Abstract
The goal of this paper is to study the consequences of non-reciprocal or unilateral altruism, that is, of altruism between individuals who have different concern for others. By contrast to what the literature usually shows-that unilateral altruists lead egoists to cooperate, that non-reciprocal altruism destroys altruism or that it generates non-desirable exploitation-we show that unilateral altruism does not forcedly lead egoists to cooperate nor it destroys altruism and that, in some situations, it can even be Pareto improving. By analyzing a simple cooperation game with other-regarding preferences, we find that unilateral altruism gives birth to a Samaritan's Dilemma where egoists predate Samaritans by free-riding on their contribution. Perhaps counterintuitively, we also show that in case "exploited" Samaritans experience a higher subjective well-being than in a classical Prisoners' dilemma. Finally, we derive conditions for the evolutionary stability of both the predators' and Samaritans' behavior.