Game Theory · GT-10
Play the same game forever, and almost any outcome — even cooperation — can become an equilibrium, backed by the threat of future punishment.
The folk theorem shows that in infinitely (or indefinitely) repeated games, a very wide range of outcomes — including full cooperation — can be sustained as equilibria, as long as players value the future enough and can credibly threaten to punish defection in later rounds.
Called the 'folk theorem' because its core insight circulated informally among game theorists in the 1950s before being formally proven; landmark formalizations came from Friedman (1971) and later refinements through the 1980s.
The Mechanism
Cooperation sustained by the shadow of the future
The one-shot Prisoner's Dilemma trap disappears once the game repeats indefinitely — the threat of future punishment (a return to mutual defection) becomes a credible deterrent, provided neither side discounts the future too heavily.
01 · THE SHADOW OF THE FUTURE
Repetition changes the incentive structure entirely
In a single round, defection dominates. Once the game repeats an uncertain number of times, a player who defects today invites retaliation across every future round — as long as future payoffs matter enough (a high enough discount factor), the total cost of one round's defection gain is outweighed by the lasting punishment.
02 · 'ALMOST ANY' OUTCOME MEANS THE THEOREM IS A DOUBLE-EDGED SWORD
Cooperation is possible, not guaranteed
The folk theorem's actual claim is unsettling as much as it is hopeful: it shows a huge range of outcomes — including bad, collusive, or exploitative ones — can also be sustained as equilibria under repetition. Cartels sustaining high prices through repeated interaction and mutual threat are folk-theorem equilibria too, not just cooperative ones.
03 · IT REQUIRES A DEFINITE END TO BE ABSENT (OR UNKNOWN)
Finite, known-end games unravel via backward induction
If both players know the exact final round, backward induction unravels cooperation entirely — nobody has anything to gain by cooperating on the last round, which unravels the second-to-last round, and so on to the very first. The folk theorem's cooperative equilibria require the game to be infinite, or to have an uncertain end.
Where It Fails / Inversion
Where it fails / inversion
The theory assumes players can actually observe defection and coordinate a credible punishment — in settings with imperfect monitoring (you can't always tell if the other side defected, or it takes time to notice), cooperation is much harder to sustain, and real-world 'repeated games' often collapse well before theoretical folk-theorem predictions because of exactly this observability gap.
How To Use It
Worked example · supplier relationships that outlast any single contract
Long-term supplier relationships often sustain better pricing and quality than any single contract could enforce, precisely because both sides know the relationship will continue — a supplier who cuts corners on one order risks losing all future orders, and a buyer who squeezes too hard risks losing supply security. The implicit threat of ending a valuable repeated relationship substitutes for a formal enforcement mechanism.
How to use it
When designing an ongoing business relationship, deliberately make defection observable and punishment credible — clear performance visibility and an explicit (even if informal) understanding that bad behavior ends the relationship does more to sustain cooperation than any single contract clause, because it invokes the same mechanism that sustains cooperation in repeated games generally.
See Also