Private Games are too Dangerous

• Authors: Ronald S. Burt
• Venue: Computational & Mathematical Organization Theory
• Year: 1999

Given the difficulty of observing interpersonal relationsas they develop within an organization, I use iterated prisoner‘sdilemma games to simulate their development. The goal is tounderstand how trust could develop as a function of private games,that is, as a function of interaction sequences between two peopleindependent of their relationships with other people. My baseline isAxelrod‘s results with TIT for TAT showing that cooperation canemerge as the dominant form of interaction even in a society ofselfish individuals without central authority. I replicate Axelrod‘sresults, then show that the results only occur in a rare socialcontext—maximum density networks. Where people form less densenetworks by withdrawing from unproductive relationships, as istypical in organizations, the competitive advantage shifts from TITfor TAT to abusive strategies. A devious PUSHY strategy wins inmoderate to high density networks. A blatantly HOSTILE strategy winsin less dense networks. Abusive players do well in sparse networksbecause their abuse is lucrative in the initial exchanges of arelationship—before the other person knows to withdraw. Wiseplayers avoiding the abusive players leaves the abusive players freeto concentrate on naive players (con men thrive in big cities). Theimplication is that what keeps abusive players at bay are friends andacquaintances warning managers away from people known to exploittheir colleagues. I reinforce the point with illustrative surveydata to conclude that private games are not only too dangerous, butalso too rare and too slow to be the foundation for trust withinorganizations. The results are an evidential call for thesociological intuition that trust and distrust cannot be understoodindependent of the network context in which they are produced.