Moral Hazard In A Principal-Agent(S) Team

We look at the principal's problem in a principal-agent(s) (p06sibly more than one agent) moral hazard problem, which, unlike most existing work, does not preclude the principal from active participation in the production process. Also, there is no uncertainty. But joint production, which renders the action of each individual in the production process unobservable, causes the moral hazard problem. The principal and all the agents playa multi-stage extensive game, called the Second Boot Game, which determines the set of individuals who actually participate in production along with the output sharing rule they follow l. Although the principal is not precluded from active participation in the production proce8B, we characterize a condition that determines whether she actually takes part in production or not. We also draw the following conclusions- (i) the principal need not look for any output sharing ,rule more sophisticated than th06C that belong to the class of commonly observed linear or piecewise linear variety (ii). the principal can completely mitigate moral hazard whenever she does not participate in production (iii) however, even when the principal does not participate in production at an optimal outcome, she may still be unable to sustain efficiency, (iv) the principal can sustain efficiency if and only if her best option in the First Best situation does not require her participation in production and (v) there are no significant changes to the results when limited liability is imposed. We argue that most of the results are driven by the deterministic production process and not by the quasilinear form of the utility functions that we use. Hence, as long as the production process is deterministic, most of the results will hold qualitatively even when individuals do not have quasilinear utilities but their utilities remain additively separable and Concave.