A Kantian explication to the prisoner's dilemma


A Kantian application to the prisoner’s dilemma:

One might think that Kant’s categorical imperative usually represents him in a very good way. Likewise, the voluntary resignation seen as a universal law is irrational, so we must cooperate. Although, if you take into account that we should not do anything that can not be a universal law independently, one might feel disgusted. The universal renunciation is irrational because the goal of all is to gain value, because if they were indifferent to this value we could rationally universalize that renunciation. So, it is attractive to mention that the objective of our own interest maintains more relevance in choosing the Kantian response to the prisoner’s dilemma than the will as a universal law. But if you consider that you should treat other people as ends and not as means, this dilemma seems to have a solution, to renounce is to treat the other agent as a means to our end. Therefore, we must cooperate.

An example about breaking promises:

If one should not do anything that can not be a universal law, the reason for the promises is to build trust because if they all broke, none would trust those unfulfilled promises and their nature would be exposed. Thus, given that you can not universalize unfulfilled promises, you should not break any promises. To give the word is to acquire a commitment and that is that the obligation acquires value. If one did not care about the value, we would not make promises.


Federico, I agree what you say about means and ends, and how considering people as ends an opportunity for a Kantian equilibrium could arise in prisoner´s dilemma. Roemer in his semminal article “Kantian Equilibrium” talks about the prisoner´s dilema. It is true what you say that in some ocassion it will be irrational to cooperate and then, a Kantian equilibrium cannot emerge. However, Roemer argues that when the utility from cheating is relatively small, cooperation can emerge and then we can find a Kantian solution to the dilemma. This is a contrast to the typical Nash equilibrium for this game, which agents never cooperate, which ecah agent consider the toher one as a mean to his “selfish” objetive to maximize his utility. Although considerering people as ends does not gurantee a cooperative soluction to the prisioner dillema, in some case, the opportunities arise for a Kantian equilibrium. In this paper we can read it in page 11 how to formalize basically what you say