Mechanisms for the evolution of prosociality

The thesis explores mechanisms for the evolution of prosocial behaviour through the lens of evolutionary game theory. The first part of the thesis deals with preference evolution. It looks at what type of utility functions are favoured by natural selection in populations with assortment, where individuals preferentially find themselves in interactions with others who are genetically similar. The central claim of this first part is that altruistic preferences – linear combinations of the payoffs to oneself and to one’s interaction partner –, fare at least as well in evolution as “Homo Moralis” preferences (Alger and Weibull, 2013) do.
The second part of the thesis considers the repeated prisoner’s dilemma, and looks at how dynamics, and selection for cooperative outcomes, change if 1) players can choose to leave their current interaction partner and re-match with a randomly drawn new interaction partner, or if 2) players’ actions are subject to execution errors. Combining analytical results with stochastic simulations of the population dynamics, this part shows that both of these modifications to the repeated prisoner’s dilemma tend to promote the evolution of cooperative outcomes.