Since the late 1980s (Grossman & Hart, 1986; Hart & Moore 1989), there has been a considerable growth in the literature known as incomplete contract theory (ICT). This literature sets about formalising and extending some of the insights from transaction cost theory (Williamson, 1975, 1985; Klein et al 1978). These include the ideas: that parties to trade fear opportunistic behaviour in the presence of specific investment; that insufficient contractual safeguards can result in inefficient levels of such investment; and that the avoidance of such inefficiencies provides a key element in the theory of the boundaries of the firm. Two assumptions are axiomatic of ICT. The first closely follows transaction cost theory (TCT) in that many important investments are observable ex post by economic agents close to a trade, but they are not verifiable in a court of law. In the jargon, they are not contractible. In particular, a contract cannot condition prices (or anything else) on ex post investments. The second is that parties to a contract cannot prevent themselves from renegotiating the terms if it is mutually beneficial to do so (Hart & Moore, 1988). Anticipating this, the parties use the contract in the context of an effective legal system to frame these renegotiations.
The name, incomplete contract theory suggests that the theory's main concern is to consider the limitations of contracts that fail to specify not only investment levels, but also many of the other contingencies that a complete contract might wish to include in an Arrow-Debreu world. The reason for this failure might be due to bounded rationality such that some contingencies cannot be imagined, or to the cost of writing complex contracts. The theory might then ask, for example: how efficient are simple contracts that can specify, at most, only one price, one product specification and one quantity? An efficient contract is one that gives the optimal incentives for both investment and trade. This characterisation of the approach suggests a fairly ad hoc limit on the ability of rational agents to write contracts. However, in practice, much of the literature has avoided this potential criticism (or aspect of reality, depending on your point of view) by adopting one of two directions that finesse the need to specify arbitrary restrictions on the content of contracts.