What is it
The Prisoner’s Dilemma is a problem studied in game theory that shows how two people might not cooperate even if it's in both their best interests to do so. It highlights the inherent tension between individual interests and a larger society. Should you pick up your trash at the lunch table? Should you push in your chair after getting up? Should you take performance-enhancing drugs? Should you preserve the earth for the next generation? John and Ken find their mutual interests with Cristina Bicchieri from the University of Pennsylvania, author of The Grammar of Society: The Nature and Dynamics of Social Norms.
Has natural selection designed human beings to cooperate? Can selfishness really be more ‘rational’ than cooperation? Ken and John ask us to consider a thought experiment (credited to Hume) where two neighboring farmers, Ethan and Duncan, each have to plow their fields. Ethan suggests that he help Duncan plow his field today, then Duncan can help Ethan the next day. But, as John points out, is it really rational for Duncan to help Ethan the next day, after his own field has already been plowed? In the cost-benefit analysis, Duncan does not get any benefit from taking on the cost of helping Ethan. This situation is modeled by game theorists as the “Prisoner’s Dilemma,” where two players must cooperate to achieve the group-optimal solution, although each individually rationally benefits from choosing to not cooperate. Cristina Bicchieri, an expert on ‘fairness’ in game theory, joins Ken and John to discuss the philosophical puzzle of cooperation versus conflict.
Why is game theory such a powerful tool? Cristina describes game theory as “simply a method for describing interactive situations.” In our world choices are often interdependent, they depend upon our expectations of how others will make choices. Cristina argues that social conventions such as ‘driving on the right’ can benefit from explanations possible through the models provided by game theory.
Cristina argues that what drives apparently rational agents is their belief in the ‘norm’ of fairness, as well as their belief that other agents will act according to that norm. An astute listener (Lee) asks our hosts to fully define what they mean by fairness. Cristina offers what she considers the three standard definitions: dividing equally, according to need, or according to deserved merit. Of course she readily admits that these three definitions often come into conflict. Do the Wall Street ‘one-percenters’ really deserve their unequal share? How much charity is fair to those below the poverty line?
John points out that we may be motivated to cooperate regardless of what we consider to be fair. After all, he says, if Ethan and Duncan are married it may definitely be in Duncan’s best interests to plow Ethan’s field, if only so that he will be able to sleep in peace. These issues and more are explored in depth, and Cristina argues that we are ‘hard-wired’ to understand norms of fairness but not to cooperate unconditionally, and these fairness preferences can be successfully preserved in a game-theoretic framework. Thus our rationality is preserved in the face of selfishness’ short-term allure.
- Roving Philosophical Reporter (seek to 5:30): Caitlin Esch explores the Prisoner’s Dilemma as it arises in ‘real-life’ scenarios. In the cancelled British game show Golden Balls, in which players compete to win a pool of money, then must eventually choose whether to ‘split’ or ‘steal’ against their competitor. If both split, then they each walk home with half the money. If they both steal, neither get anything. But if one splits and one steals, the stealer gets the entirety of the money. This situation perfectly captures the essence of the Prisoner’s Dilemma.
- 60-Second Philosopher (seek to 49:00): Ian Shoales reminds us that cooperation often appears in unpopular ways, from kindergarten teachers to corporate teamwork building exercises. After all, he says, there may be no ‘I’ in “team,” but there is most certainly a ‘u’ in “suck.”