The assurance game, often called the stag hunt, is a foundational model in game theory and international relations theory. It originates from a parable by Jean-Jacques Rousseau in his 1755 Discourse on Inequality: two hunters can jointly catch a stag (high payoff) only if both stay at their posts, but either can defect to chase a hare (low but certain payoff). The game has two pure-strategy Nash equilibria — mutual cooperation (stag, stag) and mutual defection (hare, hare) — making it distinct from the prisoner's dilemma, where defection strictly dominates.
The defining feature is that cooperation is individually rational if the other player is expected to cooperate. The problem is therefore not one of conflicting interests but of trust, expectations, and credible signaling. Players are risk-averse: fear that the partner might defect can push both toward the inferior but safer equilibrium.
In IR theory, the assurance game is central to debates between realists and liberal institutionalists. Realists like Kenneth Waltz emphasize how anarchy generates stag-hunt dynamics where states cannot trust one another to cooperate. Liberals, drawing on Robert Jervis's 1978 article Cooperation Under the Security Dilemma (World Politics), argue that institutions, transparency, and repeated interaction can resolve assurance problems by improving information about others' intentions.
Common applications include:
- Collective security arrangements where each member contributes only if others will.
- Climate cooperation, where states fear being the lone decarbonizer.
- Arms control and disarmament, where verification regimes function as assurance mechanisms.
- Standard-setting and currency regimes, where coordination on a common focal point yields gains.
Unlike the prisoner's dilemma, the assurance game suggests cooperation does not require altering payoffs — only altering beliefs. This makes communication, reputation, and monitoring institutions especially powerful.