Every alliance summit produces a capability target, and every capability target is met more slowly than the communique implies. That is not a failure of will. Thirty-two nations each run their own procurement law, their own industrial policy and their own electoral calendar, and the target sits downstream of all three. The pledge is the easy part. Making thirty-two industrial bases behave like one is the decade's actual work.
The gap shows up in the least glamorous places. A munitions line that could run three shifts runs one, because the order book stops at the end of the fiscal year and no manufacturer will finance capacity against a contract that might not be renewed. A common platform arrives in eleven national configurations, each requiring its own spares chain. An interoperability standard is agreed and then implemented against different timelines, so the interoperability exists on paper for four years before it exists in the field.
Capacity is a financing problem before it is an industrial one
Defense manufacturers are not withholding capacity out of caution. They are responding rationally to demand signals that are annual, revocable and politically contingent. Capacity costs money years before it produces anything, and the firm that builds it carries the risk alone if the order does not come. Until the demand signal is longer than the investment horizon, the shortfall persists regardless of how much is eventually appropriated.
The nations that have moved fastest have done the unfashionable thing: multi-year commitments, pooled orders across borders, and a willingness to accept that some of the capacity built will be surplus in peacetime. That last point is the hard one politically, because surplus capacity is indistinguishable from waste until the day it is needed. It is the same argument as strategic reserves, and it loses in a budget committee for the same reasons.
The alliance can decide what it wants to field. It cannot decide, by communique, that a factory will exist.
Cohesion is tested by burden, not by agreement
Cohesion is usually measured by whether members agree. A better measure is whether they still agree once the cost is distributed unevenly, which it always is. A member on the eastern flank and a member on the Atlantic face the same treaty and a different threat clock. Both are acting reasonably. Arrangements that assume identical urgency will be strained precisely when they are needed.
The practical answer is not to insist on uniformity but to design for the asymmetry: regional groupings that can move at their own pace inside a common standard, with the alliance setting what must be true rather than how each nation gets there. This is how the industrial side will have to work too. A common requirement and thirty-two routes to meeting it is achievable. Thirty-two identical programs is not.
What this asks of governments now
Three things, none of them technological. Commit for longer than the political cycle, so industry can price the risk. Decide which capabilities must be sovereign and which can be depended on from an ally, and be honest that the second list is longer than anyone wants to say aloud. And put someone in charge of delivery who outranks the program, because modernization fails at the seams between departments far more often than it fails inside them.
None of that requires a new institution. It requires the existing ones to be held to a delivery standard rather than an announcement standard, which is a harder change, and the one that will decide whether the next decade's targets read, in hindsight, as a plan or as a wish.