Evolutionary game theory (EGT) was developed in biology by John Maynard Smith and George Price in the 1970s and adapted to social science from the 1980s onward, notably by Robert Axelrod, whose 1984 book The Evolution of Cooperation used iterated Prisoner's Dilemma tournaments to show how reciprocal strategies like Tit-for-Tat can outperform unconditional defection over time.
In international relations, EGT departs from the standard rational-actor assumption that states calculate optimal strategies from fixed preferences. Instead, it treats strategies (cooperate, defect, hedge, balance, bandwagon) as traits whose frequency in a population shifts according to replicator dynamics: strategies that yield higher payoffs against the prevailing mix become more common, while underperforming ones decline. The central solution concept is the evolutionarily stable strategy (ESS), a strategy that, once dominant, resists invasion by rare mutant strategies.
EGT has been applied to:
- Cooperation under anarchy: explaining how reciprocity-based norms can emerge among states without a central enforcer, building on Axelrod and Keohane's work.
- Norm diffusion: modeling how practices like sovereignty, the laws of war, or human rights commitments spread through imitation of successful peers, complementing constructivist accounts.
- Arms races and signaling: treating hawkish or dovish postures as strategies whose fitness depends on the surrounding distribution.
- Alliance formation: examining when bandwagoning or balancing becomes self-reinforcing within a regional system.
The approach is associated with bounded-rationality and complexity-theoretic strands of IR. Critics note that states are too few, too heterogeneous, and too reflexive to behave like a biological population, and that learning in world politics is often strategic rather than blind imitation. Proponents counter that EGT is a useful heuristic for path dependence, tipping points, and the stickiness of international practices that pure rational-choice models struggle to capture. It is frequently paired with agent-based modeling in contemporary research.
Example
Axelrod's 1984 computer tournaments, in which Anatol Rapoport's Tit-for-Tat strategy won repeated iterated Prisoner's Dilemma rounds, became a foundational illustration used by IR scholars to argue that cooperation can evolve among self-interested states.
Frequently asked questions
Classical game theory assumes fully rational actors choosing best responses given fixed preferences; EGT assumes boundedly rational actors whose strategies spread through a population via imitation, learning, or selection, with equilibria emerging dynamically rather than being calculated.
Keep learning