Model offers new insight into health exchanges

MIT Management Sloan School
Wednesday, October 19, 2016


A model of health exchanges by Sloan School economist Michael Whinston and colleagues is offering new insight into a particularly challenging management issue for the exchanges.

Health exchanges are insurance marketplaces, in which insurers compete for customers, subject to regulation. The health exchanges are a prominent feature of the Affordable Care Act, and are also used by large companies that self-insure.

The issue Whinston and his colleagues address is the interplay between two key issues facing designers of the exchanges. The first is called “adverse selection,” in which an insurance purchaser has information about his health that the insurer does not. If the purchaser knows he has health problems, he can, thus, buy a comprehensive policy that benefits him, perhaps to the disadvantage of the insurer. The problem for a health exchange created by adverse selection, said Whinston, is that people in poor health would buy the most coverage, making it more expensive because of higher claims. As those insurance policies become more expensive, healthier people would buy less expensive coverage, making the higher coverage even more expensive, leading to a collapse of the exchange, wherein nearly all consumers end up buying the lowest level of allowable coverage.

The second element the model addresses is “reclassification risk,” in which the insurer could in the future require a consumer in poor health to buy more expensive insurance, once it learns of his health problem, for example through a claim.

“If a person’s health status can be priced, and insurance contracts are annual contracts, the problem is that if, for example, a person develops cancer one year, he would have to pay an enormous premium the next year, and it would be as if he didn’t even have insurance,” said Whinston.

Because the Affordable Care Act does not allow insurers to take into account such pre-existing conditions, it protects consumers against reclassification risk, but at the same time it exposes insurers to adverse selection.

The model was developed by Whinston, Ben Handel of the University of California at Berkeley and Igal Hendel of Northwestern University. The model is described in an article “Equilibria in Health Exchanges: Adverse Selection Versus Reclassification Risk,” published in the July 2015 issue of the journal Econometrica, and was recently awarded the Frisch Medal for the best applied paper published in the journal in the last five years.

Using data on individual buyers enrolled in an exchange of a large self-insured employer, the researchers modeled their behavior in an exchange under different assumptions, according to Whinston.

“Our model enabled us to explore the tradeoff between adverse selection and health reclassification. Is it good to ban pricing health conditions—because the problem it creates of adverse selection isn’t as bad as the good thing of eliminating this reclassification risk—or does it go the other way?”

“The answer, we found using the model, is that reclassification risk is a bigger problem and banning firms from pricing health conditions is a good thing,” Whinston said.

More broadly, said Whinston, the model offers a productive approach to analyzing other issues facing designers of health insurance exchanges. For example, he and his colleagues are now exploring the effects of longer-term insurance policies than the one-year policies now allowed under the ACA.