Disclosing Physician Ratings: Performance Effects and the Difficulty of Altering Ratings Consensus

Abstract

This study addresses the effects of a health care system publicly disclosing patient ratings of some of its physicians. I find evidence that the disclosure leads to: 1) performance improvement as measured by the ratings and by objective quality measures, and 2) a bias among raters, whereby raters positively weight a physician’s published average rating in forming their own subsequent ratings of the physician. This study provides the first evidence of either result in the context of the growing trend of consumer-rating disclosure. I use web traffic data to assess how the performance and bias effects vary with public attention to the disclosed ratings. I find evidence consistent with public attention reinforcing raters' bias toward echoing a physician's past rating, thus impeding rating improvement. On average, the disclosure yields a rating improvement of 17 percentile points in a national distribution of ratings, and biases ratings toward the physician’s published average by 24 percentile points in that distribution. Collectively, the results show that disclosing consumer ratings has performance benefits, but also introduces a bias that works against ratings reflecting changes in service.