The Prisoner's Dilemma and Stable Suboptimal Outcomes

1. The Story and the Structure

The Prisoner's Dilemma is typically introduced with a dramatic narrative that clearly illustrates the tension at its heart.

The story of Tanya and Cinque can be represented formally using a payoff matrix. While the "payoffs" in the story are years in prison (where fewer years are better), we typically represent payoffs as utility values (where higher numbers are better).

To analyze this game more generally, we can abstract away from the specific story and define the game by its ordinal payoff structure. Let the two strategies be Cooperate ($\text{bC}$, remaining silent) and Defect ($\text{bD}$, confessing). The payoffs are:

• $T$: Temptation to defect when the other cooperates.
• $R$: Reward for mutual cooperation.
• $P$: Punishment for mutual defection.
• $S$: Sucker's payoff for cooperating when the other defects.

The Prisoner's Dilemma is defined by a single, crucial inequality:

For the dilemma to be particularly sharp, it's also often assumed that the reward for mutual cooperation is better than the average of the temptation and sucker payoffs, $R>(T+S)/2$, which prevents players from wanting to simply take turns defecting on each other.

2. Individual Rationality Leads to Collective Failure

With the game's structure established, we can now analyze it using the concepts from our previous lessons. The key to understanding the outcome lies in identifying each player's dominant strategy.

As the video demonstrates, the strategy to Defect strictly dominates the strategy to Cooperate. Let's verify this with our $T,R,P,S$ notation for the row player:

• If the column player Cooperates, the row player can get $R$ by cooperating or $T$ by defecting. Since $T>R$, they should defect.
• If the column player Defects, the row player can get $S$ by cooperating or $P$ by defecting. Since $P>S$, they should defect.

Since Defect is the best choice no matter what the other player does, it is a strictly dominant strategy. Because the game is symmetric, this is true for both players. As we learned, rational players will always play a dominant strategy. Therefore, the unique outcome predicted by this analysis is for both players to defect, leading to the payoff profile $(P,P)$.

This outcome, (Defect, Defect), is the game's unique Nash Equilibrium. You can verify this using the best-response "underlining" method from our last lesson. In the cell (Defect, Defect), neither player has a unilateral incentive to switch. If Player A switches to Cooperate, their payoff falls from $P$ to $S$.

Now we see the paradox. The Nash Equilibrium is $(P,P)$, but the outcome from mutual cooperation would be $(R,R)$. Since $R>P$, both players would have been better off if they had both cooperated. The stable equilibrium outcome is worse for everyone than an unstable alternative.

This brings us to the concept of Pareto efficiency, a crucial tool for evaluating outcomes.

In the Prisoner's Dilemma, the Nash Equilibrium is the only Pareto-inefficient outcome. Any other outcome is Pareto efficient because you cannot make one person better off without making the other worse off. For example, moving from (Cooperate, Defect) to (Cooperate, Cooperate) makes the defector worse off (from $T$ to $R$). But from (Defect, Defect), a move to (Cooperate, Cooperate) makes both players better off. The pursuit of individual self-interest in a non-cooperative setting locks the players into a mutually detrimental outcome.

3. Historical Context and Applications

Your interest in the history of game theory is well-served by the story of the Prisoner's Dilemma. It wasn't an ancient philosophical puzzle, but a product of Cold War strategic thinking.

The game's structure appears in an astonishing variety of real-world situations, which is why it is so foundational.

Economic Competition

Consider two firms in a duopoly, deciding whether to set a high price (cooperating to keep profits high) or a low price (defecting to capture market share).

Here, the Nash Equilibrium is for both firms to increase output (defect), leading to a price war that lowers profits for both. This is collectively worse than if they could enforce an agreement to hold down output and maintain high prices.

Public Goods and The Tragedy of the Commons

The dilemma extends to scenarios with many players, a concept you might recognize from your work with multi-agent systems. A classic example is the Tragedy of the Commons, where multiple herders share a common pasture.

Other classic examples include:

• Arms Races: Two countries are better off not spending on arms, but each has an incentive to arm itself regardless of what the other does.
• Environmental Pollution: Every firm benefits from polluting (lower costs), but the collective result is a destroyed environment.

In all these cases, the logic is identical: individual incentives undermine a mutually beneficial outcome. This highlights why external enforcement mechanisms—like government regulation, binding contracts, or social norms—are often necessary to achieve cooperation.

Conclusion

In this lesson, we dissected the Prisoner's Dilemma to understand why its stable equilibrium is suboptimal. This paradox is not just a theoretical curiosity; it's a fundamental model for understanding some of the most persistent problems in economics, politics, and social life.

The key takeaways are:

• The Prisoner's Dilemma is defined by the preference ordering $T>R>P>S$.
• The individually rational choice is to play the dominant strategy, which is to Defect.
• This leads to a unique Nash Equilibrium, (Defect, Defect), which is stable because neither player has a unilateral incentive to change their strategy.
• This equilibrium is Pareto-inefficient, as both players would prefer the (Cooperate, Cooperate) outcome, illustrating a stark conflict between individual and collective rationality.

The Prisoner's Dilemma features a single, compelling, but ultimately poor equilibrium. However, many strategic situations are not so straightforward. In our next lesson, we will analyze coordination games, such as the "Battle of the Sexes" and "Stag Hunt," where multiple Nash Equilibria exist. This will force us to consider how players might coordinate their expectations to land on one stable outcome over another.