This paper establishes a relationship between the observability of common shocks and optimal organizational design under a multiagent moral hazard environment. We show that, with sucient information about common shocks, a cooperative organization can be optimal even if outputs are highly...