It is common in human interactions that we are concerned with the actions of others. How can we induce someone (an "agent") to behave in a way that we prefer? One possibility is to pay them for taking the appropriate action. However, it can be difficult to verify his action and thereby difficult to offer financial incentives for correct behavior. This chapter explores the possibility of forming an altruistic relationship with the agent so that he is directly concerned with the effects on others of his actions even without financial incentives.
Section A presents the standard principal-agent model and provides some examples. The formation of the agent's altruism for the principal - which I refer to "investments in loyalty" - is modeled in Section B. Section C applies the model to corporate loyalty and executive compensation.