Not much: European collaborative purchases of military equipment were 11 % in 2025, according to EDA – and only 19 of 27 states reported.
EU countries spent €115 billion on equipment last year, 26 % more than the year before. This is the story of one number, one footnote and eighteen years of promises.
One number and its footnote
Every year the European Defence Agency (EDA) tracks what share of military-equipment procurement countries conduct in cooperation with other European states. For 2025 it reports 11 %. The spread between countries is enormous: from zero to 55 %, with two states buying more than half of their equipment jointly. Once purchases with partners outside Europe are included – joint orders of equipment from non-EU producers, for instance – the share rises to 24 %. On these figures, European states buy together with partners outside Europe about as often as with each other.
The footnote matters as much as the number. Nineteen of 27 states sent the figure, eight did not report at all, and EDA itself writes of potential underreporting and of divergent national definitions of what counts as collaborative. The agency therefore does not publish the base from which the 11 % is calculated, so the share cannot be turned into euros. The true share may be somewhat higher. Even the most generous reading of the data, though, comes nowhere near the target Europe set itself.
Eighteen years since the promise
In November 2007 the defence ministers who steer EDA approved four collective benchmarks. One of them: at least 35 % of equipment spending to be procured together with other European states. The benchmark applies to the group as a whole, is voluntary – any country may adopt it as a national target, none has to – and carries no deadline.
The series EDA now publishes runs like this: 22 % in 2005, 27 % in 2011, 33 % in 2017, 22 % in 2021, 14 % in 2023 and 11 % in 2025. The number of reporting countries changes from year to year, so these are snapshots rather than a precise curve. Two things follow from them with certainty. The 35 % line was never reached, not even in the best year. And since 2017 the share has fallen while budgets have risen steeply: between 2024 and 2025 alone, equipment spending grew by 26 %. More money has not meant more cooperation; in the data it looks closer to the opposite, because after 2022 states plugged their most urgent gaps with whatever could be had quickly.
The full series and its budget context are in European defence in numbers.
What fragmentation means in practice
EDA puts it soberly: national approaches dominate, procurement programmes and equipment life cycles remain largely unsynchronised, and that undermines economies of scale, interoperability and long-term convergence. Translated into concrete situations.
Parallel programmes. When three countries each develop their own combat vehicle, they pay three times for development, testing and certification – and end up with three vehicles that cannot lend each other a spare part. A McKinsey analysis published in the Munich Security Report 2017, and reused by the European Commission the same year, counted 178 types of weapon system in Europe against 30 in the United States: 17 types of main battle tank against one, 29 types of frigate and destroyer against four, 20 types of fighter aircraft against six. The count dates from 2016, and the tank figure is disputed: a German defence journal counting strictly by type found 11 in the EU, and the Draghi report of 2024 cites twelve.
Short production runs. Unit price falls with the number of units built; that is as true for tanks as for phones. Whoever orders forty on their own pays the price of a short run. Whoever joins an order for several hundred pays the price of a long one. That is why the Czech defence ministry bought its Leopard 2A8 tanks in 2025 by joining the German defence ministry’s framework agreement: 44 tanks for CZK 34.25 billion, deliveries from 2028 to 2031, at a unit price the ministry says is lower than it would have negotiated alone.
Interoperability. Two brigades standing next to each other on the eastern flank that fire different ammunition, talk on different radios and carry different spares are two armies, not one. Stocks cannot be shared, repairs cannot be pooled, training cannot be combined. What exactly that costs in euros, no official statistic says. What the statistics do show is the direction: every percentage point missing from the 35 % in EDA’s chart means equipment developed, built and maintained several times over, side by side.
What is changing
Since 2025 there is an instrument that does not merely recommend cooperation in procurement but requires it. The EU’s SAFE loans, €150 billion in all, are with few exceptions available only for purchases involving at least two states. EDA reports first signs of change: countries have begun opening existing contracts to others so as to qualify for the loans, through a platform the agency built for the purpose.
It is a start, not a turn. SAFE covers a fraction of European procurement and its payments end in 2030. The 35 % benchmark remains voluntary, and nobody answers to anyone for missing it.
Why it does not happen by itself
Eleven per cent after eighteen years of promises is the logical result of an arrangement in which 27 ministries, 27 parliaments and 27 industrial policies decide on purchases, each wanting work for its own factory and none bearing the consequences when the common target is missed. A voluntary target with no deadline and no owner is met exactly as often as this one has been: not once in eighteen years.
Europe now has the money for defence – €418 billion a year and rising. It lacks a buyer. The United States procures for one army from one budget, which is why thirty types of weapon are enough. As long as Europe buys 27 times over, it will get shorter runs, poorer compatibility and a weaker industry for more money. Joint procurement for a joint defence is something a continent with one budget does automatically – and a continent with 27 budgets does only when a loan forces it to.