Essays on The Optimality of Delaying Quality Tests and the Reverse Hold-up Problem

This dissertation consists of two chapters that examine the optimality of delaying quality tests of new products and the effects of cancellation payments on the hold-up problem.

Chapter 1 analyzes the possibility of delaying quality testing of a new product when the market consists of an early adopter and a follower who receive some private information about the quality. In our social learning framework, delaying a test can lead to better informed decisions regarding conducting the test by the regulator because she, along with other market participants, gains more information about the product quality by observing early adopter's informative actions. Our results suggest that waiting can be optimal when testing costs are not extremely high or low, and when exante there is high probability that both consumers will buy a high quality product and abstain from buying a low quality product. However, once the opposite action by the early adopter is observed (e.g. buying what is likely to be a low quality product), this increases the probability of the follower taking the same action. This can result in high expected losses, and delaying the test becomes no longer optimal. It should be conducted in order to correct the follower's course of action.

Chapter 2 examines the effects of various levels of a fixed cancellation payment in a cost-plus type contract on the hold-up problem. The case of high cancellation payment that results in the agent making inefficiently high investment is referred to as the reverse hold-up problem and is of main interest. We also derive the levels of cancellation payment for which optimal level of investment by the agent can be induced and for which a standard hold-up problem arises. We report the results of the laboratory experiment designed to test our theoretical predictions. We find that, in general, participants follow the equilibrium strategy, and when the cancellation payment is set sufficiently high, the principal is held up by the agent most of the time. We find no evidence of fairness concerns that could explain participants' choices.