Four aspects of the principal-agent relationship: pure moral hazard, hiring effect under the existence of a state observable only by the agents, creation of divisions under a zero hiring effect, anti-selection when the types of the agents are independent.

Summary This thesis focuses attention on several aspects of contract theory concerning the principal-agent relationship. Chapter 1 is a review of the literature that analyzes the basic principal-agent framework of pure moral hazard. It includes three marginal contributions. Chapter 2 examines competition between hierarchies with one principal and several agents, when contracts are signed under symmetric infromation. The agents hired by the same principal are the only ones to observe the realization of a random state, correlated or not with those of rival agents. We present hypotheses that guarantee the absence of the hiring effect in the agency game. Chapter 3 analyzes games where principals (firms) choose the number of agents (independent divisions) they will hire (create) prior to the choice of contracts, when the hiring effect is zero in each possible agency subgame. We also study the number of principals in equilibrium when there is an entry cost. Chapter 4 analyzes the antiselection relation when a principal is interested in hiring a number of agents with continuous and independent types. We prove that, under certain conditions, there is no loss of generality if the principal considers only veridical mechanisms in dominant strategies. The optimal mechanism can require a higher individual performance than the optimal one of complete information to the most productive agents.