INCOMPLETE CONTRACTING: A LABORATORY EXPERIMENTAL ANALYSIS

Incomplete contracts are motivated by difficulties in specifying contingent responses to unforeseen change in long‐term trading relationships. The issues can be parsimoniously represented in a two‐period model. Parties first make transaction‐specific investment that enhances value or reduces cost. Surplus is then realized, and bargaining divides the surplus and completes the contract. The testable hypothesis is that realized surplus shares are independent of sunk investments. This is strongly rejected using laboratory experimental methods. Failure of the prediction affects the expected profitability of contracts.