Intermediatemicroeconomics
The Prisoner's Dilemma: Why Rational Choices Can Hurt Everyone
Overview
The prisoner's dilemma is the most famous model in game theory: two suspects, interrogated separately, each must decide whether to stay silent or confess. Confessing is each prisoner's dominant strategy — yet when both confess, both do worse than if both had stayed silent. This clash between individual and collective rationality explains price wars, arms races, cartel cheating, overfishing, and underfunded public goods. This lesson works through the classic payoff structure, the equilibrium, and repeated-game escapes from the dilemma.
Learning objectives
- Set up the prisoner's dilemma payoff structure and identify each player's dominant strategy
- Find the Nash equilibrium and show it is Pareto dominated by mutual cooperation
- Explain the tension between individual and collective rationality
- Apply the dilemma to cartels, arms races, advertising wars, and common resources
- Describe how repetition and punishment strategies can sustain cooperation
The classic story and payoff structure
Two suspects are arrested and interrogated in separate rooms. Each can stay silent (cooperate with the other prisoner) or confess (defect). If both stay silent, each serves 1 year on a minor charge. If one confesses while the other stays silent, the confessor goes free and the silent one serves 10 years. If both confess, each serves 5 years.
In payoff terms (using negative years as payoffs, or equivalently 'freedom points'), each player reasons as follows: if the other stays silent, confessing means freedom (0 years) versus 1 year — confess is better. If the other confesses, confessing means 5 years versus 10 — confess is better. Confessing is a dominant strategy for both. The predicted outcome is mutual confession: 5 years each, even though mutual silence would have given 1 year each.
The equilibrium is Pareto inferior
The mutual-confession outcome is a Nash equilibrium: given that the other confesses, neither prisoner can improve their own outcome by switching to silence. Yet it is Pareto dominated by mutual silence — both prisoners would be better off, and nobody worse off, if both stayed silent.
This is the dilemma's punchline. Individual rationality (each playing a dominant strategy) produces collective irrationality (an outcome everyone would pay to avoid). The structure requires the standard ranking of payoffs: temptation to defect beats mutual cooperation, which beats mutual defection, which beats being the sucker — T > R > P > S. Whenever real situations share this payoff ranking, the same trap appears.
Temptation (defect while other cooperates) > Reward (mutual cooperation) > Punishment (mutual defection) > Sucker (cooperate while other defects). This ordering defines a prisoner's dilemma.
Why cooperation fails: no enforceable agreement
If the prisoners could sign a binding contract to stay silent, both would — the dilemma would vanish. The trap exists precisely because the game is non-cooperative: any promise to stay silent is unenforceable, and each prisoner knows the other has an incentive to break it. Anticipating betrayal, each confesses defensively.
This is why explicit cartels are unstable and why price-fixing is illegal rather than merely unprofitable. OPEC members agree to quotas, but each pumps a little extra because the gain from cheating exceeds the shared gain from restraint — and each knows the others face the same temptation. The logic is identical to the interrogation room.
Real-world dilemmas: arms races, advertising, and the commons
The dilemma structure appears across economics and public policy. In an arms race, each country prefers mutual disarmament to mutual armament, but each fears being the sucker — so both arm. In advertising wars, rival firms would profit more by jointly cutting ad spending, but each gains market share by advertising while the other does not, so both spend heavily. In fisheries, each boat prefers a sustainable catch, but each profits by overfishing while others restrain themselves, so stocks collapse — the tragedy of the commons.
Climate policy is a global prisoner's dilemma: every country benefits if all cut emissions, but each prefers to free-ride on others' cuts. Recognizing the structure is the first step to designing solutions, because the fix must change the payoffs or make cooperation enforceable.
Escaping the dilemma: repetition and punishment
The bleak one-shot result softens when the game repeats. If the prisoners expect to face the same choice many times, cooperation can be sustained by conditional strategies: cooperate now, and punish defection later. The famous tit-for-tat strategy — cooperate first, then copy the opponent's previous move — performed remarkably well in Robert Axelrod's computer tournaments.
Repetition works because the future casts a shadow on the present: the short-term gain from defecting is outweighed by the long-term loss of triggering retaliation, provided players care enough about the future (a high discount factor) and the end of the game is not known with certainty. This explains real cooperation: businesses honor contracts, countries largely respect treaties, and cartels sometimes hold together — all sustained by the threat of future punishment, not by trust alone.
If both players know exactly when the repeated game ends, backward induction unravels cooperation: defect on the last round, hence the second-to-last, and so on. Sustained cooperation needs an indefinite horizon or uncertainty about the end.
Solving one-shot dilemmas: changing the game
When repetition is impossible, the only escape is to change the payoffs or the rules. Governments do this constantly: fines for polluting turn the temptation to pollute into a loss; antitrust leniency programs deliberately create a prisoner's dilemma among cartel members to induce confessions; enforceable contracts with penalties convert promises into binding commitments.
The design lesson generalizes. Whenever you see a socially bad equilibrium sustained by individual incentives — littering, tax evasion, doping in sports — ask what mechanism could alter T, R, P, or S. Mechanism design, the engineering branch of game theory, exists precisely to build rules under which rational self-interest produces socially good outcomes.
Key points
- Confessing is a dominant strategy for both prisoners, yet mutual confession is worse for both than mutual silence
- The dilemma requires the payoff ranking T > R > P > S
- The Nash equilibrium is Pareto dominated: individual rationality conflicts with collective rationality
- Cartel cheating, arms races, ad wars, and overfishing share this structure
- Repeated play with punishment strategies (like tit-for-tat) can sustain cooperation
- One-shot dilemmas are solved by changing payoffs: fines, contracts, or mechanism design
Common misconceptions
✓ Truth: It proves that incentive structures matter. The same players cooperate happily when the game is repeated or agreements are enforceable — the dilemma is about the payoff structure, not human nature.
✓ Truth: That is exactly what the dilemma refutes. Each prisoner plays rationally (a dominant strategy), yet the equilibrium is Pareto inferior to mutual cooperation.
✓ Truth: Cheap talk alone does not help: promises to cooperate are not credible because each player still gains by breaking them. Only binding commitments or repeated-play incentives change the outcome.
Practice questions
Q1. In the prisoner's dilemma, the Nash equilibrium is Pareto dominated because:
- players are irrational
- both players would be better off under mutual cooperation, which neither can reach unilaterally
- the payoffs are not monetary
- communication is physically impossible
Mutual silence gives each prisoner 1 year versus 5 under mutual confession, so it Pareto dominates. But it is not an equilibrium: either prisoner alone gains by deviating to confession, so individual incentives trap them in the worse outcome.
Q2. For a game to be a prisoner's dilemma, the payoffs must satisfy:
- R > T > P > S
- T > R > P > S
- P > R > T > S
- S > P > R > T
T (temptation: defect while the other cooperates) must exceed R (reward: mutual cooperation), which exceeds P (punishment: mutual defection), which exceeds S (sucker's payoff). This ordering creates both the dominant strategy to defect and the Pareto-inferior equilibrium.
Q3. Which real situation is NOT a prisoner's dilemma?
- Two firms deciding whether to start a price war
- Countries deciding whether to build nuclear weapons
- A student deciding whether to study alone for an exam
- Fishing boats deciding how much to catch from a shared stock
Studying alone involves no strategic interdependence with another player's choice — the payoff depends only on one's own effort. The other three all have the T > R > P > S structure with mutual harm from mutual defection.
Q4. Cooperation becomes sustainable in a repeated prisoner's dilemma mainly because:
- players forget past defections
- the threat of future retaliation makes short-term cheating unprofitable
- payoffs are divided equally by a referee
- dominant strategies disappear
Conditional strategies like tit-for-tat punish defection in future rounds. When players value the future enough, the long-run cost of triggering retaliation outweighs the one-time temptation payoff, sustaining cooperation as an equilibrium.
Click a question to reveal the answer and explanation.
Frequently asked questions
Why don't the prisoners just agree to stay silent?
They can talk, but they cannot enforce the agreement. Each knows the other profits by breaking the deal, so the promise is not credible. Without an enforcement mechanism, the dominant strategy remains to confess.
Is mutual confession really a Nash equilibrium?
Yes. A Nash equilibrium requires that no player can gain by unilaterally deviating. If your partner confesses, switching to silence raises your sentence from 5 years to 10 — so confessing is your best response, and the same holds for your partner.
How does the prisoner's dilemma relate to oligopoly?
Oligopolists face the same structure: joint high prices (cooperate) beat a price war (defect), but each firm gains by secretly undercutting while rivals hold prices high. This is why cartels are inherently unstable.
What was Axelrod's tournament and why does it matter?
Political scientist Robert Axelrod invited strategies for a repeated prisoner's dilemma tournament. The simple tit-for-tat strategy won, showing that cooperation can emerge among purely self-interested players when the future matters — a foundational result for understanding real-world cooperation.