We develop a model in which the parties to a joint production project have a choice of specifying contractual performance in terms of actions or deliverables. Penalties for noncompliance are not specified; rather, they are left to the courts under the legal doctrine of compensatory damages. We analyze three scenarios of increasing uncertainty: full information, where implications of partner actions are known; risk, where implications can be probabilistically quantified; and ambiguity, where implications cannot be so quantified. Under full information, action requirements dominate: they always induce the maximum economic value. This dominance vanishes in the risk scenario. Under ambiguity, deliverables specifications can interact with compensatory damages to create a form of ‘ambiguity insurance,’ where ambiguity aversion is assuaged in a way that increases the aggregate, perceived value of the project. This effect does not arise under contracts specifying action requirements. Thus, deliverables contracts may facilitate highly novel joint projects that would otherwise be foregone as a result of excessive uncertainty. Suggested empirical implications include the choice of contract clause type depending on the level of uncertainty in a joint development project, one application being the level of partner experience with interfirm collaborations. This paper was accepted by Bruno Cassiman, business strategy.
