Deconstructing Nepotism Sheheryar Banuri, a Catherine Eckel, a & Rick Wilson b (a) University of Texas, Dallas (b) Rice University April 18, 2011 Abstract We use a laboratory experiment in to test the strength of two motivations for engaging in nepotistic behavior: differential beliefs about the favored group member, and altruism toward that person. The first mover in a trust game is given the ability to choose the group from which the responder is drawn. Subjects can choose between a member of their primary group (in- group) and a member from the general population. We conduct two sets of treatments, one where in-group members are less productive, and another where in-group members are equally productive by design. We also vary the availability of the partner choice mechanism. We find that when nepotism is not costly (i.e., in-group members are as productive as others in the population), both the altruism and belief motivations are evident. However, when nepotism is costly (i.e., in-group members are less productive), the altruism motive is no longer prevalent, while the beliefs motive persists. Furthermore, we find that individuals with a strong group identity, and those who are more risk averse, are more likely to engage in costly nepotism. Finally, implementation of a basic anti-nepotism policy reduces trust and reciprocity, but only when nepotism is costly. Partnering with an in-group member when the choice is available is always profitable, however. These results help to explain why nepotism persists, and the conditions under which it is beneficial. Keywords: Nepotism, Corruption, Group Identity, Discrimination, Trust, Reciprocity