We develop a theory of contracts with competing agencies under symmetric information and under moral hazard. In contrast to much of the literature, we allow the agencies to choose from a general contract space. Under the assumption of supermodularity, we show that in both cases the equilibria in contracts have a characterization in terms of very simple contracts. Using these simple contracts we show that under symmetric information, the set of outcomes supportable in equilibrium coincides with the set of individually rational outcomes. While this characterization is new and possibly useful, the result is well known. However, we show that, the introduction of moral hazard (agent's action unobservable) has a striking effect -- it reduces the supportable set to only a subset of the collusive outcomes. We also show that the simple contracts are optimal, which makes them a relevant tool for policy analysis and makes possible applications to international trade, strategic R&D, oligipolistic competition with contractual delegation and taxation of oligopoly.
