The prisoners’ dilemma dilemma

You know what happens when you play the prisoner’s dilemma? Everyone loses. After all, we’re self-interested little buggers. It’s a Hobbes eat Hobbes sort of world.

Or is it? In real, one shot games, people cooperate in surprising numbers: more than a third of the time. Even when they’ve been told that the other person has cooperated, and they can thereby stiff them by defecting, people still cooperate. In fact, knowing that the other person has already cooperated actually increases rates of cooperation dramatically.

Fancy that! Cooperation, in a one shot game, with a stranger you’ll never meet again. And no gain to you from cooperating except the rosy glow of being a good egg. Well, on balance, maybe altruism pays off, even if particular episodes of altruism don’t. Wouldn’t it therefore be best to feel good about it as a general principle?

And do you know, when real prisoners play the game, they cooperate too. Those crazy game theoreticians.

One thought on “The prisoners’ dilemma dilemma

  1. The Faceless Bureaucrat says:

    Thanks, KP–very interesting. I would posit that the fallacy of the ‘rational actor’ (upon which much theory–and policy–is based) has done more damage that it was ever worth. Many scholars have based their entire worldview (and careers and doctrines) on the assumed existence of a “homo economicus” who just doesn’t exist. The ways in which ‘real people’ ‘deviate’ from the ideal of the Unitary Rational Actor are myriad: culture, gender, personality–all these ‘externalities’ conspire to render the models centred on the ideal of ‘preference maximisation’ useless, if not down-right harmful. Thanks for chipping away at the edifice. ‘Vive la crise!’

Leave a Reply

Your email address will not be published. Required fields are marked *