The Genetic Information Nondiscrimination
in Health Insurance Act of 1999:
Too Fast and Too Federal?

By Seth J. Chandler

On June 29, 2000, the United States Senate voted 58-40 to approve legislation known as the Genetic Information Nondiscrimination in Health Insurance Act of 1999 (GINHIA) http://thomas.loc.gov/cgi-bin/query/R?r106:FLD001:S56152-S56155 and to tag it onto a labor appropriation bill (S.6067). The vote was widely reported as a triumph of Republican conservatives over liberal Democrats, whose own similarly named bill on the subject, the Genetic Nondiscrimination in Health Insurance and Employment Act of 1999 (GNHIEA) (S.1322), had just gone down to a narrow 54-44 defeat. These reports miss the point, however, which is not the minor differences between the bills, but the lightening coalescence of political authority in this nation to a view that use of genetic information in health insurance is almost always wrong. For although in certain respects the Democratic bill may have been more far reaching than the Republican bill, they both contain at least one stunning feature in common -- a decision to create a federal bar on underwriting in the voluntary private individual health insurance market based on genetic information. By way of example, both the GINHIA andGNHIEA would prohibit a health insurer from knowing that the individual they are about to sign up knows that they have a genetic predisposition to breast and ovarian cancer.

To be sure, GINHIA's main thrust is to prevent use of genetic information in group insurance policies acquired through employment, the tax-subsidized vehicle by which many Americans obtain health insurance. GINHIA thus strengthens the 1996 Health Insurance Portability and Accountability Act, which had made use of genetic information difficult.

A second thrust of GINHIA, however, is to treat use of genetic information similarly in both group employment policies and the individual health insurance market. Under GINHIA, the Public Health Service Act will be amended to provide that a "health insurance issuer offering health insurance coverage in the individual market" (1) "may not use predictive genetic information as a condition of eligibility of an individual to enroll in individual health insurance coverage"; (2) shall not adjust premium rates for individuals on the basis of predictive genetic information concerning such an individual (including a dependent) or a family member of the individual"; and (3) except for HMO-like entities actually treating the patient, "shall not request or require predictive genetic information concerning any individual (including a dependent) or a family member of the individual (including information about a request for or receipt of genetic services)."

The sort of genetic information prohibited from use by GINHIA both in the group and individual markets includes not only the results of predictive genetic tests, such as those now existing for genes linked with breast cancer and Huntington's disease, but also "information about the occurrence of a disease or disorder in family members." Insurers will thus presumably be barred from continuing their common use of family history questions designed, in a primitive way, to get at individual's genetic predisposition toward various expensive disorders. GINHIA likewise bars insurers from the sneaky technique of inquiring whether an individual applicant has ever taken or received the results of a genetic test.

It is this equivalent treatment of group and individual health insurance that may come back to haunt the nation should GINHIA or something like it become law. (President Clinton has threatened to veto the labor appropriations bill to which GINHIA is attached for unrelated reasons). The two forms of insurance are very different in their susceptibility to adverse selection avalanches. Adverse selection is the process through which insurance applicants who perceive themselves (usually accurately) at higher than average risk buy insurance more frequently and buy more insurance than those who perceive themselves at lesser risk. The result is higher prices and a shrunken market for insurance. Under certain conditions, the process contains a positive feedback loop in which the higher prices further force low risk individuals out of the pool, which in turn raises prices and shrinks the market again.

Group employment insurance is less vulnerable to adverse selection because employees generally cannot significantly vary their purchases based on their perceptions as to whether the insurance is good deal for them. Insurance, at least for the individual employee, is largely compulsory. Research shows that the ability of individuals to vary the amount they spend on insurance greatly affects the potential for adverse selection avalanches. With individual insurance, however, those who perceive themselves at high risk will purchase policies more frequently than those who perceive themselves at low risk. Thus, if individuals can take genetic tests and hide the results from insurers, the results may be little different than if they could hide the results of conventional medical tests (cholesterol, blood pressure) from insurers. Cf. Thomas H. Murray, Genetic Exceptionalism and 'Future Diaries': Is Genetic Information Different From Other Medical Information? in Genetic Secrets: Protecting Privacy and Confidentiality in the Genetic Era (Mark Rothstein ed. 1997) (analyzing whether genetic information is really all that different from information conventionally used in underwriting). Those who have not been blessed with a good "roll of the dice" in the genetic lottery will purchase insurance more frequently (and purchase more insurance) than those more fortunate. Insurance premiums will rise, which will drive more of the less risk averse genetically blessed out of the market, which will in turn raise premiums. The result could be a much smaller individual health insurance market with much higher prices.

Will this adverse selection cascade occur? The truth is, we don't know. It may be that affordable genetic tests with really good predictive capabilities will be very slow to develop. It may be that, even with highly predictive genetic tests whose results can be hidden from insurers, few who find themselves blessed will in fact drop out of the individual market because, with a low safety net in place, the risks of doing so are so great. Theory suggests that fear of being unable to afford first class medical care will deter some low risk insureds from dropping out of an individual health insurance market.

And will the adverse selection problem be confined to the individual market? Again we don't know, because, with respect to voluntary dependent coverage, the employment group market somewhat resembles the individual market. Employees with genetically unfortunate offspring and spouses may choose to enroll them far more frequently than those whose genes do not project a need for expensive medical care.

But in the face of this complete ignorance as to the likely consequences, should we be leaping to action on a federal level or leaving the matter, as used to be done generally in the insurance field, to the states for experimentation? At least twenty five states have passed laws against genetic discrimination in health insurance and about ten have done so with respect to life insurance. These laws vary in their strength and scope. Should we see whether and how voluntary, private individual health insurance can coexist with a regime in which people have far more access to predictive information about their medical futures than used to be the case? Should the rules for genetic tests depend on whether the information gained thereby can be used seriously to improve health? The political consensus, as suggested by widespread Democratic support for the more radical GNHIEA and Republican support for the approved GINHIA, suggests an emergency (prompted perhaps by the recent milestone in the Human Genome Project) that cannot wait for subtleties or experimentation. If an adverse selection avalanche in fact results, however, the real emergency is yet to come.

08/03/00