Eva I. Hoppe and Patrick W. Schmitz
Abstract. In a laboratory experiment with 754 participants, we study the canonical one-shot moral hazard problem, comparing treatments with unobservable effort to benchmark treatments with verifiable effort. In our experiment, the players endogenously negotiate contracts. In line with contract theory, the contractibility of the outcome plays a crucial role when effort is a hidden action. If the outcome is contractible, most players overcome the hidden action problem by agreeing on incentive-compatible contracts. Communication is helpful, since it may reduce strategic uncertainty. If the outcome is non-contractible, in most cases low effort is chosen whenever effort is a hidden action. However, communication leads the players to agree on larger wages and substantially mitigates the underprovision of effort.
The working paper version is available for download (CEPR Discussion Paper 11002).