Economic Theory Seminar Series: The evolutionary origins of our quirky altruistic preferences: some toy models, laboratory experiments, and field evidence

Moshe Hoffman | Harvard

Why are people's altruistic preferences so strange? For instance, we give a lot but often with minimal concern for efficacy ("ineffective giving"), we give when asked but avoid being asked ("avoiding the ask"), we avoid being selfish when we know it will cause harm (strategic ignorance) but avoid determining the harmful consequences of our selfish acts, and we avoid taking actions that cause harm more so than we are willing to take actions knowing harm will thus be avoided (omission-commission distinction). Our giving is also quite sensitive to norms, subtle cues of observability, and non-binding commitments. We argue that much of our altruistic preferences evolved in response to reputational benefits. Since reputational benefits are enforced by higher order punishment or rewards (others punish or reward good behavior in anticipation that they in turn will be rewarded or avoid punishment for doing so), higher order beliefs are crucial; witnesses to a transgression (good deed) are only incentivized to punish (reward) the transgression (good deed) if they know others are aware of the transgression (good deed) as well. We argue that the aforementioned quirks are all consistent with the above account. Moreover, the above account also predicts 1) that to the extent that altruism is not driven by reputational benefits, as with kin, the above quirks should appear less 2) third party reward/punishment should likewise be sensitive to the above quirks 3) we should be able to use these insights to better design interventions that get people to give more or give more effectively.