February 19, 2007
Should Genetic Testing Be Forbidden?
Should Genetic Testing Be Forbidden?--Posner
Congress is on the verge of passing a bill that will forbid employers to discriminate against employees (including applicants for employments) on the basis of the results of genetic tests, and forbid health insurers to deny insurance or charge higher premiums on the basis of such results. (Actually, the bill tightens up an existing law that was designed to do the same things but turned out to have loopholes.) The stated rationale of the bill is that it will encourage people to obtain such tests and use the results to seek treatment or make other decisions, such as deciding whether to have children. That rationale is dubious for several reasons. First, people who suspect they have a gene that causes or predisposes them to a serious disease have a strong incentive to be tested (especially if there are treatment options), an incentive that will often override the possible adverse effect of a positive test result on employment or insurance. Second, in the absence of the law, employers and insurers could make such testing a condition of employment or insurance. Third, persons who are confident that they do not have a genetic defect have an incentive to test voluntarily and disclose their negative results to employers or insurers--and some of these persons will be mistaken and discover that they indeed have such a defect.
So while some people are doubtless deterred from testing by concern with the effect on their employability or insurability, on balance it is unlikely that there will be more testing by virtue of the new law. In a strict efficiency analysis, moreover, even if more people who are likely to have genetic defects will test for them as a result of the law, this would not necessarily be an argument in favor of the law. There is no increase in efficiency when a person conceals information (or avoids obtaining information that he fears he would have to try conceal if he did obtain it) in order to obtain a benefit that he would not obtain if he disclosed it. This would be obvious if a person who knew he was deathly ill bought a huge life insurance policy, concealing his illness from the insurer. The situation is no different if the person knows he may be deathly ill and decides not to verify his suspicion lest the confirmation of it prevent him from obtaining the insurance policy. In either case he is shifting his own expected costs (whether reduced longevity or medical expenses) to unconsenting others.
Analysis is complicated, however, by the possibility that a failure to test brought about by fear of the consequences for insurance or employment would impose costs on other people. That would happen if a prompt diagnosis would enable treatment of a genetic defect at a lower cost, assuming that treatment expenses are paid for in part at least by third parties. Then those third parties would be better off if the person tested. Suppose for example that had the person tested positive, she would not have had a child; instead she had the child, and it is badly deformed, requiring enormous medical expenses paid largely by third parties.
That would be a genuine externality, whereas if the cost of a medical treatment is merely shifted from the individual to his employer or insurer (which means, of course, to the other insureds of this insurer), the externality would be merely pecuniary. That is, it would be merely a transfer of wealth rather than an avoidable investment of scarce resources, as in the example just given where a medical expense is incurred that would not have been incurred had it not been for the failure to test. But transfers often and here are likely to have such effects, and not merely to alter the distribution of wealth. The cost of health insurance will rise if the new law goes into effect, and that rise will increase the number of persons who do not have health insurance, and their lack of insurance coverage may cause them to forgo tests and treatments that may, just as when a genetic test is forgone, avoid costlier treatments and other adverse consequences later on. Employers' labor costs will rise too, resulting in lower net wages; and health is positively correlated with income, so again the transfer will have secondary effects in the form of more ill health.
So even if the new law led to more genetic testing--which probably it would not do, for the reasons stated at the outset--its social costs, from the standpoint of economic efficiency, would probably be negative.
The law might seem defensible on noneconomic grounds as a form of social insurance, since persons who test positive for genetic defects may be unable to obtain private health insurance. The broader point is that the more that science reduces uncertainty about individuals' health, the less risk pooling there will be and the greater, therefore, the demand for social insurance. In the limit, if everyone's health prospects were known with certainty, there would be no market for health insurance at all and this would exacerbate the effects of differential health on equality of wealth; no longer would the healthy be paying to insure the unhealthy.
If social insurance is desired, the question becomes whether to finance it through taxes or, as under the proposed law, to compel private industry to provide it. The major difference is the identity of the "taxpayers": it is federal taxpayers in the first case and the members of the private insurers' insurance pools in the second. The allocative effects of the social insurance "tax" will differ because higher income taxes do not have the same behavioral effects as higher health-insurance premiums. The higher premiums cause people to leave the insurance pool; given current political concerns with the number of people who do not have health insurance, placing the "tax" on those who do have such insurance is questionable.
Eighty percent of Americans tell pollsters that they do not think that health insurers should be allowed to deny coverage or charge higher premiums to people with genetic defects. This is an example of Americans' economic illiteracy.
Comment on Genetic Testing-BECKER
Genetic testing offers great hope for future progress in treating diseases since individuals can in many instances reduce the consequences of genetic defects if they find out about them sufficiently early. For example, a woman who discovers through a simple blood that she has a BRCA1 or BRCA 2 defect has an extremely high chance of getting breast cancer if she does nothing about it. However, she can greatly reduce her risk of breast cancer by having her ovaries removed, and by other more extreme surgeries.
On the other hand, if little can be done to combat the adverse health consequences from having serious genetic defects, individuals may not want to know about them. This would explain why many persons with a high chance of having a genetic mutation that guarantees they will eventually get Huntington's Disease never take the simple test to determine whether they have this mutation since nothing is known yet about how to moderate the deadly consequences of Huntington's.
Particularly where successful health interventions can reduce the medical consequences of a genetic defect, individuals would like to have health insurance prior to taking a test to determine whether they carry that defect. Insurance prior to testing is a viable form of insurance since persons with similar risks ex ante can be pooled in determining insurance rates. Companies would be willing to offer insurance on genetic defects prior to testing since at that stage they would generally be at least as well informed, and often would be better informed, than individuals are about the chances of their having genetic defects. The premium for coverage of any particular genetic defect would then be determined by the product of the probability of having that defect multiplied by the cost of treating it. The premium would vary across different pools of individuals who have different probabilities of testing positive, and different treatment costs. The added premium to cover this risk would be small if the defect was not common, or if treatment costs were not major.
One major problem arises when individuals get tested before seeking insurance, and only take the extra insurance coverage when they discover that they carry a serious genetic defect. Of course, insurance companies would ask applicants if they have already been tested, but applicants for insurance conceal information about their HIV status and other diseases that would raise the cost of coverage if they revealed positive test results. The fear of insurance companies that they would be less well informed about these risks than those applying for insurance because applicants conceal information about testing is what makes these companies want to test many applicants before offering them insurance. Unfortunately, there is nothing to insure after testing since those individuals who test negative do not want coverage, while in the absence of controls over premiums, individuals who test positive would have to pay the full cost of any treatment.
The looming Congressional law on testing would try to handle this problem by forbidding insurance companies to raise premiums to individuals with genetic defects. This only introduces one more regulation in insurance markets, and there are already too many of them. A better way is to penalize individuals who withhold information from insurance companies about the outcomes of genetic tests. If the penalties were set high enough to effectively deter lying about test outcomes on applications for insurance, individuals would then have an incentive to seek insurance against the risk of genetic defects prior to taking any genetic test. Under these conditions, companies would offer insurance that would reflect the expected costs prior to testing of insuring individuals with different risks of genetic defects. Companies might require testing of persons after they are insured if knowledge about whether they were carriers of defects would significantly reduce the cost of preventing or moderating the disease if it eventually occurred.
One major problem remains even if insurance companies tested everyone with a reasonable risk of having genetic defects after they had insurance. Individuals who test negative afterwards would not want to continue to pay the extra premium they had been initially charged when they had a higher than average risk of having a genetic defect. They would threaten to change insurance companies if their premiums were not lowered. Insurance companies would then be left in the initially high-risk pool only with those who tested positive, and they would force these persons to pay the higher premiums appropriate for their high risk revealed ex post. Companies would then again want to know prior to providing insurance whether individuals are positive or negative, and they would then adjust their premiums according to the results.
The fundamental problem here is the inability to write long-term individual health insurance contracts that commit both insurance companies and individuals to remain together for an extended time period. This contractual limitation distorts other risks in the insurance market as well. Individuals who are revealed to be healthier than average as they age will seek insurance rates that take account of this information even though initially they paid rates that were adjusted to their expected health status. Individuals who are revealed over time to have worse health than expected will either lose their insurance coverage or face much higher premiums.
Health insurance coverage through large employee groups helps mitigate the problem caused by genetic testing. Individuals who test negative are not likely to change employers in order to take advantage of this information to get lower premiums since this adjustment in premium would generally form a small part of their overall compensation. Similarly, employees who are revealed to have genetic defects do not impose large costs on employers or other employees unless they are a large fraction of total employment. Of course, employer based health insurance under present arrangements is defective because employees who change jobs may have trouble finding employment if they tested positive for a genetic defect. The solution to this problem is to encourage a move toward a system that allows health coverage portability, so that coverage moves with employees when they change jobs.