Notes on Incomplete Contracts

Throughout the course, we have assumed that all contracts are complete. This is, of course, complete nonsense in reality but a useful starting point in theory. But why is it complete nonsense in reality? What are we missing? Many stories have been suggested. One possibility is that we can, indeed, think of all possible contingencies, but it is costly to write all of these down in a contracts (Lawyers aren’t cheap). Thus, we economize on the longshots that don’t matter much and write in the parts that do. Another possibility is that we’re simply not able to foresee all possible contingencies so that even if we wanted to write a complete contract and there were no costs in doing so, we simply couldn’t. Still a third possibility is that, while the parties engaging in the contract can observe aspects of the contract perfectly well, the courts are unable or unwilling to verify these pieces. Therefore, whether we write these outcomes down in the contract or mot avails us little in the event of breach. In the last unit on dynamics of adverse selection and moral hazard, we emphasized that renegotiation could (and sometimes did) make things worse o¤. The Principal would obviously like to have full commitment power. All renegotiation did was to constrain the set of outcomes that the Principal could credibly commit to, and this was bad news from his perspective. On the other hand, with incomplete contracts, there’s a new and more positive motive for renegotiation. An outcome has occurred that we simply didn’t contemplate or couldn’t contract upon ex ante and now we’d like to …x the incentive problem going forward. Thus, unlike the complete contracting literature, under incomplete contracts renegotiation now has the possibility of improving contractual outcomes rather than harming them. Much of this literature represents attempts to build formal models to capture the insights of transaction cost economics as described by Coase, Williamson, Klein, Crawford, Alchian , and so on. The key wedge in most of the formal models is the presence of relationship speci…c, but non-contractible investment. This leads in turn to the famous hold-up problem. Decisions by the Principal about the boundaries of the …rm and property rights more generally in response to these issues lie at the heart of many of the formalizations.