Learning and Evolution of the Workers' Altruistic Behavior

Abstract

The evolution of the workers' altruistic behavior and the evolutionary stability of the extent of altruism are studied by employing the theory of learning in games. It is assumed that the firm has two contract strategies and maximizes its profit. The workers have two strategies: altruism and egoism, and they can update their strategies following certain learning rule. The evolution of the altruistic behaviour follows the replicator dynamics of one population and stochastic system with aggregate shocks, respectively. It is shown that the evolution of altruistic behavior in the population exhibits a path-dependent property. However, the unique evolutionarily stable equilibrium is always a mixed-strategy equilibrium, which mainly depends on the population size, the utility difference of the two types of workers and the ratio of benefits that an egoistical worker gets from shirking to the loss of the firm's output. The small population size or the small utility difference between two types of workers favours altruism. These findings are consistent with the results from experimental economics.