A game theory model on authority delegation

Abstract

This thesis develops a model for allocation of effort in principal-agent relations for two special situations. In the first case the principal hired an agent but is not sure about the skillfulness of the agent. In the second case the principal is risk averse. We illuminate how the different situations influence the amount of control of the principal and the initiative of the agent. Companies significantly depend on the output and therefore effort of their employees. Initiative of their employees is the force that drives their success. The supervision of employees and the resulting effort the agents will put into fulfilling their tasks are described by a game theoretical principal-agent model. If we take for instance a research and development department, the challenge of the company's corporate management is to balance the control over the output of the agents and on the other hand leave the agent enough room for their decisions. Too much surveillance can jeopardize the motivation of the employee to create valuable results. We base our investigation on the game theoretical model of the principal-agent relation from Aghion and Tirole (Aghion & Tirole, 1997). In their paper they developed a theory of allocation of authority where they could show the balance of loss of control of the principal versus initiative of the agent. Starting from their viewpoint we modify the model and add additional parameters to see how this tradeoff of supervision of principal and initiative of the agent changes. Two situations are analyzed in detail: 1) The ability of the agent is not known Even though a company may have a very sophisticated recruiting process, the project manager is never 100% sure about the abilities of her project members. Some members are perfectly suited for the given task, but others may, for whatever reason, be 'at the wrong place'. The project manager (principal) has only a probability that her agent is skilled. She has to choose the optimal allocation of supervision depending on the possible types of the agent. 2) The principal is risk averse The ability of the agent is known, but the principal is risk averse with regard to her net benefit. We will see how her cautious behavior influences the effort of the agent. Expectation and behavior of the principal directly influences the agent who reacts on the given situation. We investigate the mechanism how the agent responds and describe a way to avoid demotivation. In the last chapter, we present three cases where companies were able to raise the motivation of their employees and at the same time raise the success of the company.