Coalition friction
A Socratic walk-through of coalition friction — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why can two allied armies fighting together achieve less than either would achieve alone?
Two allies each contribute a competent force to a joint operation. On paper the coalition has more of everything: more troops, more aircraft, more transport, more money. The addition ought at worst to be neutral.
Yet joint operations routinely underperform what either partner would manage with the same resources alone, and the phenomenon is consistent enough that professional militaries plan for it explicitly. Something is being subtracted, and it is not equipment. Since the forces are the same forces, whatever is lost must live in the connections between them rather than in either.
Reasoning it through
REASONING #Start with the most concrete layer and work outward, because the concrete problems are the ones people expect and the abstract ones do the real damage.
Technical. Radios that cannot talk to each other. Ammunition that does not fit. Fuel couplings, data formats, map datums, incompatible identification systems. These are genuine and they are expensive, but they are also the easiest to fix and the most visible, which is why alliances invest heavily in standardisation and why this layer is usually not where an operation actually fails.
Procedural. Two forces can be perfectly interoperable in equipment and still disagree about how things are done: what a warning order contains, how much detail a plan specifies, how far a subordinate may deviate from intent, what a phase line means, when fire may be called. A force whose doctrine expects subordinates to improvise toward a stated intent, working beside one that expects detailed orders and adherence, will produce mutual astonishment — each reading the other as either reckless or paralysed. Nobody has done anything wrong, and the operation still stumbles.
Decisional. Here is the layer that separates a coalition from a merely large force. In a single national force, there is one chain of command and one authority resolving a dispute. In a coalition, each contingent retains a line home. Commanders have caveats — national restrictions on where their troops may go, what they may do, when they may fire — and those caveats often cannot be waived by the coalition commander at all, because they are the price of the contribution.
Follow the consequence. The coalition commander is not commanding a force; they are negotiating with several. A plan must be acceptable to every contributor, so it converges on what all of them will accept, which is systematically less ambitious than what any of them would attempt alone. And the delay is not a one-off: it recurs at every decision, because the referral home happens again each time.
Political. Underneath that sits the reason the caveats exist. Each government is accountable to its own public, and each faces the classic problem of a shared effort: the benefit of the operation is largely common, while casualties, cost and domestic political damage are national. That structure rewards contributing enough to be seen as a good ally and little enough to limit exposure — and it rewards it for every partner simultaneously.
Now combine the layers and the arithmetic of the puzzle resolves. Adding a contingent adds capability linearly, but it adds coordination relationships combinatorially — with n partners there are n(n−1)/2 pairs to reconcile, so five partners means ten relationships rather than four. Meanwhile the plan's ambition is set not by the sum of the partners but by the most constrained of them. Capability adds up; permission takes the minimum. It is entirely possible for the minimum to fall as partners are added.
That is the mechanism in one line: a coalition sums its resources and intersects its permissions.
The analogy
THE ANALOGY #Think of four excellent cooks sharing one kitchen to produce a single dinner, where each has agreed to come but each is answerable to a different household about what may be served.
The kitchen has four times the hands. It also has one oven schedule to agree, four ideas about how far ahead to prep, and four dietary restrictions that must all be respected — so the menu is not the union of what they can cook but the intersection of what everyone will allow. Every substitution needs a phone call home. The dinner arrives late, and it is plainer than any one of them would have made alone.
Cooks can taste as they go and correct in seconds, whereas a coalition's feedback about whether the coordination is working arrives slowly and ambiguously, often only when something has already gone wrong — which is why the friction persists rather than being ironed out during the operation.
Clarifying the model
THE MODEL #Coalitions are usually formed for reasons the friction does not outweigh, and the piece should not read as an argument against them. Legitimacy, basing and overflight rights, burden-sharing, intelligence access, and the political cover of acting with others are real and often decisive. A coalition that is militarily less efficient can still be the correct choice, and frequently is. The claim here is narrower: that the military output is not the sum of the parts, and that planners who assume it is will be wrong in a predictable direction.
Friction is not incompetence, and treating it as such is the standard error. Every layer above arises from partners each behaving sensibly given their own doctrine and their own accountability. This matters because it predicts which remedies work: shared exercises, liaison officers, standing structures and common procedures reduce friction, while irritation at allies does not.
Standing alliances differ sharply from ad hoc ones, which is the natural experiment. Where partners have trained together for years with common procedures and equipment standards, the technical and procedural layers largely dissolve — and the decisional and political ones do not. That the residual friction concentrates in exactly the layers that peacetime standardisation cannot touch is decent evidence that the four layers are separable, and that the deeper ones are doing more of the work.
The evidence base is unusually interested. Post-operation accounts are written by contributors with reputations at stake, and attributing a failure to allied caveats is a convenient explanation for a national force that underperformed, just as attributing it to a partner's incompetence is convenient in the other direction. I would treat specific attributions in memoirs and official histories with more scepticism than the structural pattern, which is visible across many operations with different participants.
The falsification test. If the mechanism is intersected permissions rather than summed capability, then adding a partner with restrictive caveats should reduce the ambition of the plan even when it adds real forces — and coalitions whose partners have converged caveats should outperform equally-equipped ones whose caveats diverge. If operational ambition tracked the total forces available regardless of the restrictions attached, the account would be wrong and the friction would be purely technical.
A picture of it
THE PICTURE #How to readRead widths as shares of an imagined pooled capability, drawn to express the argument rather than measured from any operation. The upper splits are permission being intersected: each partner contributes forces, and each withholds some by national restriction, with the proportions differing by partner. The lower split is the coordination overhead taken from what remains. The point is that two separate subtractions occur before anything reaches the output, and only the second of them is what people usually mean by friction.
What became clearer
WHAT CLEARED #A coalition is not a larger army. It is several armies with one plan, and while their resources add, their permissions intersect and their coordination burden grows faster than their numbers. The technical incompatibilities that dominate the popular picture are the layer most easily fixed; the decisive constraints are that every contingent keeps a line home, that the plan must satisfy the most restricted partner, and that each government bears national costs for a common benefit. Adding a partner reliably adds capability and can genuinely subtract ambition.
Where to go next
ONWARD #- Why caveats are the price of a contribution rather than a failure of will, and how they are negotiated.
- How standing alliances dissolve the technical and procedural layers but not the deeper two.
- Why liaison officers and habitual relationships buy more than equipment standardisation does.