Essays on contracts and relationships

Abstract

This dissertation contains three independent papers. The first chapter studies the optimal contracting problem with subjective evaluation when the principal can ask the agent to revise his work. The possibility of revision benefits the principal by providing the option value of making another attempt at the work. But it also introduces a new type of incentive problem for the principal: She may ask for revision even if it is inefficient to do so. This new incentive issue for the principal also affects the incentive of the agent: He may procrastinate his effort in anticipation of excessive revision. This results in a trilemma: The optimal contract cannot simultaneously provide for efficient revision, efficient effort, and minimal ex-post surplus destruction. The optimal contract will of necessity contain at least one of the following problems: revision–the principal asks for excessive revision; procrastination—the agent shirks in the early stage; or punishment—excessive surplus destruction at low-quality final output. The second chapter examines the impacts of technology training and buyer-supplier relationships on technology adoption and quality upgrading. We randomly varied subjects of each training group across farmer–exporter clusters—farmers, exporters, both, or none—and provided training on Good Agricultural Practices (GAP). We find that training farmers enhances technology adoption and quality upgrading. Yet, the effects are much stronger when farmers and exporters are trained together. We document a plausible mechanism to explain this finding: joint training improves buyer-supplier relationship, which facilitates contract trade between farmers and exporters. We find no effect of GAP certification eligibility on technology adoption. In the third chapter, I study an infinitely repeated game between an agent and a principal, where the principal has the option to ask the agent to revise his work, and the revision is asked based on subjective evaluation. Revision can improve the output quality, but it also introduces incentive problems when the true motion of revision–whether out of efficiency or greed—is the principal’s private information and thus unknown to the agent. The optimal relational contract speaks to how the principal optimally manages the revision and how the relationship evolves over time. I show that the relationship starts with excessive revision but ends with insufficient revision. As time progresses, both the revision declines and the relationship deteriorates. Moreover, the principal gradually cedes her revision option as the agent gains autonomy on his work.