Printer Friendly

Genetics and the moral mission of health insurance.

Deciding whether genetic differences among individuals are morally relevant to health insurance requires us to ask, What kind of good is health care? and, What principles should govern its distribution? There are good reasons to doubt that "actuarial fairness" is an adequate description of genuine fairness in health insurance.

All men are created equal. So reads one of the United States of America's founding political documents. This stirring affirmation of equality was not meant as a claim that all people are equivalent in all respects. Surely the drafters of the Declaration of Independence and the Constitution were as aware then as we are now of the wondrous variety of human-kind. People differ in their appearance, their talents, and their character, among other things, and those differences matter enormously.

The commitment to equality embodied in our political tradition is not a claim that people, in fact, are indistinguishable from one another. Rather it is an assertion that before this government, this system of laws and courts, all persons are to be given equal standing, and all persons must be treated with equal regard.

Human genetics, in contrast, is a science of inequality--a study of human particularity and difference. One of the most difficult challenges facing us in the coming flood tide of genetic information is how to assimilate these evidences of human differences without undermining our commitment to political, legal, and moral equality.

The information about human differences pouring forth from the science of human genetics provides us with a multitude of opportunities to treat people differently according to some aspect of their genetic makeup. Deciding which uses of this information are just and which are unjust will require us to reexamine the ethical significance of a wide variety of human differences and the larger social purposes of a variety of institutions, among them health, life, and other forms of insurance.

Health insurance in the United States has moved from a system based mostly on community rating where, in a given community, all people pay comparable rates, to a system where the cost to the purchasers of insurance is based on the expected claims--a risk- or experienced-based system. This movement has significant ethical as well as economic overtones. Community rating was a system that reflected a notion of community responsibility for providing health care for its members, where the qualifying principle was community membership. Other differences, such as preexisting risks, did not count as morally relevant distinctions. Risk-and experience-based systems presume that it is fair to charge different prices, or to refuse to insure people entirely, if they will need expensive health care. Such systems treat predicted need for care as a morally relevant difference among persons that justifies differential access to health insurance, and through it, to health care. But this presumes precisely what is in question: what are good moral reasons for treating people differently with respect to access to health insurance and health care?

Risk- and experience-rated systems are now dominant in private health insurance in the United States, having largely replaced community rating. This trend, coupled with the rapid growth in the cost of health insurance for employers, creates a situation in which risk-oriented genetic testing could become an important, complicating feature.

We may create a catch-22 so that only people who are unlikely to need health insurance can afford it. Genetic testing may permit a much more complete and refined classification of people into risk categories, and so move us further away from sharing the financial burdens of illness and further in the direction of individualized premiums based on individual risk factors. Genetic risk testing is important because it exposes the logic of a system that provides access to health insurance to those least likely to need it.

Genetics and Distributive Justice

Distributive justice, as the term implies, concerns the distribution of social goods or ills: in its simplest formulation it holds that like cases are to be treated alike and unlike cases are to be treated differently. All depends, obviously, on how we fill in the material conditions of this purely formal statement of comparative justice. When we are asking about a particular occasion of just or unjust treatment, the question commonly takes the form, What makes these cases like or unlike in a morally relevant way? Failure to state a morally relevent reason for treating people differently opens one to the charge that one's action was arbitrary, capricious, and unjust.

Human genetics provides a large and rapidly growing set of differences among persons that may be used to try to justify unequal treatment. For many genetic differences and many distributions of social goods, the moral relevance of the difference seems transparently obvious. Height, for example, is largely determined by genetics. Does it make any sense to say that it was unfair to allow Kareem Abdul Jabaar to play center in the National Basketball Association for many years, but not me, just because he is taller than I am, and our differences in height are genetic, rather than anything we can claim credit for accomplishing? Most people would judge that to be absurd. In this instance a genetic difference--height--constitutes a morally relevant difference that justifies treating people differently. That same difference, however, would not justify treating us differently if, for example, we were accused of a crime, or being judged on our literary accomplishments, or in need of health care.

A judgment that something is or is not a morally relevant reason does not take place in a social vacuum. As in the case of height, a particular attribute or trait can count as a relevant difference under some circumstances but be utterly irrelevant in others. Whether some particular difference is relevant depends upon what conception of distributive justice applies to the specific case. Michael Walzer describes three concepts that are often taken as primary: free exchange, desert, and need.[1] Free exchange is exemplified by the market where people exchange goods, services, and money. Desert appears to be the operative principle where prizes or championships are being distributed. It is also important in, for example, determining which individuals will be given positions requiring certain skills. Need seems to be an appropriate criterion for allocating certain goods, for example food in a time of scarcity. Most would judge the hoarding of or gorging on food by people whose neighbors are starving as indecent and unjust. This would be true even if the hungry neighbors had nothing to exchange that the person with food wanted; and it would be true even if the starving people had done nothing to "deserve" food--indeed the notion of having to "deserve" a good here seems grossly out of place.

No one of these three criteria for distribution can cover all cases. Variants of one or the other, or sometimes combinations of more than one, can typically be found underlying our considered social judgments about what constitutes a fair distribution of a particular social good. The question that must be addressed here is whether and in what ways genetic (or other health-related) differences among individuals count as morally relevant reasons for treating people differently with respect to insurance, especially health insurance. That, in turn, requires asking what kind of good health care is, and what principle or principles should guide its distribution.

What Kind of Good Is Health Care?

Citizens of the United States do not speak with one voice on the question. We do not all agree as to what kind of good health care is, on what principle or principles it should be allocated, or on what count as morally relevant differences among persons in such allocation. In fact, voices can be found that call for health care to be distributed in whole or in part through the free exchange of the market,[2] desert,[3] or need.[4] Nevertheless, we can examine our public policies and public debates for evidences of what we consider just in the allocation of health care. We have two massive public programs, Medicare and Medicaid, to provide health care to groups of people who would otherwise find it difficult or impossible to obtain--the old and the poor. The concerns manifested by these two programs cannot be explained by principles of free exchange or desert; rather they appear to reflect an underlying notion of health care as a response to need.

Health insurance, whether public or private, figures in to the extent that it affects access to health care. If one's access to health care were perfectly correlated with whether one had health insurance--that is, one could get health care if and only if one had health insurance--then whatever we concluded about justice and health care would apply equally to health insurance as the instrument by which people gain access to care. In practice, the correlation is far from perfect. People with health insurance may not get the care they need for a variety of reasons, such as other obstacles to access--for example, geography. Also, lack of health insurance does not automatically mean lack of health care; people with acute health care needs can usually find an individual or institution to treat them. Such "uncompensated care" does not solve the problem of lack of access to care via public programs or private insurance, as the timing of care is often far from optimal and the struggle to get it humiliating. But it does provide a further clue that we regard health care as a response to a need, rather than something that can be left entirely to the market or treated as a matter of desert.

Imagine someone entering an electronics store and pleading to be given a VCR. They cannot pay for it (free exchange). They might try to claim they deserve it (perhaps because they actually can operate the timed-record function), but that is unlikely to persuade the store owner. They could say that they need a VCR. But in the absence of some highly unusual circumstances tying possession of a VCR into some genuine need, their request will be dismissed. People do not need VCRs, certainly not in the way they sometimes need health care. In contrast, imagine someone going to a hospital emergency room bleeding badly. We know that such a person should be treated immediately and the problem of how to pay for that treatment sorted out afterwards.

Having health insurance is a way to pay for such treatment--the cost of treating a serious illness cna easily exceed an average family's ability to pay for it. Health insurance is, for most people, the means to the end of health care. It is not the good of health care itself. But to the extent that it determines who does and who does not have access to care, and who has the peace of mind that comes with knowing that if care is needed it will be available, access to health insurance is a matter of justice.

Genetic Testing: The Challenge for Health Insurance

Research in human genetics, such as the Genome Project, is likely to increase dramatically our ability to predict whether individuals are at risk for particular diseases. There are tests currently offered for diseases such as Huntington's, where the presence of the gene assures that the individual will develop the disease if he or she lives long enough. There are tests for carrier status such as cystic fibrosis where two copies of the defective gene--one from each parent--must be inherited in order for symptomatic disease to occur. And there will be tests for diseases of complex etiology such as heart disease, cancer, stroke, lung disease, and the like. For certain relatively rare genes there will be a strong connection between having the gene and having the disease. Yet most of the common killing and disabling diseases are more likely to have a complex variety of causes, including perhaps several genes each of which has some predictive relationship with the disease. These risk-oriented genetic predictors potentially are very interesting to employers and insurers.

Genetic information, in fact, is used now by insurers. There may be considerable genetic information in one's medical record. If your policy is being individually underwritten, that entire record can be copied and shipped to the prospective insurer and that information used to justify increasing the price or denying health insurance altogether. But this begs a prior question: should information about genetic differences be used at all in health insurance?

One argument against paying any special attention to genetic predictors of risk is that insurers already use risk predictors that have genetic components. Coronary artery disease is an example. It is well known that people with higher levels of cholesterol, especially the low-density-lipoprotein component, are at higher risk of coronary artery disease and subsequent heart attacks. It also seems clear that an individual's cholesterol level is at least in part determined by genetics. Variations in individual metabolism can have a substantial impact on a person's cholesterol level, such that two people can be equally virtuous (or careless) in diet and exercise and yet have very different cholesterol levels, and, presumably, very different risks for coronary artery disease and heart attack.

In time it is likely that researchers will discover a number of genes that affect cholesterol metabolism and, presumably, cholesterol level, arterial disease, and the risk of a heart attack. We may be able to construct a genetic profile of an individual's risk of heart disease. Does such a predictive index differ in any ethically significant way from today's cholesterol test, which has not evoked similar objections?

Genetic tests differ from a cholesterol test in that the latter, even if significantly influenced by genetics, is still in some measure under the individual's control. The risk of heart attack is affected by a variety of health-related behaviors including diet, exercise, stress, and smoking. To the extent that people can be held responsible for their behavior, their cholesterol level is something for which they have some responsibility. On the other hand, people cannot be said to be responsible for their genes. An old maxim in ethics is "Ought implies can." You should not be held morally accountable for that which you were powerless to influence.

Genetic tests may also have more direct distributional consequences. Alleles occur in different frequencies in different ethnic groups; it would not be surprising to find that an allele associated with an epidemiologically significant disease such as coronary artery disease was more prevalent in some ethnic groups than in others. Alpha-1 antitrypsin deficiency, associated with lung disease, appears to be more common among people of Scandinavian ancestry. If the group in which the allele occurs more often was not historically a target of discrimination, we might not be particularly concerned. If, however, the allele was more common in a group that continues to suffer discrimination, such as sickle-cell trait in people of African descent, we would have good reason for concern. The mere fact that genetic predictors have the potential to affect differentially ethnic groups that experience discrimination does not uniquely distinguish them from other risk predictors. Hypertension, for example, is more prevalent among Americans of African heritage. But the immediate and direct tie between genetics and ethnicity may make genetic testing a more blatant use of a potentially explosive and discriminatory social classification scheme.

A third response to the claim that we need not worry about genetic risk testing because it is essentially similar to things like cholesterol testing is to question the premise that people know about the genetic component of cholesterol. Discussions of cholesterol in the media emphasize the things people can do to lower it. Reminders that cholesterol level is also significantly affected by genetics appear less frequently, and it may well be that most people are unaware that cholesterol level has a substantial genetic component. If people did understand that, perhaps they would be less tolerant of the widespread use of cholesterol testing to determine insurance eligibility, precisely because it was to that extent outside of individuals' control.

There is yet another possibility: that the central notion underlying commercial health insurance underwriting--the greater the likelihood of illness, the more one should pay for coverage--is morally unsound.

Actuarial Fairness

Insurers take a particular view of fairness: actuarial fairness. Actuarial fairness claims that "policyholders with the same expected risk of loss should be treated equally.... An insurance company has the responsibility to treat all its policyholders fairly by establishing premiums at a level consistent with the risk represented by each individual policy-holder."[5] This definition of fairness begs the question: Why should we count differences in risk of disease as an ethically relevant justification for treating people differently in their access to health insurance and health care?

Actuarial fairness does have a realm of application in which it seems reasonable. Call it the Lloyds of London model: if two oil tanker companies ask to have their cargoes and vessels insured, one for a trip up the Atlantic to a U.S. port, the other for a voyage through the Arabian Gulf during the height of the war in Kuwait and Iraq, the owner of the first ship would cry foul if she were charged the same extraordinarily high rate as the owner of the second. Most of us, I suspect, would agree that charging the two owners the same rate would be unfair. What makes it so?

For one thing, the two ships are exposed to vastly different risks, and it seems only fair to charge them accordingly. (The process of assessing risks is called underwriting.) Furthermore, the risks were assumed voluntarily. Third, the goal of both owners is profit, and it seems reasonable to ask them to bear the expense of voluntarily assumed risks. We could also ask how commercial insurance divides up the world. In this hypothetical it divides it into those who prefer prudent business ventures and those willing to take great risks. That does not seem to be an objectionable way to parse the world for the purpose of insuring oil tankers.

In practice, insurers do not behave as if actuarial fairness were an iron-clad moral rule. Valid predictors may not be used for a variety of reasons, typically having to do with other notions of fairness--for example, not discriminating on the basis of race, sex, class, or locale, even though these characteristics are related to the likelihood of insurance claims. Deborah Stone, who has studied insurance practices for HIV infection, dismisses the idea of actuarial fairness and argues instead that:

insurability is the set of policy decisions by insurers about whom to accept. It is not a trait, but a concept of membership.... Treated as a scientific fact about individuals, the notion of insurability disguises fundamentally political decisions about membership in a community of mutual responsibility.[6]

In a related vein, Norman Daniels points out that:

we allow many considerations, both of justice and of other goals of social policy, to override appeals to actuarial fairness, suggesting that we do not treat it as a basic requirement of distributive justice in insurance contexts after all.[7]

According to actuarial fairness--the Lloyds of London model--insurers would be justified in using genetic tests that predicted who was likely to file a claim. But is actuarial fairness an adequate criterion of fairness in health insurance?

Underwriting and the Social Purposes of Insurance

The threat genetic testing poses to the future of insurance for health-related risks--including health, life, and disability insurance--compels us to reexamine the social purposes served by insurance. Two points are obvious: first, the different types of insurance can have different purposes; and second, that the purpose of a particular form of insurance must be understood within its social context.

Life insurance, for example, is meant to provide financial security for one's dependents in the event that one dies. In the contemporary United States we must evaluate the role of such insurance in the context of a not particularly generous social welfare system that would otherwise leave the surviving dependents of a deceased breadwinner in very poor financial condition. The typical purchaser is an individual with one or more dependents who are unlikely to become financially independent in the immediate future. The benefits from life insurance are intended to tide survivors over until they can become financially self-sufficient, or live out their lives decently; they are not meant to provide windfalls to friends of the deceased. To the extent that life insurance is perceived as serving a need rather than being merely a commodity, we are likely to regard it as something that ought to be available to all. Our public policies toward life insurance suggest we view it otherwise, however. We prohibit certain actuarially valid distinctions such as ethnicity in setting life insurance rates. But we do not require that all persons, whatever their age, employment, or health, be permitted to buy life insurance at identical prices or at all. In consequence, the financial dependents of a person unable to obtain life insurance may suffer devastating changes in their life prospects if the principal earner dies.

Does the Lloyds of London model fit health care? Despite the current enthusiasm for tying voluntary behavior to health, most illness and disability is neither chosen nor in any sense "deserved," distinguishing it from the risks of shipping oil in a war zone. Neither is the goal of health care for those who seek it profit. Daniels argues that "justice requires that we protect fair equality of opportunity for individuals in a society." Reasonable access to health care in the contemporary United States is a necessary condition for fair equality of opportunity to pursue other goods that life affords. The social purpose of health insurance, understood in this way, is to provide access to the health care that people need to have a fair opportunity in life.

Lastly, how does underwriting in health insurance divide the world? It sets off the well from the ill and those likely to become ill. For insurers, the concept of actuarial fairness provides a rationale for charging much higher rates or declining to insure persons with a substantial possibility or illness or disability, reasoning that such persons should bear the costs associated with their particular risks. Persons at risk could find it difficult to obtain insurance at affordable rates, or at all.

Because most persons with insurance get it through their employer rather than through individual policies, the impact of such individual underwriting by commercial insurers would be diminished. The Office of Technology Assessment estimated that 14.5 million individual or family policies are written for people under age sixty-five.[8] This is a considerable number, to be sure, and people denied coverage from this source of insurance are greatly disadvantaged. But a far greater number of privately insured persons will be affected by what employers and group insurers do.

The pressure on employers to reduce costs may affect the insurance opportunities of employees. Although individual underwriting is still the exception rather than the rule in group policies, a recent OTA survey showed that 82 percent of commercial health insurers used some form of individual underwriting in their small group policies, and 48 percent did the same with late applicants in their large group policies. Blue Cross/Blue Shield plans used similar measures less frequently: 57 percent in small groups and 34 percent with late enrollees in large groups. It is reasonable to assume that such individual underwriting in group policies may continue to increase as a response to cost containment pressures. Another possibility is that self-insuring employers, who are largely exempt from state regulation, may use such testing to limit the employee to a policy that covers the employee alone, or the employee and spouse, but not an affected child.[9]

Implications for Policy

The era of predictive genetic testing coincides with a period of grave public concern about health care. In a recent survey, 82 percent of the public expressed such concern.[10] For the first time, health care is the number two political issue in the United States. Some unfamiliar numbers help account for this intense and growing worry: an estimated 22 million people are refused care in each year[11] and some 63 million people were without health insurance for at least part of the year--more than one in four Americans.[12] Many others are afraid to change jobs because of waiting periods, exclusions, or other penalties for changing insurers.

Much has been made of the alleged distaste Americans have for government involvement in health care. Yet a recent survey shows that 60 percent of the people want the government to take the primary role in health care.[13] There is little doubt that the current ragged system of private and public programs, with its many holes and frayed edges, must be changed. The conviction that health care ought to be available to those who need it seems to be widely shared. That conviction, together with a growing sense that the current patchwork is failing, may be strong enough to overcome the citizenry's hesitations about government inefficiency. Indeed, it seems likely that private health insurance would not have survived this long if not for government intervention. Tax subsidies for employer-sponsored health insurance programs amounted to $39.5 billion in 1991.[14] In addition, we provide direct government coverage for the health needs of people that commercial insurers want to avoid: Medicare, for those much more likely to need health care; and Medicaid, for some of those unable to pay for their own insurance.

Public programs such as Medicare and Medicaid tell us something important about our moral convictions on health care. They suggest that we are not content to allow the old and the poor simply to languish without access to care. Had we not passed such legislation, we well might have overturned or radically restructured the existing system of commercial health insurance decades ago.

There are good reasons to doubt that actuarial fairness is an adequate description of genuine fairness in health insurance. It may be a sufficient principle for commercial insurance against losses of ships at sea, but even a brief inquiry into the social purpose of health insurance suggests that apportioning by risks, as actuarial fairness dictates, fails to accomplish the primary social goals of health insurance. Genetic tests, like other predictors of the need for health care, are not good reasons for treating people differently with respect to access to health insurance.


1. Michael Walzer, Spheres of Justice: A Defense of Pluralism and Equality (New York: Basic Books, 1983).

2. H. Tristram Engelhardt, Jr., "Allocating Scarce Medical Resources and the Availability of Organ Transplantation," NEJM 311 (1984):66-71.

3. Robert M. Veatch, "Voluntary Risks to Health: The Ethical Issues," JAMA 243 (4 January 1980):50-55.

4. Norman Daniels, "Health Care Needs and Distributive Justice," in In Search of Equity, ed. Ronald Bayer, Arthur L. Caplan, and Norman Daniels (New York: Plenum, 1983).

5. Karen A. Clifford and R. P. Iuculano, "AIDS and Insurance: The Rationale for AIDS-Related Testing," Harvard Law Review 100 (1987):1806-24.

6. Deborah A. Stone, "AIDS and the Moral Economy of Insurance," American Prospect 1 (1990):62-73.

7. Norman Daniels, "Insurability and the HIV Epidemic: Ethical Issues in Underwriting," Milbank Quarterly 68, no. 4 (1990):497-525.

8. Daniels, "Insurability and the HIV Epidemic."

9. The Americans with Disabilities Act (ADA) prohibits non-job-related medical screening. The act is not explicit about risk-oriented genetic testing; nor does it require that employers offer any particular set of health benefits, or any health benefits at all. The Genome Project's Working Group on Ethical, Legal, and Social Issues requested that the U.S. Equal Employment Opportunity Commission revise its proposed interpretive rules for the ADA to make genetic discrimination more difficult; the EEOC refused all of the Working Group's suggestions.

10. Mark D. Smith, Drew E. Altman, Robert Leitman, et al., "Taking the Public's Pulse on Health System Reform," Health Affairs 11, no. 2 (1992):125-33.

11. Smith et al., "Taking the Public's Pulse."

12. Madison Powers, "Justice and the Market for Health Insurance," Kennedy Institute of Ethics Journal 1, no. 4 (1991): 307-23.

13. Smith et al., "Taking the Public's Pulse."

14. John K. Iglehart, "The American Health Care System: Private Insurance," NEJM 326 (1992):1715-20.
COPYRIGHT 1992 Hastings Center
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:includes related articles
Author:Murray, Thomas H.
Publication:The Hastings Center Report
Date:Nov 1, 1992
Previous Article:Insurance for the insurers: the use of genetic tests.
Next Article:Oregon's disability: principles or politics?

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters