Testing the Hirshleifer-Riley Model: The Values of Information Sources for a Future Hospital Stay

This study tests whether the Hirshleifer-Riley (HR) model predicts the values of information sources for a future hospital admission. The main testable prediction of that model concerns the values of information sources for those who intend to choose the same hospital again and those who intend to choose a different hospital. Satisfaction with the prior choice should be negatively correlated with the values of information sources for intentional "stayers," but positively correlated with the values of information sources for intentional "switchers." The authors had a dataset comprising a sample of employees and spouses at a large employer who had been hospitalized during the past year. Respondents were asked to name the hospital(s) they would consider for a future overnight stay, as well as the values of three information sources: their physician's recommendation, family or friends' recommendation, and quality ratings comparing hospitals in the community. Analysis of the responses showed that moderately and highly satisfied consumers who intend to use the same hospital have lower values of quality ratings and that moderately and highly satisfied consumers who intend to switch hospitals have weakly significant, higher values of a physician's recommendation. Otherwise, the HR model's predictions are not supported. There is broader support for the idea that consumers who care about the attributes of the hospital-reputation, medical services, and amenities-have higher values for information sources. The findings suggest that "report cards" comparing hospital quality will be used by only a subset of consumers.