This paper provides a signaling-game theoretical foundation for empirically testing the effects of quality report cards in the U.S. health care industry. It shows that, when health care providers face an identical distribution of patient illness severities, a trade-off between multidimensional measures in the existing report cards renders them a mechanism that reveals the providers' qualities without causing them to select patients. However, non-identical patient type distributions between providers, attributed to the referring physician, may force the high-quality provider to shun patients in order to signal himself. Despite this imperfection, the existing report cards cause the minimum provider selection compared with alternative report mechanisms. Since the report cards not only may cause providers to select patients, but also cause patients to select providers, the single difference-in-differences estimates used in previous studies are not su¢ cient to indicate providers' selection behavior, and cannot capture the report cards' long-run welfare effect with short-run data. In an updated empirical framework, a treatment effect will be estimated once every period.
