Starmer Shows his Hand on Defence Spending

Standing up in the House of Commons on 25 February, Prime Minister Keir Starmer finally met his manifesto commitment to 'set out a pathway' to spending 2.5% of GDP on defence, stating that this would be achieved by 2027, sustained throughout the rest of this Parliament and followed by an ambition to reach 3% during the course of the next Parliament. The statement was welcomed by the House, but accompanied by questions as to whether this would be ‘enough’, something that has been a topic of debate for several months given the state of the Armed Forces and with the 

Strategic Defence Review (SDR) underway. The prime minister’s statement was clearly intended to pave the way for his visit to Washington, given the new US administration has made the case that for Europe to do more it should be spending 5%.

Calculating 2%

The obsession with defence spending expressed as a percentage of GDP is a long-standing spectator sport within NATO, and the current focus extends back to the 2014 NATO Summit in Wales, where the Allies agreed to keep spending at that level as a minimum, or to move towards it within a decade – all with an aim of meeting NATO capability goals. The problem is that as a headline, the target is a political statement of intent, and it is often forgotten that the capability goals and commitments towards NATO are the actual measure of success. The current focus on percentage figures as a test of political virility risks obscuring the fact that what NATO and Europe in particular require are credible armed forces capable of fighting and winning, such that they deter an adversary from even attempting to use force. Meeting 2.5% will not be ‘enough’ to help UK defence if it becomes the goal in its own right, nor will 3% or any other suspiciously round number apparently plucked from the air.

Image credit: House of Commons Library

Any percentage figure can be problematic because the NATO criteria make clear they are quite broad, and can include costs that might not be related to direct military output. Some portion of the budget for the UK’s intelligence agencies (the Single Intelligence Account) has often been included in defence spending: many military personnel are seconded to the agencies, especially for signals intelligence and cyber activity as part of the National Cyber Force. In his statement, the prime minister appeared to suggest that a greater proportion (of what was £4.2 billion in 2024/25) would be counted as defence spending in the future, bringing the current percentage to 2.6%. This was justified as a response to the ‘changing nature’ of the threat, but it will need to be subject to some scrutiny in the future if there is further pressure within the Ministry of Defence (MoD).

The UK’s relationship with the 2% figure has been fraught; in a worrying example of inconvenient timing, the cuts to defence spending after the 2010 Strategic Defence and Security Review risked taking the UK below the 2% threshold around the Wales Summit. As the Commons Defence Committee identified a few years later, this fate was avoided by including some items (allowed under the NATO criteria) in defence spending which had previously not been used. It is worth remembering that the criteria aren’t just calculated against the core defence budget, and that other spending such as the cost of operations drawn from the Treasury Special Reserve can also be included. Figures in the current government who point out that the UK last hit 2.5% around 2009 and 2010 would do well to remember that was because Special Reserve funding going towards the considerable cost of simultaneous operations in Iraq and Afghanistan was included. Today, the annual commitment of £3 billion for Ukraine, which covers purchasing equipment, training Ukrainians and the cost of some operations in Europe, also comes from the Special Reserve. This isn’t unreasonable: it is still defence spending, but it is another example of why the 2.5% figure can’t be used as a measure of UK core military capability or outputs. 

How Much is ‘Enough’?

Given this, there is sometimes a suggestion that comparisons should be made to Cold War spending levels as a benchmark. Defence spending was around 7% of GDP at the time of the Suez crisis and then dropped to around 4% by the 1970s, jumping back towards 5% in the 1980s and after the Falklands War. Against this, 2.5% seems paltry. But this set of figures only tells half the story, not least as GDP is a moving target. The most obvious response to spending ‘enough’ and comparisons with what was ‘sufficient’ during the Cold War would be to compare defence spending in real terms. This shows that the high point in the late Cold War was 1984/85, where the figure (in 2023/24 prices) would be £55.3 billion. In other words, given the current defence budget of £56.9 billion and with £59.8 billion planned for 2025/26 under the (now superseded) Autumn Budget, the UK was already spending at ‘Cold War highs’. For another example of how the statistics can mislead, the UK is currently at 2.3% of GDP on defence spending, whereas Germany has finally crept up to 2.1% in the past year. But this still involves a German defence budget that is actually several billion pounds larger than the UK’s.

To be sure, the NATO spending target still has some value, as it represents the relative prioritisation of a country’s willingness to spend upon defence. But it can’t be used in isolation as a target for spending ‘enough’. A better question to ask is what the UK wishes to achieve, and what the requirement therefore is (followed rapidly by whether it is affordable). Just standing still – or rather, sustaining the current size and capability of the Armed Forces – is likely to require a sizable proportion of what might amount to around £6 billion a year in extra spending above the existing budget. 

Pressures on the Defence Budget

First, nuclear issues will continue to consume a large and probably growing proportion of the defence budget. The Defence Nuclear Enterprise covers more than just the deterrent submarines and Trident missiles; it also covers the Atomic Weapons Establishment, costs for the procurement and maintenance of the (nuclear-powered) attack submarine force, and the disposal of the Navy’s decommissioned nuclear submarines stretching back decades. With the Astute-class attack submarines still being constructed (painfully slowly) alongside the construction of the new Dreadnought deterrent submarines, and studies on a new warhead, the Nuclear Enterprise has been allocated nearly £110 billion over the next 10 years in the Equipment Plan alone: 38% of the plan overall, or just under 20% of all existing defence spending. And this data is more than two years old, with growth in the nuclear programme the area that Permanent Secretary David Williams previously identified as most concerning. If the current contingency in the nuclear programme proves insufficient, it could eat into spending elsewhere on conventional capabilities. 

Second, the current defence budget is under considerable stress beyond that introduced by nuclear issues. The last National Audit Office report also identified gaps in the Equipment Plan of nearly £7 billion over financial years 24/25 and 25/26. Overall, the plan was regarded as unaffordable, with a variance estimated at £16.9 billion but which could have been as low as £7.6 billion and as high as £29.8 billion. If a firm grip isn’t exerted on the plan then a large section of the increase could be consumed by overspending or dealt with through substantial cuts (and thus a reduction in capabilities or activities) or significant delays to procurement programmes to move expenditure into later years. The scale of this existing pressure was seen with the early retirement of a range of capabilities in November 2024, ‘only’ netting an estimated saving of £500 million. 

Image credit: House of Commons Library

Outside the equipment plan, personnel costs are also expected to rise even as the forces shrink, given that the public sector pay announcement last year potentially added more than £1 billion in costs, and service personnel already account for around 20% of the defence budget. Civilian personnel cost around 3.5% of the defence budget, so are not fertile ground for huge savings by comparison.

Potential Priorities for Investment

As ever, potential priorities competing for any new spending significantly exceed available resources. Climbing to 2.5% over the next two years might mean that by 2027/28 the defence budget is over £70 billion. This is in excess of most expectations in advance of the announcement, but in the face of the above inflationary pressures, the MoD can’t be profligate. There is a case to be made that alongside investment in new technology, AI and experimentation, the foundations of defence must be fixed in ways which are fundamental, if unflashy – reversing the trend towards 'hollow forces' which have good but thin capabilities or which are not as impressive as they appear on paper.

One priority might be personnel. The Armed Forces continue to lose more people than they recruit each year. Partly, this is down to the deliberate shrinking of the forces to more ‘affordable’ levels, but official surveys of personnel also show worrying levels of dissatisfaction with pay, pensions, allowances, career management and accommodation. Investing in these areas as well as targeted recruitment and improved training to address the burden on overworked specialists might be necessary to avoid the vicious circle of logjams in training, excessive deployments and short-notice moves which disrupt both careers and families and encourage more resignations, thus increasing the pressure on the remaining personnel.

Other priorities would include basing and infrastructure, necessary to improve the UK’s resilience by reducing its reliance on a small number of locations, or to address vulnerabilities in specific facilities which provide maintenance and support. If the UK – like much of Europe – wanted to improve its deterrent credibility, then having an obvious and proven ability to fight for an extended period of time is also a requirement, necessitating expanded ammunition stockpiles of all kinds. At the moment too much rests on holdings that are either growing too slowly or have been reduced through donations to Ukraine, while replacements remain years away. As well as complex weapons, artillery remains an issue: at the start of the conflict in Ukraine, it was calculated that at a similar scale the British Army would expend its stocks in a matter of weeks. Not only has subsequent analysis shown a continued high rate of expenditure, but the Army has donated its self-propelled guns (AS90) to Ukraine, and has been left with a stop-gap of 14 Archer guns: more capable but very limited until new self-propelled artillery is purchased. 

Image credit: House of Commons Library

These are not the only areas where gaps need to be filled: logistics, medical and intelligence capabilities are all present but thin. They all stand as examples of the challenge of readiness and deployability facing the Armed Forces right now. The forthcoming SDR will rightly look at the role of new technology, including a wider variety of better drones than currently deployed by the Armed Forces, AI, sensors and electronic warfare capabilities. All will need some kind of investment to ensure that the UK remains competitive. But if the SDR is about the longer-term transformation of the UK’s military, the near-term threat will require improving the value and impact that can be achieved with the forces the UK has now, not the wish list it wants to obtain a few years hence. It could be argued that not a single new tank, combat jet or warship should be ordered until the UK has improved the readiness and effectiveness of the forces that it currently has. To that end, reform of existing procurement and investment in the defence industry, coordinated with Europe, will be necessary, which is why the defence secretary’s emphasis on reforming the MoD should be not regarded as just an exercise in bureaucracy. As if to highlight this, and the issue of confidence in the MoD’s ability to control its spending, it has now missed production of an Equipment Plan for two years in a row, drawing the ire of the Commons Defence and Public Accounts Committees. Moreover, the 2015 Strategic Defence and Security Review and the 2021 Integrated Review offer cautionary tales: both offered reasonably substantial increases in defence spending which either turned out to be based on over-optimistic spending or were eaten away by inflation and procurement problems.

Do as the US Says, Not as it Does?

None of this is to say that headline figures are without value: the prime minister will need to go to Washington with a convincing message on how the UK will play a part in Europe doing more to ensure its own security, which by necessity will require a response to the Trump challenge on GDP figures. Politically, by taking the money from international aid, talking about 3%, and creating some internal party difficulties in doing so, the prime minister has also changed the tone of the discussion around defence spending and where it sits as a priority. Yet this can only be the start of the debate and can’t obscure the need for targeted investment. If the prime minister needed any more evidence of the superficial nature of the percentage target, he need only look at the NATO spending figures since 2014. Only one NATO member has reduced the proportion of GDP it spends on defence in the decade since the 2014 Wales Summit. That country is the US. 

© RUSI, 2025

The views expressed in this Commentary are the author’s, and do not represent those of RUSI or any other institution.

For terms of use, see Website Ts&Cs of Use.

Have an idea for a Commentary you’d like to write for us? Send a short pitch to commentaries@rusi.org and we’ll get back to you if it fits into our research interests. Full guidelines for contributors can be found here.