Verification and reputational concerns: An experiment

I investigate the interaction between two solutions to the moral hazard problem in experience goods markets: Quality verification and reputational concerns. I set up an experimental market where a long-lived seller interacts with a sequence of short-lived buyers and where buyers observe previous decisions by sellers. I test the effect of letting buyers verify product quality at a cost in two distinctly different environments: One in which reputational concerns for sellers are induced by way of computerized sellers, and one without induced reputational concerns. My equilibrium analysis shows that with induced reputational concerns, costly quality verification may fully crowd out reputational concerns and lead to inferior market outcomes. However, when reputational concerns are not induced, verification can improve market outcomes. Results from the experiment to some extent corroborate these predictions. First, in the environment where reputational concerns are induced, short-run incentives created by verification almost fully crowd out the reputational concerns. Second, in the environment without induced reputational concerns, verification improves market outcomes by increasing quality and overall market surplus.