Agency Problems and Reputation in Expert Services: Evidence from Auto Repair


  • I thank Steve Berry, Justin Fox, Don Green, Justin Johnson, Dean Karlan, Josh Lustig, Steve Nafziger, Philipp Schmidt-Dengler, Fiona Scott Morton, Harsha Thirumurthy, Michael Waldman, various seminar participants, and the Editor and anonymous referees for comments, and the Yale Institution for Social and Policy Studies for financial support. I also thank Michael Coughlin at Premier Subaru, Romana Primus at Whaling City Ford and Paul Trembley at East Rock Auto Repair for background information about the auto repair industry, and George Iny and his colleagues at the Automobile Protection Association for invaluable assistance in designing the experiment and interpreting the outcomes.


Using a field experiment involving undercover visits to auto repair garages with a test vehicle, I first examine how asymmetric information between mechanics and motorists over auto repair service quality affects outcomes. I then examine whether reputation mitigates these problems via a matched-pair treatment in which undercover researchers appeared as either one-time or repeat-business customers. The results indicate that under and overtreatment are widespread, and that reputation via a repeat business mechanism does not improve outcomes significantly.