

The arithmetic is simple. Defence spending will rise from 2.3 per cent of national income now to 2.5 per cent in 2027. It will be paid for by cutting foreign aid from 0.5 per cent of national income to 0.3 per cent. That is all there is to it: a transfer of 0.2 per cent of national income from aid to defence. That 0.2 per cent will be worth about £6bn a year in 2027.

There is a wrinkle in the definitions, in that the defence spending target is expressed as a share of GDP (gross domestic product), whereas the aid target is a percentage of GNI (gross national income). GNI “includes the income earned by a country, regardless of where it is earned”, according to the Office for National Statistics, but the two numbers can for most purposes be treated as the same.

The armed forces have long had their pet projects on which each service would like to spend more money, but most analysts seem to agree that most of the increase in spending will go on personnel – paying existing service people more and recruiting more.

The losers will be aid programmes in some of the poorest countries in the world. The effect will be greater than the 40 per cent cut implied by reducing the aid budget from 0.5 per cent of national income to 0.3 per cent, because 28 per cent of the budget is spent on housing asylum seekers in Britain. Starmer said on Tuesday that this money would be freed to be spent abroad as the asylum backlog is cleared, but this will not happen quickly. He also said that aid to Sudan, Gaza and Ukraine would be protected, meaning even deeper cuts to aid in the rest of the world.

Another part of Starmer’s announcement on Tuesday was the aim to continue to increase defence spending so that it reaches 3 per cent of GDP “in the next parliament”.

This is a big increase, although reclassifying intelligence as defence spending will make it easier, because it will take the 2027 figure up from 2.5 per cent of GDP to 2.6 per cent.

At the same time, Starmer says that he wants to restore aid spending to the 0.7 per cent target that was reached under David Cameron’s government. One of those targets is more likely to be met than the other.