Three Essays on Guilt Aversion : Theory and Experiments.

Summary The purpose of this thesis was to question the sphere of influence of guilt aversion by investigating (i) the direction of guilt: can individuals have guilt aversion toward someone who is not affected monetarily by their decisions, or only toward someone whose income is affected? (ii) some necessary conditions for the emergence of guilt aversion: does the vulnerability of the person towards whom individuals may feel guilty have an impact on the emergence of guilt? (iii) the robustness of guilt aversion to selfish biases: are individuals strategic in their acquisition of information about the expectations of others in order to avoid triggering their guilt? On the one hand, our work revealed for the first time that people can be guilt averse even toward people who are not financially affected by their decisions. This finding, obtained in the context of an embezzlement game (Chapter 1), was extended to new contexts where guilt aversion was systematically observed toward gamblers regardless of their vulnerability (Chapter 2). On the other hand, although guilt aversion appears to generalize to a variety of situations, we have shown that its robustness can be challenged in situations where decision makers have the opportunity to avoid the tension between their monetary incentives and belief concerns (Chapter 3).