Working Paper: NBER ID: w9919
Authors: Mark V. Pauly; Yuhui Zeng
Abstract: This paper investigates a possible predictor of adverse selection problems in unsubsidized stand-alone' prescription drug insurance: the persistence of an individual's high spending over multiple years. Using MEDSTAT claims data and data from the Medicare Survey of Current Beneficiaries, we find that persistence is much higher for outpatient drug expenses than for other categories of medical expenses. We then use these estimates to develop a simple and intuitive model of adverse selection in competitive insurance markets and show that this high relative persistence makes it unlikely that unsubsidized drug insurance can be offered for sale, even with premiums partially risk adjusted, without a probable adverse selection death spiral. We show that this outcome can be avoided if drug coverage is bundled with other coverage, and briefly discuss the need either for comprehensive coverage or generous subsidies if adverse selection is to be avoided in private and Medicare insurance markets.
Keywords: adverse selection; prescription drug insurance; insurance markets
JEL Codes: G22; I18
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| persistence of high outpatient drug expenses (H51) | likelihood of adverse selection (D82) |
| high outpatient drug expenses (H51) | seeking insurance coverage (G52) |
| skewed distribution of expected expenses (D39) | adverse selection death spiral (G52) |
| lack of subsidies or bundling (D43) | sustainability of unsubsidized standalone drug insurance (G52) |
| high required premiums (G52) | disincentivization of low-risk individuals from purchasing insurance (G52) |