Allocating Control in Agency Problems with Limited Liability and Sequential Hidden Actions

I discuss the optimal organization of sequential agency problems with contractible control actions under limited liability. In each of two stages, a risk-neutral agent can choose an unobservable effort level. A success in the first stage makes effort in the second stage more effective. Should one agent be in control in both stages (integration), or should different agents be in charge of the two actions (separation)? Both modes of organization can be explained on the basis of incentive considerations due to moral hazard, without resorting to commitment problems or ad hoc restrictions on the class of feasible contracts.