You are here

Prisoner's Dilemma: Cooperation or Competition?

15 January, 2016 - 09:10
Available under Creative Commons-ShareAlike 4.0 International License. Download for free at http://cnx.org/contents/05c97be4-3ad0-47f2-b5a7-a75d0ad90ab7@3.72

Scholarly debates on corporate governance have turned on the advocacy of different approaches, many of which can be modeled mathematically. Two approaches are based on the concepts of agency and stewardship. (See Davis et. al. in Clarke 2004) To enter into this debate, you will reenact the "Prisoner's Dilemma." Imagine that two patriotic spies, A and B, have just been captured by the enemy. Both are placed in separate interrogation cells and are being pressured to confess and provide details about their spying activities. A and B would like to coordinate their actions but the enemy has kept them apart to prevent this. Their objective is to pit A against B another in order to get the desired information. To do this, they have set forth the following systems of motivations, i.e., punishments and rewards.

Options for the Prisoners

  • If both A and B confess. A and B are put in jail for five years each. The net loss in this scenario is 10. This is the least desirable alternative from the collective standpoint.
  • If one confesses and the other does not. The confessor is released immediately while the non-confessor gets seven years in prison. This maximizes the confessor's self interest but severely punishes the patriotic, non-confessor. Net loss is 7.
  • If both do not confess. After six months of half-hearted interrogation (most of this time is for processing the prisoners' release), both are set free for lack of evidence. While not maximizing self interest (this lies in confessing while the other remains silent) this does maximizes overall welfare by producing a net loss of only 1.

Prisoner Dillema Options Summarized

Prisoner A / Prisoner B

Confess

Not Confess

Confess

Both go to jail for 5 years (Net loss is 10)

A goes to jail for 7 years. B is released. (Net loss is 7)

Not Confess

B goes to jail for 7 years. A is released (Net loss is 7)

Both held for six months, then released. (Net loss is 1.0)

Assumptions in the Prisoner Dilemma

  • Cooperation produces the best collective option and the second best individual option. This, in turn, assumes that cooperation produces more social welfare than competition.
  • Free riding (competing) on the cooperation of others produces the most individual gains (for the free rider) but the second worst collective results. Society suffers loses from the harm done to the trusting, non-confessor and from the overall loss of trust caused by unpunished free-riding.
  • Unlimited, pure competition (both prisoners confess) produces the worst collective results and the second worst individual results.
  • Multiple iterations of the prisoner's dilemma eventually lead to cooperative behavior. But what causes this? (1) The trust that emerges as the prisoners, through repeated iterations, come to rely on one another? Or (2) the fear of "tit-for-tat" responses, i.e., that free riding on the part of one player will be punished by free riding on the part of the other in future iterations?
  • Does the Prisoner's Dilemma assume that each player is a rational, self-interest maximizer? Are the players necessarily selfish in that they will seek to maximize self interest even at the expense of the other players unless rewards and punishments are imposed onto the playing situation from the outside?

The Prisoner's Dilemma is designed to model the reality of corporate governance where the directors/owners of a corporation delegate responsibility for the corporation's operations to managers who are charged with pursuing, not their own interests, but those of their directors. The problem of corporate governance is how this cooperative arrangement is institutionalized. Can managers be left alone and trusted to pursue the best interests of the corporation? This is implied in stewardship theory. Or is it necessary to design a system of controls to keep the managers from diverting the operations of the corporation toward their exclusive, self-interests? This is the approach taken in agency theory. Modeling this in terms of repeated iterations of the prisoner's dilemma, does cooperation emerge as the most reliable strategy in the long run? Or does it need to be manufactured by introducing a system of incentives such as fear of tit-for-tat strategies? The Prisoner's Dilemma models the central problems of corporate governance by asking whether cooperation naturally emerges between managers and directors or whether it needs to be manufactured through a system of punishments and rewards.

The Prisoner's dilemma is discussed throughout the literature in business ethics. For a novel and insightful discussion in the context of corporate responsibility see Peter A. French, 1995 Corporate Ethics from Harcourt Brace College Publishers.