Are Payoff Maximisers More Pro-social? Evidence from a Prisoner's Dilemma Experiment

Abstract

Are high earners more likely to sacrifice their own payoff to help others? Our paper provides a formal answer by experimentally studying cooperative behaviour in a finitely repeated Prisoner's Dilemma. The subjects play the game in two phases against different types of opponents: other human players (capturing pro-social behaviour) and a known computer algorithm (capturing a player's ability and willingness to earn monetary rewards). Subjects who played a payoff-maximising strategy against a computer algorithm are more cooperative when paired with human subjects, compared to those who did not play a payoff-maximising strategy. The difference in cooperation rates increases as the subjects gain experience with the game.

Keywords: Prisoner's Dilemma, Repeated Games, Payoff Maximisation, Pro-social Behaviour

JEL Classification: C72, C73, C91, H49