In response to:
Will Clinton's Plan Be Fair? from the January 13, 1994 issue
To the Editors:
Professor Dworkin correctly observes [NYR, January 13] that a standard of care derived entirely from the rescue principle is an absurd one and that no sane individual would wish to bankrupt himself or his society in order to provide “heroic treatment of improbable value…” What he does not speak to, however, is the case of heroic (and costly) treatments of probable value. It is precisely in this area that the decisions of the “prudent individual” might very well be in conflict with those of the society as a whole. Long-term renal dialysis for otherwise healthy patients is a classic example of a procedure that might be termed heroic but effective, as is bone marrow transplantation for certain forms of leukemia. Whereas a utilitarian perspective might argue that it is not prudent to provide insurance for such care, as it can consume considerable resources that might otherwise be used to benefit larger groups of individuals, it is likely that most individuals would want such treatments available to them. It seems to me that in instances in which medical intervention can prevent an early death and restore more or less normal functioning, most people probably do make decisions on the basis of a rescue principle, i.e., a willingness to bear any cost to save life and/or health. I would assume that in these instances the “prudent insurance” principle would mandate coverage for such catastrophic expenditures, but I would be interested in Professor Dworkin’s views on the subject.
Professor Dworkin rightly acknowledges that it may not be sufficient to leave all decisions about a basic benefit package to the “prudent” decision-making of the average individual. That is it may be desirable for the government to mandate that certain kinds of coverage be included even if they would not be selected by the average consumer. He fails, however, to articulate criteria for deciding when to override individual decision-making. To take an example from my own area of expertise, most individuals tend to rank coverage for mental health care as a low priority, reasoning that such care is for “crazy” people, a category to which they do not belong. In fact, however, a surprisingly high percentage of the population requires such services at some point in their lifetime. Failure to provide coverage for such services may deter many from seeking treatment, even in times of crisis. Furthermore, there is a potential social cost to such a failure, as suicide, violence, drug addiction, child abuse, or other anti-social behavior may result in at least a small percentage of cases from the failure to receive treatment. It would therefore seem appropriate for the Health Board to mandate mental health coverage as part of the basic package. In general, however, on what basis would a Health Board make such determinations? Would the test be the possibility of harm to society at large (as with violent mental patients or carriers of infectious diseases), or would there be a kind of in loco parentis rule in which the government would protect the individual from the consequences for his or her own foolishness? To some degree debates such as these are similar to those over helmet laws for motorcyclists; most observers agree that wearing helmets represents a prudent safety precaution when riding a motorcycle; there is confusion over a clearcut rationale for mandating such behavior.
In his discussion of what kinds of coverage the prudent individual might select, Professor Dworkin seems to suggest a probability of success test as one of the criteria. While this would seem to be a reasonable suggestion, I fear it is unworkable in the real world. Success rates are based upon large groups of patients with widely differing characteristics. Thus, a procedure that has a “10 percent chance of success” based upon aggregate statistics in fact may have almost no likelihood of success in most patients, but a very good chance of a favorable outcome in a select few. To deny the few a treatment that may save their lives because it would be futile to prescribe it for most patients would seem unfair, but it is difficult to see how a universal package of coverage can be structured without either including or excluding procedures for all patients. In Canada, for example, the criteria for approving certain procedures tend to be global and objective (e.g., age) and not take into consideration the specifics of a given case. The alternative, some form of case management by insurers, has proved to be unworkable both fiscally and medically, as it has created an expensive and unwieldy bureaucracy and tends to pit the insurer against the patient.
My final question concerns the definition of the “prudent individual.” Professor Dworkin notes that the actual decisions of any particular individual would not necessarily be “prudent,” but he offers no guidelines for how to determine prudence. Does one assume that the actual decisions of a large number of individuals shall define prudence? Or does some panel of experts determine what should be prudent, as was done in Oregon? A further problem is that the decisions of any given individual change over the life course. Professor Dworkin’s repeated references to a mythical twenty-five-year-old individual suggest that he views prospective choices as more relevant than those made retrospectively. I question this assumption. A healthy twenty-five-year-old might determine that the quality of his life for the next forty years takes precedence over prolonging it after age sixty-five years. He might therefore opt not to have coverage for more expensive treatments for ailments common to older individuals. The same individual at sixty-five, however, faced with the prospect of a fatal heart attack, might willingly have given up expensive vacations or other luxuries in order to be able to have coronary bypass surgery that might prevent an early death. In some sense, this is analogous to the “prudent” decision of a twenty-five-year-old to smoke cigarettes on the grounds that the improvement in his quality of life outweighs for him the risk of developing lung cancer or other chronic respiratory ailment later in life. At sixty-five, the same individual may deeply regret his earlier choice. Interestingly, it is some variant of the rescue principle that dictates that the decision not to smoke (or to purchase coverage for cardiac surgery) is more prudent on the grounds that health is a greater value than simple pleasure.
One point that Professor Dworkin makes (albeit in a footnote) that deserves underscoring is the importance of not discouraging health care outside of the system as is done in some systems. An approach such as the “prudent insurance” one dictates that decisions about a basic health care package derive their legitimacy from a recognition that they represent average sentiment and that individuals may be expected to vary in their needs and priorities. The system needs to retain the freedom for those individuals with greater need to purchase coverage to meet these needs. Currently, many health maintenance organizations provide powerful disincentives (including in some cases contractual prohibitions) to the provision of so-called “off-plan” services. This practice is inconsistent with the model of health care envisioned here.
Professor Dworkin has provided a valuable service by elevating the level of debate about health care and the Clinton Plan and offering a framework within which specific proposals can be evaluated. The points raised in this letter reflect concerns about specific problem areas or aspects of implementation rather than any disagreement with the underlying philosophical approach.
Bruce L. Smith, Ph.D.
Assistant Clinical Professor of Medical Psychology
University of California
To the Editors:
It is very well to consider what medical expenditures a prudent individual would insure him or herself for, but when faced with medical need it is not as easy as Professor Dworkin appears to think to discern where justice lies in making or, what is more to the point, requesting or approving the actual expenditures. He seems to feel that a certain likelihood of benefit should be the standard that determines the willingness to provide insurance coverage for medical modalities. I would suggest, however, that the 25 percent threshold for a prudent investment in the payment of a premium gets quickly revised downward when actual patients calculate whether they could benefit from a procedure. Based on my experience with patients, it is just this gap between what people are willing to pay for when they are well and what they wish done when they are sick that accounts for much of the rapid growth in health care costs. The problem as I see it is not that doctors adhere to the rescue principle but that patients do. I do not see how the assumptions Professor Dworkin asks his readers to make so as to allow them to decide in an unbiased way what they would wish to pay for will make patients sensitive to cost at the moment of medical decision. As a thirty-nine-year-old said to me just a couple of days ago when faced with a choice between therapies whose likelihood of success were none versus slim, “I’ll do anything to live.”
Interestingly enough, many of the elderly say the same thing. Furthermore, it offends their notions of equity to have paid their Social Security and other taxes for forty or more years and, in the case of most of my patients, served their country honorably, often in time of war, only to be told that they must (or, at least, ought to) forgo some therapies in favor of those who have longer to live. In fact, neither they nor I am affected much by knowing that a major (emphasis mine) fraction of medical expense is expended during the last six months of life. What else would one expect in the case of Medicare, a program specifically designed to pay for medical care for the elderly? Just within the last few weeks, I have had two men over the age of sixty-five elect to have kidney dialysis treatments despite life expectancies due to underlying pulmonary and cardiac disease that can only be measured in months. I would rather these men had decided for themselves, when they were young and vigorous, which classes of treatments they would have been willing to forgo when they became debilitated by age and chronic disease. The prudent, as defined by Professor Dworkin, would draw up such advanced directives and, I like to think, would often eschew expensive treatments whose major effect is to prolong the process of dying.
I wonder where Professor Dworkin has gotten the idea that health care is rationed on the basis of money. In fact, in most localities with which I am familiar, the truly poor, i.e., the underclass, have better access to health care through various forms of governmentally subsidized care than those who work at jobs that provide no or inadequate health insurance. Furthermore, it is inefficiencies and misuse (by inappropriate hospital emergency room visits, for instance) of these subsidized systems such as Medicaid that accounts for a significant amount of waste in health care expenditures. This can only be improved by increasing the numbers of primary care physicians willing to serve such populations and elevating the general level of medical sophistication found there. However, both of these goals would require more expenditures if anything, at least up front, to save money, perhaps, in the future.
Finally, at the risk of seeming self-serving, I might say that I find it interesting that it is health care that faces so much criticism regarding what it costs rather than, for instance, the legal, entertainment or professional sports industries. I dare say that the rate of increase in the amounts spent by Americans in these areas rivals that of health care and also far outstrips the general rate of inflation. Would the prudent elect to spend the $100 or more that is required for a family to attend a professional football game or would they choose to put those resources into health care, education or, even, invest them? I wonder.
Jack G. Kleinman, M.D.
Professor of Medicine
Medical College of Wisconsin
Ronald Dworkin replies:
I am grateful to the many readers who raised objections and questions about my article on fairness in health care, and though I shall concentrate on the points that Drs. Smith and Kleinman make, I shall also reply to some made in other letters. I argued for a “prudent insurer” test, which requires government to guarantee only the treatment that most prudent people of average means would insure to provide for themselves, early in their lives, if they had state-of-the-art medical knowledge, and if neither they nor insurance companies could predict that they were either more or less liable than others to contract any particular disease or suffer any particular accident.
As I said, officials applying that prudent insurer test would have to answer a variety of difficult and to some degree speculative questions. Dr. Smith is rightly worried about the most difficult of these. Suppose it is true, as he suggests, that the average twenty-five-year-old would not in fact buy insurance providing for mental health treatment because he would not count himself as among those likely to go crazy. On what basis could officials nevertheless decide that it would be “prudent” for him to buy such insurance and, in effect, buy it for him?
I suggested the following answer: these officials would have to put themselves in his shoes, to think more clearly, on his behalf, than he would himself. So they would have to ask whether he would have insured against mental illness if he had reflected on his own best interests, if he was aware of the “surprisingly high percentage” of people who develop such illness, and if he had the clear-headed imagination needed to contemplate his situation if he himself did. Political officials often have to make such decisions on behalf of those they represent: they have to decide, for example, whether it is prudent for most people, given their long-term preferences, to wear seat-belts, which is different from asking whether most people would actually wear them if given the choice.
As I said, since people differ, even after reflection, in what they take their best interests to be, insurance decisions that would be prudent for one young adult would not necessarily be prudent for another. It would be prudent for almost all young people, I think, if they had average means to buy mental health insurance. This is not necessarily true of everyone, however. A few people might believe passionately in the wisdom of living for the moment; they might think, even after deliberation, that it would be degrading to buy long-term insurance at the expense of such present pleasures as (in Dr. Kleinman’s example) football tickets. The prudent insurer test, applied just to them, would not recommend buying insurance on their behalf.
Some readers feared that such diversity would undermine the prudent insurer approach—that it could work only if everyone (or nearly everyone) would choose the same hypothetical insurance policy. But that is a misunderstanding. I do think that the great bulk of prudent people would buy at least most of the coverage that the Clinton plan now explicitly provides, including, for example, routine hospitalization and routine surgery. I also think, as I argued, that the great bulk would reject insurance that would provide expensive medical procedures for patients in a persistent vegetative state, or expensive treatment that could prolong their lives by only a few months. But people would have sharply different opinions about other coverage. Perhaps nearly as many people would insure to provide costly neo-natal treatment for extremely premature babies as would not, for example.
In such closer cases, a further and more difficult judgment about fairness would be necessary to decide whether the benefit in question should be provided universally. In my own view, fairness would not require universal coverage unless a substantial majority of prudent people would insure to provide it. If only a bare majority would do so, it would seem unfair to provide it for everyone, making almost half the population buy insurance that they did not want. (Though in such cases the additional coverage should of course be available as supplementary insurance at an additional premium.) I recognize, however, that other people might make this further judgment of fairness differently. They might think that everyone should be entitled and required to have what even a bare majority would buy for itself if it could. They and I would agree that the prudent insurer test provided the best structure for deciding what fairness required in health care, but we would disagree in detail about how that test should be applied.
Dr. Smith offers an important argument of a different kind, however, for making at least some kinds of coverage universal even when a large number of people would not buy it for themselves. If we permitted people not to have any health insurance, and an uninsured individual developed serious mental disease, the community would have to provide treatment anyway, out of concern for others who might be affected by his conduct, as well as for him. Since in these and other similar cases we would in fact provide certain forms of care for everyone, whether they had insurance or not, it seems fair that we require them to pay, directly or through taxes, to insure for that care.
One of Dr. Smith’s other questions is particularly valuable. He wonders why I am so concerned with the choices that young adults would make prospectively rather than those that people would make later in life, particularly after they have fallen ill when, as he and Dr. Kleinman report, they want any treatment that might conceivably save or extend their lives, however remote the chance. The most basic purpose of the prudent insurer test, however, is to force people to decide which risks it is better to run—the risk of less than maximal provision for health care later in life or the risk of inadequate investment in other resources, like jobs and education, earlier in life, when such resources make a great difference to the whole life one can lead. That choice can only sensibly be made from the perspective of youth, when both these risks are real ones. Of course, someone who chose to invest more in education when he was young may very well want, when old, treatment that he cannot then afford. But it does not follow that he would or should regard his earlier decision as a mistake. Dr. Kleinman, I think, agrees: he says that though older patients appeal to the rescue principle when they are sick, they would not have thought it prudent to follow that principle in deciding, when “young and vigorous” how much to set aside to guarantee medical treatment later.
It is not true, however, that the prudent insurer test would always recommend more limited coverage than other approaches would. As Dr. Smith notices, a strict cost-benefit approach, in a utilitarian spirit that allows terrible losses to one person to be canceled out by small gains to many more, might well recommend against very expensive but very effective life-saving procedures, like long-term renal dialysis or bone marrow transplants in certain cases. A substantial majority of Americans, on reflection, might nevertheless think it prudent to pay the premiums that would provide such care for themselves in circumstances in which its value was clear. People who made that decision would not be endorsing the rescue principle, as Dr. Smith suggests they might be—they would not think it sensible, for example, to provide for speculative procedures with slim chances of success, which the rescue principle would recommend. They would be endorsing the different and more attractive view that added life has an importance to people, one by one, that it does not have in familiar versions of aggregate cost-benefit calculations.
I am uncertain how far a workable universal plan could take into account another important fact Dr. Smith mentions: that the probability of success of particular treatments is much greater in some cases than in others. An optimal insurance policy would reflect this fact: it would allow individuals to insure for treatment normally not covered if, in the opinion of specialists, that treatment had a substantial chance of significantly improving or extending life, and it would charge, for such qualified insurance, a much lower premium than would be necessary for guaranteeing the treatment even when its benefit would be more doubtful. As Dr. Smith remarks, however, such provisions would be expensive to administer and would provoke agonized conflict between patient and insurance company in particular cases. Perhaps it would be feasible, nevertheless, for a national health policy to allow appeals to a panel of doctors for coverage of treatment normally denied as of doubtful value, in which the burden of proof would be on a patient’s doctor to show that the prospects of success in his case were much higher than average.
Dr. Kleinman questions my suggestion that health care is now rationed by money in America. It is true that those who qualify for Medicaid receive better coverage than many of those who do not. But those in the latter group are mostly poor or unemployed or unlucky: in any case they lack the care they would buy for themselves if they were better off, which is, I take it, what we mean by the claim that we now ration medical care by wealth. Professor Thomas Scanlon, however, has pointed out to me that my use of the term “rationing” was inaccurate. Rationing in the strict sense—in the sense in which food and fuel were rationed in World War II—is a matter of preventing people from buying what they can afford, not a matter of not giving them what they cannot. The point is a substantial one because saying that a national health plan must “ration” health care suggests that it must deny people what they otherwise have a right to have. It is better to say, not that government must ration health care in some manner or other, but that it must provide everyone with all but only the health care that fairness demands.
Some readers misunderstood an even more basic point that I should now try to clarify. One wrote, for example, that the prudent insurer approach would give the wrong result in some cases because though prudent people might well buy insurance covering expensive and speculative transplants, since such coverage would add only a small extra premium to each policy, the government should not guarantee such coverage because the overall cost of covering all such transplants would be too high. But if it were true that covering all possible transplants would add only a small amount to each premium, and that prudent individuals would therefore buy such coverage, then it would follow that the overall cost was not too high. It would follow if my basic assumption is sound: that we ought to spend collectively what it would be prudent for each to spend individually if wealth were fairly divided.
Other readers worried that asking officials to decide what people would do if they were prudent would tempt these officials to impose their own moral or ethical values on others. Deciding what prudent people would want is, as I said, a familiar responsibility of government in many other areas. But that does not make the danger less serious, and that is why I recommended diversity among the officials making the decisions, and a series of informed public discussions to improve their sense of what people really think is in their own best interests. It is also one reason why I recommended that people be free to purchase additional insurance. If a great number did so, it might suggest that the officials had made an important mistake.
A number of letters suggested that my approach to the question of fairness in health care is an application of John Rawls’s more general theory of justice, and some criticized me for varying the features of his method. In fact, there are important differences between the two approaches that should be noticed. Rawls imagines a social contract among people each of whom is ignorant of his or her own resources, talents, status, preferences, and values. My argument, on the contrary, focuses on the decisions that real people would make, for themselves one by one, if they had full and reflective knowledge of their actual resources, age, position, preferences, and values. It aims to elicit very concrete information about people’s preferences so that government can make decisions that are rational in the light of these preferences. It is a mistake to think that every philosophical argument that relies on hypothetical decisions is the same kind of argument.
I should like to take this opportunity, finally, to comment on two much-publicized New Republic articles by Elizabeth McCaughey, a fellow of the Manhattan Institute, which sharply criticized the Clinton plan. McCaughey has made three main charges: that under the plan people will find it “difficult” to consult doctors of their choice, that they will not be permitted to pay doctors extra for better care if they wish, and that the National Health Board will dictate to them and to their doctors what treatments are “necessary and appropriate.”
This is a misleading account of features of the Health Security Act that I described in my article. The act requires all states (other than those that have elected single-payer plans, as few if any are likely to do) to offer fee-for-services insurance policies that allow patients to choose their own doctors. Such policies will be more expensive than HMO plans that do not allow such choice, of course, but since a growing number of employers are requiring employees to use HMO’s now, more people might be in a position to choose their own doctors if the act were passed than now are.
McCaughey points, however, to the provision I mentioned that allows health care alliances to reject any fee-for-services plan that charges more than 120 percent of the average premium in the alliance; she says that that provision will mean few such plans. She provides no evidence for her claim. But if compelling evidence were available, or experience proved her right, this provision would presumably be changed, because the administration has made plain that fee-for-services policies are an essential part of its program.
It is true that the act forbids doctors to accept higher payments than the doctors have agreed with the local health alliance to accept. But that applies only to services that are covered by the universal benefits package; and the disadvantages of allowing doctors to solicit higher fees from some patients for services that are meant to be available at a lower cost to everyone should be obvious. The act leaves people free, as I emphasized, to buy supplemental insurance to pay for procedures not included in the basic benefits package. McCaughey, in replying to the many commentators who corrected her, suggested that this feature was unimportant, because it would leave people free to buy “cosmetic surgery, psychotherapy or other uncovered services.” But these “other uncovered services” would include exactly the more comprehensive regular medical examinations and more expensive diagnostic procedures that people now pay higher premiums or special charges to secure. They would also include exactly the treatment McCaughey says she fears a national health board would exclude from the basic coverage as not necessary or appropriate.
Her arguments are, in that way, inconsistent. She complains that Clinton’s plan would prevent people from spending extra to buy special medical care, and then complains that a group of doctors would have the power to decide what special medical care they must spend extra to have. She does argue that since supplemental insurance would have to be offered to everyone, regardless of medical history, only the seriously ill would seek it and insurers would not offer it. But whether insurers did offer any particular supplemental coverage, and at what premium, would depend on how many people beyond those with bad medical histories would want it. If many Americans would think a particular form of extra coverage prudent, in spite of the fact that it was not provided universally, the “seriously ill” would not form a sufficiently large part of the market so that adverse selection would make providing such insurance unprofitable.
But such coverage would be more expensive, it is true, than it would be if insurers were free to reject those who needed the insurance most. Clinton’s plan is, after all, a plan for making the delivery of health care in America fairer than it now is, and that would undoubtedly mean that some people would pay more in insurance premiums for the highest-quality care than they do now. McCaughey and those who have hailed her arguments show little understanding of that goal. If they succeed in blocking any significant reform, many of the people they frightened into opposing it will one day find that they are among the “seriously ill” or the unemployed or the millions of others for whom the high-tech health care she praises is only a sick joke.