The Impacts of Performance Pay for Hospitals: The Readmissions Reduction Program

US health policy increasingly ties payments for providers to performance on quality measures, though little empirical evidence guides the design of such incentives. I deploy administrative Medicare claims data to study a large federal program which penalizes hospitals with high rates of repeat hospitalizations (`readmissions'). The research design exploits introduction of the penalty and policy-driven variation in penalty across hospitals for identification. I construct a measure of hospitals' expectation of exceeding the penalty threshold in each year, based on their observed readmission rate in previous years. OLS using this constructed measure is potentially biased, with mean reversion being a particular concern. I circumvent this problem by using an instrumental variable approach, which also mitigates concerns due to measurement error in constructing hospital beliefs. I use two alternate instruments to implement this strategy and find similar results. I find that hospital responses to the penalty account for two-third of the observed decrease in readmissions over this period, as well as a 2% decrease in one-year mortality. To place this in context, if this penalty was extended to apply to all Medicare hospital stays, nearly 70,000 hospitalizations would be avoided annually and save approximately $620 million just in Medicare spending. Several key facts point to the causal interpretation of these results. First, there are no differential pre-trends across hospitals at different levels of penalty incentive. Second, the timing of the decrease coincides with the introduction of the penalty. Third, I find small and statistically insignificant effects on the readmission rate over 31-60 days, which was not penalized. Although not a strict placebo test, it reassures us that the estimated effect on thirty day readmissions is not driven by macro trends or other contemporaneous changes in hospital payments. Finally, these results are robust to changes in key assumptions underlying the analysis. Applying the same research design, I quantify the role of two mechanisms. Quality improvement accounts for only about half of the decrease in readmission, the remainder is explained by selective admission of returning patients. The change in admission decisions appears to be driven by the penalty since I find evidence of a large decrease in the probability of hospitals readmitting their own, penalty-inducing patients when they return to the ED within thirty days, but no corresponding effect on readmission for non-penalty inducing patients originally admitted at a different hospital. Changes at the ED appear to have been effected through the increased use of observation status for penalty-inducing patients.