Plate iii in print

The Commitment

Schelling; the strategy of conflict, and the paradox that binding one's own hands — burning the bridge, throwing away the choice — can be a source of power.

The surprising advantage in a contest often falls to the side that can least afford to yield — the army that has burned its bridge, the bargainer whose hands are visibly tied — for the capacity to bind oneself, to throw one's own choices away, can become the capacity to move another.

— the argument of this plate, after Thomas Schelling, 1960

The prisoner's dilemma showed self-interested players trapped in a bad equilibrium, unable to reach the cooperation that would serve them. The first route out is to change the game — to alter the payoffs, or the players' beliefs about one another, so that the equilibrium itself shifts. Thomas Schelling made this shaping of the game the center of strategic thought, and he uncovered a paradox at its heart: that in a contest of strategy, the power to get one's way often comes not from having more options but from giving them up — from the ability to commit, irrevocably, to a course of action. Where the prisoner reasons in isolation, Schelling's player reasons about how to shape the other's expectations; and the sharpest instrument for shaping them, he saw, is to bind one's own hands.

In most of life, having more options is an advantage and keeping them open is wise. Schelling saw that strategic interaction inverts this homely truth. Consider an army advancing across a river: if it burns the bridge behind it, it destroys its own ability to retreat — and precisely thereby grows stronger, for the enemy now knows it must fight to the end, and may decline the battle. Consider a bargainer who declares, publicly and irreversibly, that he cannot pay above a certain figure: his instructions forbid it, the offer is final, his hands are tied. By demolishing his own freedom to concede, he forces the other to take his terms or walk away. What constrains the adversary, in each case, is the player's own self-binding — the deliberate destruction of his freedom to do otherwise. The move that looks like weakness, the throwing-away of options, is converted into strength, because it changes not what the player can do but what the other can expect him to do.1

1. The paradox: in ordinary life more options are better, but strategy inverts this. An army that burns its bridge cannot retreat — and so the enemy, knowing it must fight to the end, may decline. A bargainer who irrevocably ties his hands forces the other to accept or walk. What constrains the adversary is one's own self-binding, the destruction of one's freedom to yield. The apparent weakness becomes strength, by changing what the other can expect.

The mechanism is indirect but precise. A player's payoff depends on what the other does; what the other does depends on what he expects the player to do; so if the player can change that expectation, he changes the other's action, and thereby his own outcome. Commitment is the manipulation of that expectation. A threat — do this, and I will answer in a way costly to us both — moves the other only if he believes it will be carried out; but executing a threat is often costly to the threatener too, so a bare announcement may be dismissed as bluff. By committing irrevocably — making the response automatic, public, or otherwise impossible to retract — the player makes the threat credible, and the other, now believing it, complies. The same logic governs promises, whose keeping might later run against the promiser's interest until he has bound himself to keep them. Strategy, in Schelling's hands, becomes the art of credible commitment: of arranging matters so that what one has said one will do, one visibly cannot avoid doing.2

2. How commitment works: my payoff depends on your action, your action on what you expect of me; change your expectation and you change your action. A threat moves you only if believed — but executing a threat often hurts the threatener too, so a bare announcement is dismissed as bluff. Committing irrevocably — automatic, public, irretractable — makes the threat credible. Promises work the same way. Strategy becomes the art of arranging things so that what one has said one will do, one cannot avoid doing.
⁂

The deep difficulty Schelling exposed is that the most useful threats and promises are exactly the ones hardest to believe. A threat that would hurt the threatener to execute — a parent's threat to abandon the outing, a superpower's threat of mutual annihilation — invites the reply, you would never actually do it, and so fails to deter. A promise that the promiser will later wish to break — to repay, to reward, to refrain from exploiting a weakness — invites the reply, you will renege when the moment comes, and so fails to induce cooperation. In both cases the obstacle is the player's own future rationality, which will counsel backing down. Commitment devices are the remedy: the doomsday machine that retaliates automatically and cannot be called off; the contract whose penalty a court will enforce; the public pledge that would cost more in reputation to break than to keep. By delegating the action to a mechanism beyond one's own later discretion, the player makes credible what reason alone would leave hollow. The credibility problem here is the very one The Level met in monetary policy, where a central bank must bind itself to be believed.3

3. The deep difficulty: the most useful threats and promises are the least believable. A threat costly to execute ("you would never do it") fails to deter; a promise the promiser will later wish to break ("you will renege") fails to induce cooperation. The obstacle is the player's own future rationality. Commitment devices remedy it — the automatic retaliation, the court-enforced contract, the reputation-staking pledge — delegating the act beyond one's later discretion. The same credibility problem bound the central bank of The Level.

Schelling's second great insight concerned a gentler kind of game, in which players want the same thing but cannot communicate. Asked to meet a stranger in a city on a given day, with no chance to confer, most people name the same place and hour — the obvious landmark, the round number on the clock — not because logic dictates it but because it is salient, prominent, the natural place for expectation to settle. Schelling called these focal points: solutions made compelling not by their payoffs, which may be matched by many others, but by some psychological, cultural, or historical prominence that lets each player guess that the others will guess it too. Focal points reveal that coordination is solved by shared expectation as much as by calculation — that a convention, a tradition, a round figure, or a natural boundary can succeed where pure reason, confronting many equally good options, would be paralyzed. The mutual reading of what stands out does the work the arithmetic cannot.4

4. Focal points: when players want the same thing but cannot communicate, they converge on the salient solution — the obvious landmark, the round number — because each expects the others to expect it. The prominence may be cultural, historical, psychological; the payoffs of the focal option need be no better than many rivals'. Coordination is thus solved by shared expectation, not calculation alone: a convention or natural boundary succeeds where pure reason, facing many equal options, is paralyzed.

With commitment and focal points, Schelling enlarged game theory beyond the pure conflict that von Neumann had analyzed. Most strategic situations, he argued, are mixed in motive — the players share some interests and oppose others, as in bargaining, deterrence, or a marriage — and in such games the decisive moves are not the choice of a strategy from a fixed menu but the shaping of the game itself: the commitments made, the threats and promises issued, the expectations manipulated, the focal points seized. This is the strategist's art, and it lit up everything from labor negotiations to the nuclear standoff that hung over his century. Yet Schelling's commitments raise a question he pressed but a younger theorist would formalize: which threats and promises are actually credible, and which are bluffs that a clear-eyed opponent should ignore? The answer required looking forward through the game and reasoning backward, and in finding it the theory sharpened its very notion of equilibrium. The next plate is the credibility.5

5. Schelling enlarged game theory past von Neumann's pure conflict to mixed-motive games — bargaining, deterrence — where players share some interests and oppose others, and the decisive moves shape the game: commitments, threats, promises, focal points. The strategist's art lit up labor talks and the nuclear standoff alike. But which threats are credible, and which mere bluff? The answer, found by reasoning backward through the game, sharpened the idea of equilibrium itself. Next: the credibility.

The first escape from the dilemma is to change the game, and Schelling found the master tool: credible commitment. In a contest of strategy, he showed, surrendering options can be a source of power, for by binding one's own hands — burning the bridge, tying the bargainer's hands, automating the retaliation — one changes what the other expects, and so changes what he does. Threats and promises work only when believed, and the most valuable are the least believable until commitment makes them irrevocable, just as a central bank must bind itself to be trusted. And where players seek to coordinate without speaking, focal points let shared expectation settle on a salient solution that calculation alone could never single out. Schelling made game theory the study of mixed-motive interaction and of the shaping of the game itself. But a commitment must be credible to do its work, and the next plate asks which threats a clear-eyed player should believe. The next plate is the credibility.