The Republic of Agora

Rearmament Plans


Piling Up Trouble?

Paul O’Neill | 2023.09.06

With attention turning to rearmament as the global security situation deteriorates, NATO members will need to decide on the best model for replenishing their weapons and ammunition stocks should a major war occur. Stockpiles may not offer the best solution.

The war in Ukraine has exposed the poor state of Western stockpiles and the difficulties of sustaining high-intensity conflict, even if fought by others. The support provided to Ukraine, which has been essential, has been achieved by drawing on already small numbers of weapons systems and low ammunition stockpiles. Attention across NATO has turned to rearmament – not only to replace stocks given to Ukraine, but to grow stockpiles to levels more suited to the threats that NATO and national security strategies identify.

The UK’s refreshed Integrated Review and Defence Command Paper in 2023 stress the need to grow stockpiles and expand industrial capacity. In the 2023 Budget, the Ministry of Defence received an extra £1.9 billion to replace items given to Ukraine and to invest in munitions infrastructure. In addition, production capacity has to grow. Both are necessary, but neither provides the answer. A third avenue could be to adopt an idea of prototyping reversionary capabilities that are simple enough to produce at scale in the event of war using a non-specialist industrial base – a policy of dissimilar rearmament.

Stockpiles Will Never be Enough

Democracies rarely (if ever) go into a major war with the armed forces they need to win, especially if that war is protracted and attritional. While Adam Smith observes that the first duty of government is the protection of its people, he notes that investment in the armed forces cannot be larger than a nation can afford to maintain. Defence ministries will always compete with other priorities, such as health, education and social security. And money tied up in stock is often seen as wasteful against commercial accounting procedures. Consequently, budgeting processes can incentivise the reduction of stockpiles to lower book costs. Unless the incentives and accounting rules change, stockpile growth now is likely to result in reductions later on – the disposal by 2020 of PPE stocks bought after the 2010 National Security Strategy which identified pandemics as a Tier 1 risk is indicative of the challenge of maintaining stocks for contingencies. Covid also highlights how stocks have a shelf-life, so storing them could result in them expiring before they are needed. Moreover, uncertainty over the character of a future war can make investment in stocks of a singular capability meaningless – Mastiff, which was very effective against IEDs in Iraq, is less so in Ukraine against artillery or anti-tank weapons.

A solution could be to create more international stockpiles where countries pool their stocks and draw on the larger pool as required. NATO would be an obvious organisation for coordinating this, but it would require member states to go beyond interoperability and to do more to standardise their weapons, munitions and policies for storage, carriage and certification. Reinvigorating the NATO Standards Organisation is a start, but this will raise difficult questions regarding industrial sovereignty and it will take time to transition to the common standards, so it is not a quick fix. The Vilnius Summit offers some hope through the commitment to materiel standardisation and encouraging multinational cooperation, but whether this will be delivered remains to be seen.

Industrial Capacity

If large stockpiles are not the sole answer, expanding production capacity to ensure replacement weapons and munitions can be produced quickly is logical. But industrial capacity that lies fallow most of the time is expensive, and industry would have to be compensated for its preservation. This would be a large bill for governments, and whether paid by defence or industry ministries, the taxpayer would ultimately fund it; Adam Smith’s affordability warning remains relevant.

Even allowing for faster production in wartime, it is implausible that combat losses of major equipment will be replaced at anything like a “speed of relevance”

While growth in defence industrial capacity is sensible, it is always likely to lag behind what is actually needed in a major war, and to be potentially focused on sophisticated capabilities for which workforce skills and supply chains may atrophy when production lines close – the US order for Stinger missiles to support Ukraine in May 2022 will probably see the first missiles roll off the production line in 2026, despite having brought back retired workers and needing to reengineer obsolete electronic components.

An understanding of the supply chain and industrial ecosystem, including the skills base, is crucial; it is not enough to maintain capacity at the level of the defence prime if the suppliers do not exist or have capacity issues that will disrupt production flows. For example, there are very few steel mills in Europe capable of producing large quantities of armoured steel, and these rely on anthracite coal, the supply of which traditionally came from Donbas but has been disrupted by the war. Expanding production lines for tanks, therefore, is pointless if the armoured steel is not available. The same is true of stocks of long-lead items or the clamours for more additive manufacturing presses unless stocks of powders that the machines use to print are available.

Costs and capacity issues can be mitigated through exports, but this requires UK manufacturers to produce things the rest of the world wants to buy – probably not exquisite capabilities, but affordable weapons designed for armed forces from Day 2 of a conflict onwards. It also requires government to take industrial engagement with partners more seriously and to take a longer-term view – closer to that of the French approach that uses persistent political engagement, rather than the UK’s tendency for high-level ministerial engagement only when large contracts are ready to sign.

The Limits of Symmetrical Rearmament

Industrial capacity will never be enough, but even where capacity exists, it will take time to replace battlefield losses. Like-for-like replacement of sophisticated modern weapons systems will not be quick. Accepting that the timeline today is a peacetime one, the order for the second batch of Type 26 frigates was placed in late 2022, with construction of the last ship to be completed by the mid-2030s. Even simpler systems, like tanks, are too slow to deliver, with the upgrade of 148 Challenger 2 tanks to Challenger 3 taking over 6 years: Russia has allegedly lost over 4,000 tanks in the 18 months since invading Ukraine. So, even allowing for faster production in wartime – something defence, industry and the workforce are not practising – it is implausible that combat losses of major equipment will be replaced at anything like a “speed of relevance”. Given this reality, symmetrical rearmament is a chimera. And it is unlikely that networks of shadow factories can deliver modern platforms – this may have worked for Spitfires, but F35s will not be coming out of garages and coachbuilders because the production facilities, tools and skills are vastly different today.

Dissimilar Rearmament

Dissimilar rearmament envisages simpler weapons systems, perhaps made smarter with AI or more robust electronic components, that are mass-produced quickly at simple manufacturing sites and by lower-skilled workforces, or even additive manufacture. Individually less sophisticated than the weapons they replace, collectively they would provide the ability to sustain a fight until such time as better solutions can be found. The Ukraine war provides plenty of examples where dissimilar rearmament has given an advantage to Ukraine, and indeed Russia, that they would otherwise have lacked, such as uninhabited surface vessels, old tanks, cardboard drones and quadcopters able to carry RPGs and other explosives.

While not a panacea, the idea of dissimilar rearmament is worthy of further exploration to test its viability, alongside replenishing depleted stockpiles and strengthening industrial capacity

There might be three levels of dissimilar capability. Level 1 would use pre-existing platforms for military roles, much as China is doing where new merchant vessels are built with the capacity for military use, or sensor/weapons pylons that can be fitted to commercial aircraft to provide standoff reconnaissance or strike capabilities in extremis. Level 3 would consist of commercial off-the-shelf capabilities with all of their frailties, but which could be manufactured rapidly. Level 2 would sit between these and would be based on either standard commercial products with some adaptation for defence purposes, or military-grade designs where the cost and complexity required remains favourable. Level 2 would therefore be a broad church, and there would be no single approach within – let alone across – domains.

Dissimilar rearmament could offer a systematic model that connects science and technology, innovation and defence-industrial partnerships though continuous prototyping of reversionary weapons. Companies would have responsibility for specific capabilities, for which they would develop and test prototypes that could be rapidly manufactured, including under licence – a sort of Defence analogy of the ventilator challenge during Covid, but taking place in advance, because the four ventilator types chosen for mass production were based on pre-existing designs. Defence would purchase the intellectual property and experiment with the capability to ensure appropriate tactics, techniques and procedures, but the reversionary capability would not go into full-scale production except in wartime. The design authority would iterate the design so that the blueprints always represented viable capabilities. They might even be paired with pre-identified licensees who could be remunerated for preserving capacity and skills to be activated in a time of crisis.

Experimentation units within the armed forces would work with the design authorities to ensure not only that the technology was ready, but that the military had a pre-prepared ability to rush such items into service, including training users and maintainers on the equipment. These experimentation units might be operated by the Reserves as a way of ensuring that first-echelon forces are able to focus on the primary equipment in the inventory, and to bring an openness to non-traditional uses of equipment – a form of organisational ambidexterity where the regular units represent conventional (or traditional) capabilities, and the Reserves the unconventional ones. It would also ensure that the equipment could be used by those with less formal military training, which is likely to be necessary in the event of a major war where the first echelon might be expected to suffer significant levels of casualties. The second echelon, therefore, might be designed, trained and equipped differently to the first echelon: the first echelon focusing on competitive advantage through premier capabilities, and the second echelon configured for fighting with what would realistically be available to it in wartime.

There are, of course, problems with the notion of dissimilar rearmament. The nuclear deterrent seeks to avoid the need for the UK to field massed citizen armies; should its activation be necessary, deterrence will have failed, but it is uncertain whether the prime minister would resort to the use of nuclear weapons. Short of nuclear war, industrial incentives must still exist, and will therefore compete with funds for more capable weapons. Dissimilar rearmament also requires greater partnership between defence ministries and industry, and among companies who in other respects may see themselves as competitors. Finally, it does not solve the problem that defence will be paying for something that it may never need to use in anger, even if it is cheaper than the more traditional alternative of large stockpiles or preserving industrial capacity that lies dormant for much of the time. However, there is precedent: the Royal Aircraft Factories between 1911 and 1918 produced numerous designs, many of which were intended as research aircraft, to keep pace with the rapid developments in the emerging technology of powered flight.

While not a panacea, therefore, the idea of dissimilar rearmament is worthy of further exploration to test its viability, alongside replenishing depleted stockpiles and strengthening industrial capacity. There is a balance to be struck, but current plans appear to overly prioritise approaches that are unlikely to give the UK what it needs for a prolonged conflict.


Paul O’Neill is Director of Military Sciences at the Royal United Services Institute. His research interests cover national security strategy, NATO, and organisational aspects of Defence and security, including organisational design, human resources, professional military education and decision-making.

Made with by Agora