Expert decision-making: A Markovian approach to studying the agency problem

In this paper, we study the agency problem in an organisation within a Markovian framework. More specifically, the paper presents the case of a principal imposing an incentive-control structure upon an agent to force him to follow the principal's interests for which he was hired, against the tendency of the agent to follow his own interests. Findings point toward the principal's difficulty in controlling the behaviour of the agent through incentives and monitoring; instead, best results are obtained when hiring agents who care for their reputation and refrain from unprofessional behaviours. The implication is that if we consider that it might be difficult to identify this characteristic at the time of the agent's hiring, the best criterion will be to look for low levels of greed in the agent. This conclusion goes in some way against current practices of looking for aggressive agents for the generation of higher profits. Nevertheless, it should be noted that these potential benefits might actually fade away if the agent follows his own interests, instead of the principal's. Another interesting result points to the restricted, although necessary, role of monitoring to control the agent's behaviour, a result that goes against current research interests on measures of corporate governance. The paper is a contribution to expert decision-making.