Abstract
Cooperation is acting in the interests of one’s social group, often at a cost to yourself. When the level of cooperation is observed in the laboratory, people cooperate more often, and at higher levels than are predicted by standard theories. In this thesis I find two novel ways in which cooperation on multilayered populations is increased. These models contribute to an understanding of how people cooperate in real-world social situations, and help us to explain why people cooperate as much as they are observed to do. In each study I model the tension between the individual and the group using the public goods game. This game is played on a structured population defined by a multilayered network. Each layer represents a different sphere of influence on the player’s decision to cooperate or defect. The first model studies the effect of a player choosing whether to cooperate or defect on either all layers simultaneously (synchronously) or on one layer at a time (asynchronously). Updating asynchronously leads to increased cooperation across a number of different parameter regimes. This demonstrates a new way in which cooperation can be increased in a system with multiple influences, and also helps to understand exactly why cooperation is increased in multilayered systems. Inspired by empirical examples, the second model adds to the standard model of the public goods game on networks in two ways. The first is to include conditional cooperators, and the second is the addition of a layer of social influence. This combination of economic and social influence has not been considered in previous models of the public goods game, and I find that this additional layer of influence results in high levels of cooperation. In the final chapter, I study these dynamics on more realistic network structures, with results echoing empirical findings under certain parameters.