Credence goods, consumer feedback and (in)efficiency

We analyze the effects of consumer feedback on a credence goods market. We present a model inspired by Dulleck and Kerschbamer (2006) where consumers sequentially visit a monopolistic expert. Each consumer faces a problem which can be either minor or major. The expert performs a diagnosis that may or may not reveal the severity of the problem faced by each consumer. He then implements a treatment which can solve the problem or not. After visiting the expert, each consumer reveals the received treatment and its outcome, i.e., whether it solved her problem. Each consumer receives the feedback from all previous consumers and uses it to update her belief about the informativeness of the expert's diagnosis. She then decides whether to visit the expert. We show that consumer feedback can lead to inefficiency. More precisely, when the diagnosis fails, the expert overtreats consumers whereas the probability of a major problem is sufficiently low. This behavior does not arise without consumer feedback.