Contracts as Reference Points — Experimental

493 In a recent paper Hart and Moore (2008) develop a theory which provides a basis for long-term contracts in the absence of noncontractible investments. The theory is also capable of rationalizing the employment contract, which fixes wages in advance and leaves discretion to the employer. However, the theory rests on strong behavioral assumptions that lack direct empirical support and deviate in important ways from the assumptions made in standard contract theory. For this reason, and because the theory has the potential to cast new light on the theory of the firm, it seems important to test it. In this paper, we provide such a test by carrying out a controlled laboratory experiment. In doing so, we identify new behavioral forces that cannot be explained either by traditional contract theory or by existing behavioral models. These forces are, however, predicted by the Hart-Moore notion that competitively determined contracts serve as reference points. It is useful to start with some background and motivation. According to the standard incomplete contracts literature, trading parties find it difficult to write a longterm contract because the future is hard to foresee. As time passes and uncertainty is resolved, the parties can complete their contract through renegotiation. The typical model supposes symmetric information and no wealth constraints, so that Coasian bargaining ensures ex post efficiency. However, there is a hold-up problem: as a consequence of renegotiation, each party shares some of the fruits of prior (noncontractible) investments with the other party. Anticipating this, each party underinvests.