Book Description

I consider models in which contracts are written over the verifiable actions taken by an agent in multiple decisions. The principal's preferences over actions depend on underlying states of the world, but only the agent observes the states. The principal cannot audit the agent's information or punish her ex post for having taken inappropriate actions. Moreover, the principal is uncertain about the agent's preferences conditional on the states. Chapter 2 extends the concept of a quota contract to account for discounting and for the possibility of infinitely many periods: a discounted quota fixes the number of expected discounted plays on each action. Discounted quotas are optimal contract forms, even if arbitrary dynamic transfer payments are available, if the agent is assumed to have state-independent preferences. I recursively characterize the optimal discounted quotas for an infinitely repeated problem with independent and identically distributed states. Then I give a more explicit description of these contracts in the limit as interactions become frequent, and when only two actions are available. In Chapter 3 I allow the agent's preferences to depend on the states of the world. Under a variety of assumptions on the timing of the game and on the set of possible agent utility functions, I solve for the max-min optimal mechanisms -- those which maximize the principal's payoff against the worst possible agent preference type. These mechanisms are characterized by a property which I call "aligned delegation." Max-min optimal mechanisms may take the simple forms of simultaneous ranking mechanisms, sequential quotas, or budgets.