User-Generated Physician Ratings and Their Effects on Patients' Physician Choices: Evidence from Yelp

We examine whether user-generated physician ratings from online sources signal physician quality information and affect patients' physician choices. We collect physician rating and review data from Yelp, a leading online physician rating platform, and match the rating data with Medicare claims data. First, we empirically examine information included in physician reviews by using machine learning algorithms and Amazon Mechanical Turk. We find that although reviews primarily describe physicians' service quality and interpersonal skills, ratings are positively associated with important conventional measures of clinical quality, including physicians' credentials, their adherence to clinical guidelines, and their patients' risk-adjusted health outcomes. To examine the effects of online ratings on patients' physician choices, we examine the effects of ratings on patient flow, measured by physicians' patient revenue and patient volume. By instrumenting for physicians' ratings with reviewers' leniency (or harshness) in rating other businesses, we find that a 1-star increase in a physician's average rating increases physicians' patient revenue and volume by 1.9% and 1.2%, respectively. Overall, we find that online physician rating platforms can promote efficiency by disseminating important quality information to patients and directing patients to higher-quality physicians.