When we study individuals interacting in laboratory environments, it is clear that some consistently observed behaviours are at odds with pure material self-interest (Fehr and Schmidt (2003) and Camerer (2003) present good surveys of the experimental evidence). Several motivations have been proposed to explain such departures from the homo oeconomicus paradigm. Amongst them we can find altruism (Fehr and Fischbacher (2003)), inequity aversion (Fehr and Schmidt (1999)) and preferences for efficiency (Charness and Rabin (2002))). We might place these explanations under the category of ``other regarding concerns''. Subjects may also take into account their social image (Bénabou and Tirole (2006)), or simply want to avoid being punished for not following a social standard. Finally, they might have moral concerns of a Kantian flavour (Alger and Weibull (2013)). That is, they might care about the hypothetical scenario where their actions are imitated by other subjects (universalisation) and also be able to put themselves in their counterpart's shoes. In spite of the large body of research on pro-social preferences, few efforts have been devoted to disentangling them. This is particularly relevant inasmuch they may lead to similar behaviours in a given context, but produce strikingly different results in another, due to their differing underlying logic. In this paper we contribute to filling this gap and specifically aim to distinguishing other-regarding from Kantian moral concerns. Our experimental design has subjects deciding between two actions that affect their own payoff and that of another participant with whom they have been paired. Crucially, other-regarding preferences make individuals choose the action that leads to their preferred final payoff distribution. In turn, Kantian moral concerns lead them to evaluate actions in light of the fact that roles could have been reversed. We aim to exploit this difference in order to distinguish the importance of other-regarding concerns and Kantian moral motivations. More specifically, we employ a treatment where we make it explicit to some participants that they could have found themselves in the other player's position, making them decide behind the ``Veil of Ignorance'', and compare their decisions with those of subjects who are just asked to make a choice given that they are in a decider role.

Invitation Link