What is the virtue of a proportionate response?

The question is a line of screen dialogue. In 1999 a fictional president asks it of his Chairman of the Joint Chiefs. The question is worth taking seriously because of the answer it gets. The Chairman does not defend proportionality. He concedes the point, “It isn’t virtuous Mr. President. It’s all there is”, and the argument then continues on other grounds.

Six defences are offered for proportionate response. Two of them defend something other than what is under attack. Three hold. The sixth fails on its own terms, and that failure is the strongest ground the argument against proportionality has.

Two things are granted throughout. Adversaries who structure themselves around a known rule are real and documented. A rule the adversary has already planned around is not working as a constraint on him. The question is what follows from that.

Core takeaways

• The legal rule compares expected civilian harm to the military advantage anticipated. It does not compare harm to harm.
• Matching a blow with a blow is a political and strategic practice, not a legal one. Much of the public argument attacks the wrong rule.
• The deterrence defence fails. A response that can be worked out in advance from the provocation can be evacuated and budgeted for.
• Three defences hold, and they hold on cost: allies, the options left for later, and what a state’s announced rule tells an adversary about its own captured personnel.
• Escalation research names human shielding as a predictable response to an adversary attempting escalation dominance. The conduct cited against restraint is partly produced by attempts to escape restraint.
• The formula “a head for an eye” entered the 2026 exchange as an Iranian threat months before it was adopted as American policy.

What the law actually requires

The most common defence fails on inspection, because the legal rule and the popular one are not the same rule.

Proportionality in the law of armed conflict forbids an attack whose expected incidental civilian harm would be excessive relative to the concrete and direct military advantage anticipated. It compares harm to military gain. It does not compare harm to harm. It has never required matching a blow with a blow, and nothing in it obliges a state to answer a transmitter with a transmitter. The tit-for-tat convention, which is what the question is asking about, is a political and strategic practice, not a legal one.

What this establishes and does not. An argument against eye-for-eye symmetry does not touch the legal rule it is usually taken to attack. The defence fails as stated, and the attack is aimed at the wrong rule. Both sides of the public argument are frequently discussing different objects. What survives is a narrower and harder question about the legal rule itself: whether “concrete and direct military advantage” can be assessed honestly against an adversary who has arranged the battlefield so that every advantage is expensive.

Military capability placed inside protected sites

The related claim is that proportionality is what stops a defender from being able to act at all when military capability is placed inside protected sites: hospitals, schools, mosques, tunnels beneath cemeteries. Documented cases of exactly this exist and are not in dispute here.

The rule does not do what this defence says it does. A protected site used for military purposes loses its protection. That is settled law and not a loophole. What remains is the proportionality calculation on incidental harm, which is a real constraint but a different one, and the party that placed the capability there bears responsibility for the consequences under the same body of law. Perfidy, which is often reached for at this point, is narrower than its colloquial use. It means killing, injuring or capturing an adversary by inviting his confidence that protection is owed and then betraying that confidence. It does not mean simply operating among civilians.

What this establishes and does not. The defence is confused about its own content, and so is the attack on it. The law does not forbid striking a mosque used as an armoury. Proportionality also does not leave a defender unable to act at all. The genuine problem is that the calculation gets harder, slower and more contestable exactly where the adversary has made it so. That is a claim about friction, not about prohibition.

Whether it deters

This is the defence that fails, and the failure is the strongest ground the disproportion argument has.

A rule the adversary knows is a rule he can plan around. In the drama, the four targets in the proportional package have been abandoned for four days, the satellites show it, and the strike will land on empty ground because the response was predictable enough to evacuate. “They did that, so we did this, it’s the cost of doing business, it’s been factored in.” The Chairman’s answer is “No sir, you’re right sir.” A response whose size and shape can be derived in advance from the provocation does not deter. It is a known cost, and an adversary budgets for a known cost.

The same structure appears in the field. Where a rule of engagement is known and stable, a sufficiently adaptive opponent will reorganise around it: uniforms abandoned, capability moved beneath protected sites, operations timed to the constraint. At that point the rule has stopped deterring anything. It has started deciding who holds the advantage.

What this establishes and does not. Predictability defeats deterrence, and the defence does not survive. It establishes that the practice of matching a blow with a blow buys less than it is assumed to buy. It does not establish that the alternative buys more. That is a separate question, and the remaining three defences answer it.

The cost in allies

Here the defences begin to hold, and they hold on cost rather than on virtue.

The disproportionate option in the drama is priced out loud: thousands of civilian dead, a region’s medical supply and drinking water crippled, and then the rest of the cost, no allies, no Western coalition, no Britain, no Japan, no Congress. “You’ll have doled out a five thousand dollar punishment for a fifty buck crime.” The president is not talked out of it on moral grounds. He is shown what it costs, asks for a cigarette, and takes the proportional plan.

What this establishes and does not. This is a real constraint, and pointing out that the enemy is worse does not answer it. It is contingent. It binds a state that needs allies and does not bind one that has decided it does not need them. So the defence binds hardest on a country whose strategic position is weakest.

The options left for later

In the Chairman’s phrasing, the proportional response does not use up the options available in the future. A maximum response spends the maximum threat. What remains afterwards is either repetition, which is no longer maximum, or nothing.

Escalation research supports the mechanism and adds a sharper finding. Attempts at escalation dominance produce reciprocal adaptation: opponents seek asymmetric strengths, prolong the conflict, and adopt precisely the behaviours that make the original problem worse. Human shielding is named in that literature as a predictable response to an adversary attempting dominance, which places it partly downstream of the strategy that cites it as justification.

What this establishes and does not. The defence holds, and it is the most damaging one, because it makes the case against proportionality partly circular. The conduct offered as proof that restraint fails is in part produced by the attempt to escape restraint. That does not make the conduct acceptable, and it does not mean restraint caused it outright. It means the causation runs in both directions, and an account that gives only one direction is incomplete.

What the rules tell an adversary about our own people

The last defence is the one nobody in the public argument answers, and it is not moral in form. Reciprocity is why a captured soldier is supposed to be treated as a prisoner rather than as an object. A state that has publicly announced it will answer an eye with a head has said something about what its own personnel can expect. The objection is usually made by former officers rather than by lawyers, which is worth noting. It is an argument about protecting one’s own forces, made in legal terms.

What this establishes and does not. It holds, and it is narrower than it sounds. Reciprocity binds where the adversary is capable of reciprocating in kind and cares to, which describes states more than it describes the irregular opponents the argument is usually about. The defence is weak against an adversary who was never going to observe the rule. It is strong against the next war, which may be against someone else entirely. The precedent set does not stay in the theatre it was set in.

Where this leaves the question

The concession is genuine and it should be granted: proportionality is not a virtue. It is not morally superior to have answered a barracks with two transmitters, and the deterrence claim made on its behalf does not survive an adversary who can read the rule. Anyone defending the practice on those grounds is defending it badly.

What survives is entirely consequential. The practice persists because the alternative costs allies, spends the escalation ladder, and sets a precedent that outlives the conflict it was set for. It also persists because the specific behaviour that most justifies abandoning it is partly a response to previous attempts at abandoning it.

Two further facts bear on the question. Ratio retaliation has been tried as formal policy, most explicitly in a 1941 occupation order setting fifty to a hundred lives against one, and the officer who signed it was prosecuted at Nuremberg for that order among others. The specific formula now circulating in English, a head for an eye, entered the 2026 exchange as an Iranian threat months before it was adopted as American policy, and its clearest theoretical defence was given by a hardline member of the Iranian parliament: “The defense doctrine of ‘an eye for an eye’ is wrong because it is not deterrent. The defense doctrine of ‘a head for an eye’ is correct because it is deterrent.” The argument is not a Western insight about an adversary who only understands force. Both sides reach for the doctrine, in the same words, in the same month.

How to practice this

1. When someone defends proportionality, ask which rule they mean. Notice whether they mean the legal calculation or blow-for-blow matching.
2. Take a rule your side follows and ask whether an opponent could derive its size and shape in advance. If he could, notice that it is a cost he budgets for rather than a deterrent.
3. Before choosing the larger response, list what it spends: allies, the options left for later, and the precedent. Notice which of those your position can afford to lose.
4. Check whether the conduct you cite as proof that restraint fails came after an attempt to escape restraint. Notice when the causation runs in both directions.
5. Ask what your announced rule tells an adversary about your own captured people. Notice that the precedent does not stay in the theatre it was set in.
6. Sort the defences you are given into claims about virtue and claims about cost. Notice that only the cost claims survive here.

Sources

• The West Wing, season 1 episode 3, “A Proportional Response”, written by Aaron Sorkin, aired NBC 6 October 1999. Transcript: westwingtranscripts.com; official podcast transcript, thewestwingweekly.com.
• Additional Protocol I to the Geneva Conventions, 8 June 1977, Articles 37, 44(3), 51(5)(b), 52. United Nations Treaty Series vol. 1125.
• Morgan et al., Dangerous Thresholds: Managing Escalation in the 21st Century, RAND MG-614-AF, 2008.
• Order for the suppression of communist insurrectionary movements in the occupied territories, 16 September 1941; Nuremberg document 389-PS.
• UN Fact-Finding Mission reporting on the Dahiya doctrine as declared by an Israeli Northern Command chief.
• Hamid Rasaei, member of the Iranian Majlis, 23 July 2026, Iran International.
• Full research bank, including the counter-case and the empirical record on reprisal and insurgency: 01 - Workbench/Proportional Response - Research Bank.md.