April 03, 2017

The Evolution of Fairness

Jeffrey Schank, UC Davis

There are many theoretical approaches to explaining fairness, but explaining fairness without leverage (e.g., genetic relatedness, punishment, or retaliation) is especially challenging. The dictator game (DG) is a fairness game without leverage. One player, the dictator, is given a divisible quantity of some resource (typically money) and must decide how much to give a second player, the recipient. The recipient has no counter strategy, so the obvious rational solution to the DG is for the dictator to keep all of the resource and give none to the recipient. In experimental applications of the DG, to control for possible forms of leverage (e.g., relatedness, friendship, etc.), anonymity is maintained between players. Nevertheless, hundreds of experiments using the DG across cultures have shown that people, on average, share nearly 30% of a resource with an anonymous recipient. Theoretically, this result has proven difficult to explain because of the lack of leverage. For example, Hamilton’s rule, which is standardly used to explain altruism and cooperation at the level of the individual, does not formally apply. Using an agent-based model, I show that fairness can evolve among agents that naively and anonymously play the DG. I also show that this framework can be extended to its close cousin the ultimatum game, providing a better explanation of the available Empirical data than models that only rely on leverage.