Quality Disclosure and Regulation: Scoring Design in Medicare Advantage

Abstract:

Regulators often generate quality scores to help consumers with limited information about product quality, as in schooling, healthcare, and financial markets. When designing scores, regulators must not only anticipate how they will influence consumer choices but also the resulting impact on firms’ incentives to invest in quality. In this work, I draw on theoretical insights, econometric strategies, and computational methods to develop an empirical scoring design methodology. I apply it to a large health insurance market and find an alternative policy which vastly improves the market’s performance. The new design coarsens consumers’ information about lower-quality insurance options but refines it for higher-quality ones. Changes to available product information generate a shift in demand towards higher-quality plans, triggering additional firm investments and making consumers better informed about a menu of superior options. The new design also optimally aggregates different quality dimensions, tackling a multitasking moral hazard problem. The friction is due to firms’ (agent) private incentives to attain scores using cost-efficient investments instead of consumer-valued ones, preferred by the regulator (principal). Overall, the alternative policy increases welfare by $669 per enrollee per year. The analysis reveals that simple scores can be remarkably effective if well-designed and provides a method to construct them.