Evolution of In-group Favoritism
Ethnocentrism is a phenomenon found commonly in human behaviour and involves a wide range of actions and behaviours. One such aspect of ethnocentrism is in-group favoritism, in which individuals preferentially cooperate with others who share something in common. Evolutionary game theorists consider these aspects “tags”, which can range from ethnicity and physical characteristic (hence the greenbeard theory) to religion and behaviour.
To help explain this phenomenon, researchers have built computational models to study in-group favouritism using evolutionary game theory. One such model by Axelrod and Hammond (2006) used a prisoner’s dilemma framework with a spatial structure. The world consists of 50 x 50 spaces, each which may be occupied by an agent. Each turn, agents with random genetic makeup enter the world. Agents then interact, and the payoffs are roughly outlined as follows:
Let C be the cooperative strategy and D be the defect strategy. Payoff is fitness.
(C1,C2): Both agents gain fitness
(C1,D2): 2 gains fitness, 1 loses fitness
(C2,D1): 1 loses fitness, 2 gains fitness
(D2,D2): both payoffs are 0, no change in fitness
Subsequently, agents reproduce with frequency proportional to their fitness. Thus, strategies which are more successful in the model will have more representative agents.
We see immediately from the game theory learned in the Network course that there is a clear Nash equilibrium of (C1,C2). However, this is not all there is to the model. Agents have their strategy decided when they are introduced to the population via immigration, and always stick to that strategy until they die. It then is possible to imagine that all the agents in the model will be unconditional cooperators. However, this is not the case when ethnic “tags” are added to the population. Altruists which unconditionally cooperate with all others always lose when interacting with defectors. However, in-group favouritists which only cooperate with others of the same race quickly form populations resistant to agents which never cooperate.
Hence, these studies report the spontaneous emergence of the in-group favoritism strategy (some call it the ethnocentric/racist strategy, but these terms encompass more than just in-group favoritism). This is an example of how game theory may help explain real-world phenomena, in this case the deep engraining of in-group favouritism in human history.
Link to article: https://doi.org/10.1177%2F0022002706293470