‘Adverse selection’ and ‘moral hazard’ are phenomenons which affect any analysis of the insurance market. For instance, Cutler and Zeckhauser (1997) speak of an adverse selection ‘death spiral’ which made untenable the continued offering of a generous health insurance benefit at Harvard University. In their NBER paper, Finkelstein and McGarry (2003) attempt to estimate the effect of private information in the long-term care insurance market. In 2000, the authors state that long-term care expenditures in the U.S. comprised 7.5% of all medical expenditures and 1% of GDP for a total of $100 billion. Thus, analyzing this segment of the health insurance market is non-trivial and it will only increase in importance as baby boomers continue to retire in larger numbers.
The first item Finkelstein and McGarry investigate is whether there is a positive correlation between the amount of long-term care insurance purchased and the likelihood of using the insurance benefit. If adverse selection is a problem, one would expect people who purchase large amounts of insurance will also be the ones who incur expensive long-term care costs for the insurance companies. Surprisingly, they encounter no evidence of this. Using a more specific test for adverse selection, however, the authors do find proof that adverse selection is a problem. Examining a supplement to the Health and Retirement Study (HRS), they find that people who expect to enter a nursing home in the next five years are 1) more likely to enter a nursing home and 2) more likely to purchase long-term care insurance.
So what explains the lack of correlation between the amount of long-term care insurance purchased and the likelihood of entering a nursing home if adverse selection is a problem? It turns out that preferences explain the residual. The authors posit that risk averse individuals are more likely to undertake preventative health care measures such as having a flu shot, cholesterol test, mammogram or prostate screen. Using these measures as a proxy for risk aversion, Finkelstein, et al. find that risk adverse individuals are 1) more likely to buy long-term care insurance, but 2) less likely to need the benefit. We see that the effects of adverse selection are offset by the fact that healthier, risk averse individuals also purchase excess amounts of insurance. Thus, in aggregate–at least for the long-term care insurance–this market can reach a stable equilibrium.
Finkelstein and McGarry (2003) “Private information and its effect on market equilibrium: new evidence from long-term care insurance,” NBER WP #9957.