Contracts
We Didn't Think of That...

It is usually argued that a contract facilitates trade between two parties who have a long-term relationship involving large specific investments. Once these investments have been made, competition will have little impact on the terms of their trading, so these must be governed instead by contractual provision. A major problem facing the drafters of the contract is to anticipate and deal appropriately with the many contingencies which may arise. In fact since it does not pay to plan for every conceivable eventuality, or one simply cannot do so, contracts will typically contain large gaps. In Discussion Paper No. 60, Research Fellows Oliver Hart and John Moore consider how these gaps might be filled during the course of the trading relationship.

Hart and Moore study contracts in which there are 'gaps', due to the inability of the parties ex-ante to identify the objective events which will ex-post determine their benefits and costs, or how these benefits and costs will be affected by objective events. Inability to specify these objective events is important, because the contract cannot be based directly upon the realized values of benefits and costs. These are only known (jointly) to the parties themselves, and not to the enforcers of the contract - the courts.

This incompleteness in the contract may be less important if it contains a mechanism for revising the terms of trade after the contract is signed, and as information is received concerning benefits and costs. Ex-ante, it will typically be in the parties' interests to prescribe, as far as possible, the sorts of revisions that can later take place. For example if, having signed the contract, they each privately take actions which affect the probability distributions of benefits and costs, then the ex post division of surplus matters because it will affect these actions. Or the parties may be risk averse, in which case the ex post terms of trade will determine the extent of risk- sharing.

The parties can rescind the original contract and negotiate a new one ex post. Hart and Moore discover that this places severe limitations on the price revision schemes that are feasible ex ante. They demonstrate that these limitations depend crucially on what kind of communication mechanism the parties have at their disposal.

For example, suppose the messages sent between parties cannot be publicly recorded. Thus if one party receives a message, he can choose to reveal it in the event of a dispute, but is under no obligation to do so since he can always deny that he received it. In this case, Hart and Moore find that the only feasible price revision scheme is very crude: given that there is a single item to trade, the contract would simply specify a no-trade price and a trading price, which if necessary the buyer can raise and the seller can lower. If on the other hand, a public record exists of all messages sent and received, then other forms of contractual revision are feasible, but they are nonetheless still surprisingly restrictive.

Hart and Moore consider applications of their analysis to two situations: first where the parties are risk averse, and second where they take actions which affect the probable size of the ex post surplus. If messages cannot be publicly recorded, then in both situations the fact that the surplus can only be divided crudely almost always means that the contract is very inefficient. If messages can be publicly recorded, however, then contracts can be written which share risk efficiently in the first model.


Incomplete Contracts and Renegotiation
O Hart and J Moore

Discussion Paper No. 60, April 1985 (ATE)