The Bargaining Model of War

In brief

In 1995 James Fearon published "Rationalist Explanations for War" and reframed the oldest question in international politics. Instead of asking why states fight, he asked why they ever fail to avoid fighting. His starting point is simple arithmetic. War is costly: it burns lives, treasure, and time, and its outcome is uncertain. So if two states could see the eventual result in advance, they could agree today on a division of whatever they are fighting over that matches the expected outcome, minus the costs of the fight, and both would prefer it to war. Such a peaceful settlement almost always exists on paper. War is therefore a bargaining failure. The interesting theoretical work is not explaining aggression but explaining why rational leaders cannot reach the deal that would leave both of them better off. Fearon isolates three mechanisms that can block it: private information combined with incentives to misrepresent it, commitment problems, and issue indivisibility.

The full treatment

The puzzle it answers

Traditional accounts of war reach for human nature, ideology, misperception, or the anarchy of the state system. Fearon does not deny these matter, but he points out that most of them do not survive contact with rational choice. If war is worse than a negotiated settlement for both parties, then explanations resting on anger, greed, or the security dilemma leave a gap: they tell you why a state wants more, not why it cannot get most of what it wants without fighting. The bargaining logic makes this precise. Imagine two states disputing a piece of territory worth 100. Suppose war would give state A a 60 percent chance of taking all of it, but fighting costs each side 20. State A's expected value from war is 60 minus 20, or 40; state B's is 40 minus 20, or 20. Any division of the territory that gives A between 40 and 80 leaves both sides better off than fighting. That range of mutually preferable deals is the bargaining range, and it is created by the very costs of war. The puzzle is why states so often fail to locate a point inside it.

The first mechanism: private information and incentives to misrepresent

States have private information about their true military strength and their true resolve, meaning how much they are willing to bear to get what they want. If both sides knew these facts, they could compute the bargaining range and settle. But each has a reason to bluff. A state that overstates its strength or resolve can push the settlement in its favor, so simply announcing "we are strong and determined" carries no information: a weak state would say exactly the same. Credible signals are costly ones, actions a bluffer would not take. Mobilizing troops, signing a defense pact, or accepting a public commitment that would be humiliating to abandon can separate the resolute from the pretenders. But these signals shade into the danger they are meant to avoid. Sometimes the only way to prove you mean it is to fight, and war becomes the mechanism by which hidden information is finally revealed.

The second mechanism: commitment problems

Even with perfect information, a deal is worthless if neither side can be trusted to keep it. In domestic life a court enforces contracts. Between sovereign states there is no such enforcer, so every settlement must be self-enforcing. Two situations wreck this. The first is shifting power. If state B is rising and will be far stronger next decade, any promise it makes today about the future is empty, because when it is strong it will have no reason to keep an old bargain. State A may then prefer preventive war now, while it can still win, to a peaceful deal it knows will be renegotiated at gunpoint later. The second is first-strike advantage. If striking first confers a decisive edge, each side fears the other will attack, and the fear itself pushes both toward attacking. Robert Powell argued in 2006 that most rationalist causes of war ultimately reduce to this family of commitment problems over shifting power.

The third mechanism: issue indivisibility

The bargaining logic assumes the stake can be sliced. Some goods resist slicing. A throne, a holy site, national sovereignty over a capital, or the identity of who rules cannot easily be split into a 60 to 40 division. If the object is genuinely all or nothing, no settlement inside the bargaining range physically exists, and the parties may fight over the whole. Fearon treated this as the weakest of the three, because clever statecraft can usually manufacture divisibility through side payments, time-sharing, or linkage to other issues. Later work has taken indivisibility more seriously where domestic politics or religious meaning make a good genuinely impossible to compromise.

Lineage

The model grows out of the game theory that Thomas Schelling brought to strategy in The Strategy of Conflict (1960), where bargaining, credible commitment, and the manipulation of risk were first treated as formal problems. It shares its intellectual DNA with the study of mixed-motive games such as the prisoner's dilemma, where players who would both gain from cooperation fail to reach it. Fearon's contribution was to apply this apparatus to the specific question of war and to show that the standard non-rational explanations were incomplete. The approach was extended by R. Harrison Wagner in War and the State (2007), which recast war itself as an ongoing bargaining process, and by Robert Powell, who unified much of the field around the commitment logic. The tradition sits inside political realism's assumption of self-interested states in anarchy, but sharpens it into a testable account of when that anarchy actually produces violence.

The strongest case for it

Its power is parsimony with reach. From a single premise, that war is costly and its outcome uncertain, it derives a short and closed list of conditions under which rational states fight, and it explains why the vast majority of international disputes are in fact settled without war. It disciplines loose talk: "misperception" becomes the specific problem of private information, and "the security dilemma" becomes a commitment problem about first-strike advantage. It also generates predictions. Wars should be more likely between states with hard-to-observe capabilities, during power transitions, when offense dominates defense, and over indivisible stakes. And it reframes prevention as a design problem: reduce the costs of revealing information, build institutions that make commitments credible, and find ways to divide the indivisible.

The strongest case against it

The sharpest objection is that leaders are not the unitary rational calculators the model assumes. Domestic politics can make war profitable for a ruler even when it is ruinous for the state: a leader facing overthrow may gamble on conflict, a phenomenon studied under the label of diversionary war. Jack Snyder and others in the study of the domestic sources of aggression argue that logrolling coalitions and imperial myths drive states to overexpansion that no rational unitary actor would choose.

A second line, from the psychology of decision, holds that the model's rationality is a fiction. Robert Jervis, in Perception and Misperception in International Politics (1976), documented systematic, non-random errors in how leaders read one another, errors the bargaining model folds into "private information" but which are really failures of cognition, not strategic secrecy. Prospect theory adds that leaders facing losses take risks a rational expected-value calculation would forbid.

A third objection targets indivisibility and identity. Constructivists argue that whether a good is divisible is not fixed but socially constructed, and that treating states as fixed preference-maximizers misses how conflicts over meaning, status, and identity generate wars the model cannot price. Stacie Goddard's work on indivisible territory pushes this point.

Finally, critics note the empirical difficulty: private information and resolve are largely unobservable, which makes the theory hard to falsify cleanly, and the "war reveals information" claim risks becoming true by definition.

Where it stands now

The bargaining model is the dominant framework in the rationalist study of international conflict and the standard starting point in graduate training in the field. Its vocabulary, the bargaining range, private information, commitment problems, has become the common language even of scholars who reject its assumptions. Debate has moved inside the framework: how power shifts drive preventive war, how alliances and domestic institutions change the credibility of commitments, and whether nuclear weapons, by making the costs of war catastrophic and first strikes less decisive, widen the bargaining range enough to explain the long peace between great powers. What the model has done is turn the question of war into one that can be argued with a shared and precise structure.

Test yourself

Pick a war you know something about. Ask not why one side wanted to fight, but why the two sides could not find a deal both preferred to fighting. Was the block hidden strength and the temptation to bluff, a promise neither could be trusted to keep, or a prize that could not be divided? If you cannot fit it to any of the three, you have either found a gap in the theory or a fourth mechanism worth naming.