Andrea Marietta Leina (University of East Anglia)

Seminar series “CLOSER, but not quite” – “Cooperation in the Helping Game”

Venue: Campus Luigi Einaudi, room 3 D1 01

Abstract: Using a lab experiment, we examine which of the reputation-based mechanisms proposed in the theoretical literature perform better in the decision to help strangers in two conditions of the so-called “helping game”. We compare the “good standing”, a binary score with a recursive feature, and the “image scoring”, a numerical score that takes into account past actions. We theoretically argue that the “good standing” mechanism creates an incentive to discriminate between subjects who do not help someone that did not help in the past (‘justified punishers’) and subjects who did not help without a justification (‘unjustified non-helpers’). We find that good standing seems to work for discriminating subjects, supporting reciprocal helping behaviour. However, image scoring leads to higher cooperation rates, irrespective of the subject’s reputation.