Indiana Law Review Law Review Volume 35 2002 Number 3 BOOK REVIEWS Moral Principles and Legal Practice Leslie Pickering Francis* Review of David Orentlicher," MATTERS OF LIFE AND DEATH, Princeton, Princeton University Press (2001). In Matters ofLife andDeath, 1 David Orentlicher pursues a glorious subject: a moral theory for the translation of moral principles into moral practice. Such a theory would contribute to a central debate in applied ethics about the plausibility and structure of what is called "principlism." 2 Not surprisingly, principlists contend that moral principles play a major role in reasoning to conclusions about what to do in the messy circumstances of actual situations, such as whether to remove a feeding tube from a patient in the late stages of Alzheimer's disease when there is no clear indication of her prior preferences. A defense of translation principles would help to show how to bridge the apparent gaps between theory and practice which are the basis for central criticisms ofprinciplism. Orentlicher's subjectwould also contribute to growing interest in what is called "partial compliance theory," 3 the idea that additional or different moral principles may be called upon in deciding what it is right to do under circumstances of injustice. Principles for the just distribution of health care, for example, might be different in a world in which racial discrimination is rampant, than in a world of racial justice, or so defenders of affirmative action contend. 4 In much of the book, however, Orentlicher's subject is far more familiar: * Alfred C. Emery Professor of Law, University of Utah College of Law; Professor of Philosophy, University of Utah. B.A., 1967, Wellesley College; Ph.D., 1974, University of Michigan; J.D., 1981, University of Utah College of Law. * * Samuel R. Rosen Professor of Law and Co-director ofthe Center for Law and Health, Indiana University School ofLaw—Indianapolis. A.B., 1977, Brandeis University; M.D., 1981, Harvard University Medical School; J.D., 1986, Harvard University Law School. 1 . David Orentlicher, Matters of Life and Death (2001 ). 2. See, e.g., Margaret Olivia Little, On Knowing the "Why": Particularism and Moral Theory, 3240 HASTINGS Center Rep., July-Aug. 2001, at 32-40. The best known exemplar of principlism in bioethics is Tom L. Beauchamp& James F. Childress, Principles of Biomedical Ethics (5th ed. 2001). 3. The term is from John Rawls, Theory of Justice 212-13, 215-18 (rev. ed. 1999). 4. E.g. , Leslie Pickering Francis, Affirmative Action and the Allocation ofHealth Care, 66 Mt. Sinai J. Med. 241 (1999). 990 INDIANA LAW REVIEW [Vol. 35:989 how moral principles should, or should not, be reflected in the law. This debate is far older, going to the roots ofthe conflict between the positivist's conception separation of law and morality and the natural law theorist's defense of some variety ofnecessary connection between the two. A common positivist view has been that law consists of a set of rules, identifiable by their "pedigree," that is, by how they were adopted. 5 In this way, principles of law can be distinguished from principles of morality. Orentlicher develops arguments for the law to employ bright-line rules that differ from the recommendations of moral theory. Although Orentlicher does not characterize his contribution in this way, his achievement might be regarded as a moral defense of positivism, through an account of why the law might justifiably employ distinctions that seem indefensible from the perspective of moral theory. 6 This Review first develops Orentlicher' s argument in some detail, showing how its principal achievements concern the translation ofmoral theory into legal practice in some highly controversial policy areas. In this respect, Orentlicher' s book draws original links between philosophy of law and applied ethics, links that are woefully under-explored by scholars. The book is less successful, however, as a general theory for translating principle into practice, partially because of its case-based mode of argument. The final section of the review argues that some ofOrentlicher' s claims are better understood within the domain ofpartial compliance theory-that is, as strategies for translating moral principles into practice in an unjust world. I. Three Models of Translation Moral principles are famously abstract. "Do unto others as you would have others do unto you." "Thou shalt not kill." "Respect autonomy." 7 "Actions are right when they will produce more happiness on the whole than available alternatives." 8 What do such principles portend for daily dilemmas in medicine, such as whether to inform patients about very small risks of morbidity from prescribed drugs? What do they recommend for more difficult decisions, such as whether to allocate organs for transplantation based on medical need or 5. See, e.g., Hart's POSTSCRIPT (Jules L. Coleman ed., 2001); H.L.A. HART, TheCONCEPT of Law (2d ed. 1994); Jules L. Coleman & Jeffrie G. Murphy, The Philosophy of Law: An Introduction to Jurisprudence (rev. ed. 1990). 6. In a recent monograph, Larry Alexander and Emily Sherwin draw linkages between the debate about the imperfections of rules and the conflict between legal positivists and natural law theorists. Legal positivists focus on what is needed for rules to provide authoritative guidance; natural law theorists look to the effort of law to achieve moral goals. Larry Alexander& Emily Sherwin, The Rule of Rules 184 (2001). 7. Respect autonomy is a principle applied in many fields. It is a fundamental principle in bioethics and comprises a number of medical codes. See, e.g., WORLD Medical Ass'n, Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects (1964); American Medical Ass'n, Principles of Medical Ethics (2001). 8. See RAWLS, supra note 3. 2002] BOOK REVIEW 991 potential years of life saved? At the very least, abstract moral theory owes ordinary moral actors an account ofwhy it is sometimes hard to apply principles to practice. One popular account is that difficulties in applying principles to practice lie in our ignorance about the facts. Fuller knowledge of the circumstances of a particular situation, for example, should demonstrate what action will produce the most good on the whole. Orentlicher's thesis is, instead, that at least part ofthe problem of translation is moral. Practice does not always reflect what would be expected from straightforward application of theory, for important moral reasons. In the book, Orentlicher develops three methods for translating principle into practice, each in the context of a particularly thorny issue in bioethics. An advantage of such context-specific development is that it contributes to the debate over each ofthe issues Orentlicher treats; a disadvantage is the failure to develop systematic themes that recur across a range of problems of translation. The first translation method rejects individualized decisions in favor ofgenerally valid rules. Orentlicher argues persuasively that the line between withdrawing or withholding care on the one hand, and active assistance in dying on the other, may be more sensible than it seems to some moral theorists. The second method recommends avoiding perverse incentives in translating principles into practice. Otherwise justified moral rules may create problematic incentives in practice, and so should not be followed. For this method, Orentlicher's example is the imposition ofmedical treatment on a pregnant woman for the benefit ofthe fetus she carries. While such compulsion might appear clearly warranted in the circumstances ofa particular case, the knowledge that such practices occur risks poorer overall outcomes for fetuses, because pregnant women avoid medical care. The third method recommends disguising the use ofprinciple when "tragic choices" are at stake. When faced with deep moral conflicts, society may be justified in avoiding clear statements of principle. Here the example is that life- sustaining treatment may be denied as "futile," an apparent judgment that it would not work, in lieu of a more public decision to ration scarce medical resources. The remainder ofthis section develops each ofthese methods in more detail, showing how they both do, and do not, connect into a more general theory oftranslation. The discussion also explains how they are more plausibly viewed as strategies for defending legal practice against the charge that it is morally unprincipled, than for translating moral theory into moral practice. A. Translation by Means ofGenerally Valid Rules Orentlicher's first method for translating principle into practice is the use of generally valid rules rather than case-by-casejudgment. There are many reasons for relying on rules, either in morality or in law: lack of the time required for careful assessment, risk of bias, or the need to correct for inadequate information. 9 In law, public promulgation and protection of reliance interests 9. see, e.g. , frederick schauer, playing bythe rules: aphilosophical examination of Rule-Based Decision-Making in Law and in Life 135-66 (1991). 992 INDIANA LAW REVIEW [Vol. 35:989 lend further support to the use ofrules. ,0 Despite these advantages, rules may not achieve their underlying moral purpose when they are applied to particular cases. When the general moral goal is protecting autonomy, for example, a rule that physicians should respect the informed choices of competent patients may be both over- and under-inclusive. A rule is over-inclusive when it has "false positives," cases in which applying the rule to respect individual decisions would not in fact further autonomy, perhaps because there has been a failure to notice problems with competence. A rule is under-inclusive when it has "false negatives," cases in which intervention would be permitted, but a ban on intervention would further autonomy. For some rules, difficult moral choices attend whether false positives or false negatives are the more serious risks to be avoided. A system that imposes stringent due process requirements on capital convictions, for example, represents the judgment that a false positive (undeserved execution) is worse than a false negative (undeserved leniency). The imposition of more stringent standards of competence on respect for patients' choices represents the judgment that a false negative (respecting an incompetent patient's risky decision) is worse than a false positive (failing to recognize the risky decision ofa competent patient). ' ' With respect to end-of-life decision making, a commitment to life recommends that false positives (allowing unjustified deaths) are worse than false negatives (not permitting justified deaths). In contrast, a commitment to allowing individuals to control how they die and the memories which survive them suggests that false negatives are the more serious risks. This last difference lies at the heart of the disagreement between Chief Justice Rehnquist's insistence that states may require clear and convincing evidence ofthe patient's prior wishes before care may be withdrawn from an incompetent patient, and Justice Brennan's conclusion that the stringent evidentiary standard impermissibly burdens the patient's right to die with dignity. 12 When rules are over- or under-inclusive, case-by-casejudgments may be the better strategy for achieving underlying moral goals. But this move raises difficulties of its own. Suppose, with respect to end-of-life decisions, that the moral goal is to prevent "unjustified" but permit "justified" deaths. Case-by-case efforts to decide whether death is justified may misfire, just as may the application ofrules. Moreover, such individualized inquiries may bring divisive social scrutiny into the private lives ofindividual patients. Orentlicher's strategy is to bypass much of the disagreement about when a death is "justified," by arguing that two principal accounts ofjustified death each independently support reliance on a general rule, the rule that consensual withholdings or withdrawals of care are permissible, but consensual aid-in-dying is not, despite the fact that many commentators have argued convincing that there is no defensible 10. Mat 139-41. 1 1 . This view is taken by Allen E. Buchanan & Dan W. Brock, Deciding for Others: The Ethics of Surrogate Decision Making (1989). 12. Cruzan v. Mo. Dep't of Health, 497 U.S. 261 (1990). 2002] BOOK REVIEW 993 distinction in moral theory between the two. 13 Orentlicher's arguments here are complex and worthy of careful attention. For he does not employ the standard objections ofslippery slopes (where is American health care, anyway, on slopes tilting towards or away from problematic forms ofeuthanasia? 14 ) or exploitation ofthe vulnerable (aren't those who are not allowed to die as vulnerable as those who die prematurely as a result of inadequate health care?). Instead, Orentlicher argues that on two principal accounts ofjustified death-the autonomy view and the no-real- life- left view-case-by-case decision making is morally flawed. In the end, however, Orentlicher's argument is more persuasive as a defense of the law's use of a bright-line rule distinguishing withholding or withdrawing care from aid-in-dying, than of reliance on such a rule in moral practice. Orentlicher first defends the general rule in light of the autonomy view of justified death. On this view, deaths are justified when, and only when, they are genuinely chosen by competent patients. Any given instance of withholding or withdrawing care, or assisted death, is only justified if genuinely chosen by a competent patient. False positives would result from judging wrongly that a decision was genuine, perhaps from failure to identify coercion, depression, or problems with competence. Bayesian analysis holds that the ratio offalse to true positives is likely to be high when the background probability of true positives is low, and low when the background probability oftrue positives is high. Thus if the background probability of a suicide's being genuinely chosen is low, the probability that a positive is a false positive would be higher than if the background probability of a suicide's being genuinely chosen is high. Orentlicher believes that these differences in background probabilities are indeed the case: withholdings or withdrawals of care are likely to be genuinely chosen, suicides much less so. Defenders of autonomy, he concludes, would be well advised to adopt the general rule that withholdings and withdrawals should be permitted, but aid-in-dying should not be. This argument is elegant, but deceptively so, since Orentlicher is not entirely clear about the relevant contrast classes. If his argument contrasts all withholdings or withdrawals of care from all cases of physician aid-in-dying, then all cases of each are the relevant contrast classes. If his argument contrasts all withholdings or withdrawals ofcare for patients who are irreversibly ill from all cases of physician aid-in-dying for patients who are irreversibly ill, these are the relevant contrast classes. In asserting that withholdings or withdrawals are likely to be genuinely chosen, but suicides are not, however, Orentlicher contrasts withholdings oflife-sustaining treatment (a category which builds in the fact that the patient is at least ill, in virtue of the need for life-sustaining treatment) with all suicides, including those ofthe depressed young. 15 But if the relevant contrast classes are held constant, as Orentlicher to some extent 13. Perhaps the seminal article in this vein is James Rachels, Active and Passive Euthanasia, 292 NEW ENG. J. MED. 78 (1975). 14. See, e.g., Wibran Van den burg, The Slippery Slope Argument, 102 ETHICS 42 (1991). 1 5 . Orentlicher, supra note 1 , at 64. 994 INDIANA LAW REVIEW [Vol. 35:989 recognizes, 16 the false positive rates may not be so different. Indeed, some commentators have argued that, in the United States, the false positive rate may be as high, higher, or ignored altogether for withholdings or withdrawals from ill patients, as it is for assisted suicides. 17 In the end, Orentlicher recognizes that if terminally ill patients seeking to have care withheld or withdrawn are compared to terminally ill patients seeking assisted deaths, the false positive rates may be comparable. Thus he concludes that a limited exception to the legal prohibition ofaid-in-dying for the competent, terminally ill, as in Oregon, may be justified. 18 But to apply such an exception more generally raises the second moral reason for favoring a bright- line distinction between withholding or withdrawing care and aid-in-dying. Orentlicher foregrounds this reason from the perspective ofanother, commonly- accepted account ofjustified death: that death is justified when the remaining quality of life is limited or low. When the moral goal is to permit these, but only these deaths, case-by-case decisions will rest on judgments about the quality of life for individual patients. If these are to be legal judgments, they will require assessment ofthe individual's quality oflife by an outside entity, such as a court, an ethics committee, or some other regulatory body. However, on what might be called the Mill principle, 19 that the individual is the best judge of his/her own interests, such outside judgments should be resisted, Orentlicher contends. 20 Outside entities are in less favorable epistemological positions than individuals themselves; moreover, for outside entities to try to gather sufficient information about individual cases would significantly intrude on privacy. Such intrusive assessments could be avoided if the law were to adopt a bright-line rule. Once again, Orentlicher contends, the line should be drawn between withholdings or withdrawals and aid-in-dying, because ofthe likelihood that the former, but not the latter, reflect accurate judgments about the quality of life left. But this argument is problematic for some of the same reasons as the first argument. When the contrast classes are held constant, so that withholdings or withdrawals and aid-in-dying are compared for similar groups of patients, it is not clear that the former group will contain a higher percentage of accurate quality-of-life-leftjudgments than the latter. Indeed, disability advocates contend thatjudgments about quality oflife are notoriously erroneous in both cases. 21 To be sure that decisions are made which reflect the moral goal of permitting only justified deaths will require more particularized inquiry into individual circumstances. Indeed, if the Mill premise is correct, it will be ill-advised for outside authorities to engage in this scrutiny. But this argument only supports the 16. Id. 1 7. E.g., Margaret P. Battin, Euthanasia: The Way We Do It, the Way They Do It, 6 J. PAIN & Symptom Management (1991). 18. Death With Dignity Act, Or. Rev. Stat. §§ 127.800-127.897 (1999). 1 9. From John Stuart Mill, On Liberty (David Spitz ed., Norton 1 975) ( 1 859). 20. Orentlicher, supra note 1, at 67. 21. see, e.g., anita silvers et al., disability, difference, discrimination: Perspectives on Justice in Bioethics and Public Policy 87 (1998). 2002] BOOK REVIEW 995 use of a general translation principle to avoid bringing in the law, not to ensure more accurate moral decisions. Orentlicher correctly points to moral objections to making quality of life judgments for others, but he has not shown why such judgments are required when moral principle is translated into moral, rather than legal, practice. B. When Moral Principles Go Wrong: The Problem ofPerverse Incentives Orentlicher' s second model for applying principle to practice is the avoidance of perverse incentives. Moral principles should not be put into play in ways that are self-defeating or that threaten to undermine other moral values. Distinguishing between translation outcomes into moral and into legal practice is central to the evaluation of this model. When assessing a translation into moral practice, the concern is whether the principle (or other moral goals) would be undermined by its direct application. There is extensive literature, for example, on whether applying the principle of utility in every day moral life would encourage people to behave in ways that do not promote the overall good. 22 When assessing translation into legal practice, the coercive and distributive effects of the law come to the fore. Thus it is more difficult to defend the claim that pregnant women have legal obligations to undergo medical care for the benefit ofthe fetuses they carry, than the claim that they have moral obligations to accept such care. 23 In discussing perverse incentives, Orentlicher makes clear that his goal is translating moral principle into legal, rather than moral, practice, Orentlicher develops two approaches to the dilemma. 24 One is to hold that pregnant women have limited legal obligations to accept medical treatment that would benefit both them and their fetuses. The other approach accepts, regretfully, that pregnant women have no such legal obligations to accept medical intervention, because the requirement may create incentives for pregnant women fearful of coercion to avoid prenatal care altogether. Orentlicher concludes that which alternative is preferable is an unresolved question. 25 His argument, however, attends less fully than it might to some ofthe general moral concerns about translation that inform his defense of the use of generally valid rules. The perverse incentives concern is that a general practice of forcible intervention for the benefit ofthe fetus will deter pregnant women from seeking medical care. At least for the limited situation in which the care is beneficial to both the pregnant woman and the fetus, Orentlicher thinks this risk is real, but manageable. The risk is magnified when decisions need to be made quickly, by 22. A central work in this vein is D.H. Hodgson, Consequences of Utilitarianism: A StudyinNormativeEthics andLegalTheory (1 967). See also R.M. Hare, MoralThinking: Its Levels, Method, and Point (1981). 23. See, e.g., Bonnie Steinbock, Maternal-Fetal Conflict and In Utero Fetal Therapy, 57 Alb. L. Rev. 781,791(1994). 24. Orentlicher, supra note 1 , at 89-90. 25. Id. at 90, 119. 996 INDIANA LAW REVIEW [Vol. 35:989 judges who lack information about the circumstances of a particular case or knowledge about medicine generally. 26 Orentlicher believes these difficulties can be mitigated by the availability of appellate review, which can set out general standards for use in such cases; a clear and convincing evidence standard for intervention, for example, would help to guard against any bias of the medical profession in favor of intervention. 27 He does not, however, explore the possibility that inequalities in access to legal services will skew the protections of appellate review. 28 Nor, with the exception of litigation involving Jehovah's Witnesses, does he explore the possibility that the perverse incentives concern will be more powerful for some groups-racial and ethnic minorities particularly-than for others. A further noticeable gap in Orentlicher's discussion of perverse incentives is his failure to explore why there is such a lack of evidence, on one side or the other, about perverse incentives claims. There are many other examples of possible, but unstudied perverse incentives hypotheses. Psychiatrists are concerned that duties to warn will deter patients from seeking care in an honest fashion. 29 Infectious disease specialists worry that duties to report positive HIV tests will discourage testing and reporting that are invaluable from a public health perspective. 30 Geriatricians voice the concern that duties to report conditions that increase risks ofdriving, such as Alzheimer's disease, will unfairly burden those who seek care.31 Lawyers defend stringent obligations of confidentiality as necessary to adequate assistance ofcounsel. And so on. None ofthese empirical contentions have undergone rigorous scrutiny, perhaps because of ethical and other pragmatic difficulties in designing the research. (Would it be morally permissible to randomize defendants to confidentiality-protecting and non- confidentiality-protecting lawyers, and assess the resulting quality ofthe lawyer- client communications?) Orentlicher could, but does not, explore the troublesome point that moral considerations, about experimentation and the delivery of professional services, partially explain both the power and the insubstantial ity ofthe perverse incentives concern. 32 Ironically, while supporting the concern, moral considerations also at least partially explain our inability to assess how serious it really is, and thus Orentlicher's ability to determine its genuine weight. 26. A well-known example of such difficulties is the case of Angela Carder, a pregnant woman dying of osteosarcoma. She was compelled to undergo a caesarean section by court order, but neither she nor the infant survived. See In re A.C., 573 A.2d 1235 (D.C. Cir. 1990). 27. Orentlicher, supra note 1, at 1 16. 28. This issue is discussed more fully in Part II, infra. 29. See, e.g., Tarasoffv. Regents of the Univ. of Cal., 551 P.2d 334 (Cal. 1976). 30. See, e.g., Ronald Bayer, Private Acts, Social Consequences: AIDS and the Politics of Public Health 1-72 (1989). 3 1 . Bruce Jennings, Freedom Fading: On Dementia, Best Interests, and Public Safety, 35 Ga.L. REV. 591,602(2001). 32. See, e.g., ORENTLICHER, supra note 1, at 90 (stating that the conclusion will depend on "how one estimates unknown empirical data"). 2002] BOOK REVIEW 997 The other alternative is that pregnant women have a legal duty to accept medical care, at least when it is beneficial both to them and to their fetuses. This alternative conflicts with the legal doctrine that competent individuals do not have a duty to accept medical care, even when it clearly would benefit them.33 But the situation of a pregnant woman who plans to carry the pregnancy to term is different, Orentlicher contends, because parents give up liberty rights in deciding to have children. Parents, for example, are not free to travel without considering their children. 34 Orentlicher admits that the legal duty of parents to care for their children has not been extended to the obligation to undergo medical treatment. Indeed, there are no reported cases of parents being legally required to undergo even the most minimal medical procedures, such as donating blood, to preserve the life or health oftheir children. 35 However, Orentlicher contends, the crucial feature of this legal limit is that the parent's interests are in some degree of conflict, however minimal, with the child's. When the pregnant woman would also be benefitted, Orentlicher contends, there might be a case for imposing the legal obligation to accept the care on her. Here, his analogy is to asking a parent to step aside so that a physician can offer needed medical care to the child. 36 This limited legal obligation, Orentlicher contends, would be consistent with other legal obligations which society imposes on people when both they and others would benefit. 37 Orentlicher admits that his argument here employs what some might regard as a "narrow view ofharm and benefit to the pregnant woman." 38 Although the pregnant woman would not be forced to undergo medical care that puts her health at risk, she would be forced to undergo care that might adversely affect her non- medical interests: in liberty, in dignity, in religious commitment, or even in body image. Invasions of such other interests, Orentlicher counters, are frequently countenanced when special relationships generate duties ofcare. The mere fact of a bodily invasion should not matter, he writes: "[I]t is hard to see why bodily invasions are worse than other limitations on autonomy, beyond the fact that they pose a real risk to one's health." 39 His critics might reply that it is this very invasiveness which matters. His critics might also point out Orentlicher here abandons his reluctance to substitute external, legal judgments for the 33. See, e.g., In re Duran, 769 A.2d 497 (Pa. Super. Ct. 2001); Stamford Hosp. v. Vega, 674 A.2d 821 (Conn. 1996); Pub. Health Trust of Dade County v. Wons, 541 So.2d 96 (Fla. 1989); Bouvia v. Super. Ct., 225 Cal. Rptr. 297 (Cal. Ct. App. 1986). 34. Orentlicher, supra note 1 , at 99. 35. Orentlicher simply cites a constitutional law text in support of his claim. Id. at 205 n.52. Although it is always difficult to prove a negative, the stronger claim advanced here is based on a LEXIS search of combined state and federal case law and all law review data bases, with the following search: (parent! or mother or father) and (fore! or compel! or request!) and ((donat! or giv!) W/2 (blood or (bone marrow))). 36. Id. at 108. 37. Mat 110. 38. Id. at 112. 39. Id. 998 INDIANA LAW REVIEW [Vol. 35:989 individual'sjudgment ofher own best interests, which he used to support a bright line between withholding or withdrawing care and aid-in-dying. Applying Orentlicher's second approach both requires a decision about whether the imposition of medical care is in a pregnant woman's health interests, and grants that invasions of health interests are especially serious. Critics might well question these judgments about interests, and urge instead drawing a bright line against legal coercion of pregnant women to undergo medical treatment, even when it is supposed to be for their own benefit. This line ofcriticism is stronger still if there are reasons for concern that some judgments here are tainted by injustice based on race or sex, as Part II below explores. C. Tragic Choices andMoral Deception Democratic societies find it hard to make open decisions in the face of deep moral conflicts and high stakes for individual lives. So such societies sometimes disguise the decision rules they actually use. "Tragic choices" 40 of life and death may be made by replacing explicit value choices with apparently neutral rules and procedures, such as markets, lotteries, aresponsible agencies, or technical experts. Judgments that further medical care would be "futile," in Orentlicher's example, frequently represent the subterfuge of transforming an evaluative judgment that rationing is justified into an apparently technical determination that a patient is not a candidate for given therapy. 41 The moral justification of such subterfuge is Orentlicher's third translation problem. Judgments that further medical care would be futile are notoriously plastic. 42 These judgments range from the simple factual claim that care will not achieve its intended result (an antibiotic will not work against a virus) to evaluative judgments that given goals are not worthwhile (why maintain a patient in a persistent vegetative state?), to rationingjudgments that care is too costly in light ofpossible results (why maintain an anencephalic infant at great state expense?). Despite careful efforts to establish conceptual clarity about futility, 43 judgments intertwining these concepts remain. Deliberately cloaking ajudgment to ration in the language of science is morally troubling because it may generate mistrust of medicine and disenfranchise families and patients from challenging 40. The term is from Guido Calabresi & Philip Bobbitt, Tragic Choices ( 1 978). 41. Orentlicher, supra note 1, at 130-31. Another example may be the current trend towards using "evidence-based" medicine in making decisions about what procedures should be covered by various health insurance programs. See, e.g., Alan M. Garber, Evidence-Based Coverage Policy; Insurers Can Borrowfrom Research into Medical Effectiveness to Help Them Allocate Medical Resources Wisely, HEALTH Affairs 6282 (200 1 ); J.D. Swales, Science andHealth Care: An Uneasy Partnership, 355 LANCET 1637-40 (2000). 42. See, e.g. , Robert D. Truog et ai., The Problem with Futility, 326 NEW ENG. J. MED. 1 560 ( 1 992); Steven H. Miles, InformedDemandfor 'Non-beneficial ' Medical Treatment, 325 NEW ENG. J. MED. 512 (1991); Lawrence J. Schneiderman et al., Medical Futility: Its Meaning and Ethical Implications, 1 12 ANNALS OF INTERNAL MED. 949 (1990). 43. See sources cited supra note 42. 2002] BOOK REVIEW 999 supposedly technical judgments. But if decisions by physicians to ration can themselves bejustified-and Orentlicher believes they sometimes can-then there may be moral reasons which support keeping them under wraps. Rationing decisions are inevitable, Orentlicher argues, and physicians, familiar with the nuances ofparticular cases, will inevitably be involved at some level in making them. 44 In a discussion remarkably at odds with his original support for general rules, Orentlicher notes that rules are indeterminate, 45 physicians will not tolerate guidelines with which they disagree, 46 and physicians in any event are likely to make decisions that are more sensitive to the circumstances of individual cases, 47 Orentlicher offers no explanation for his divergent treatment of rules; perhaps he is simply more comfortable with physicians making judgments to ration care than with physicians making judgments to aid death. If so, his conclusions seem ad hoc, suggesting the appropriateness of further reflection on when resort to generally valid rules is helpful. Part II below suggests that concerns to counter the effects ofbackground injustice may help to explain resort to rules in some contexts but not in others. Orentlicher then surveys the moral reasons that might support disguising the decisions to ration as judgments about futility. The first is that apparently scientific judgments of futility are needed to counter stringent rules about withholding or withdrawing care, in particular the use in some states ofthe clear and convincing evidence standard. Since physicians are not obligated to provide care that is not medically indicated, futility judgments might serve to avoid invocation of the strict evidentiary standard. As a matter of fact, however, Orentlicher believes, futility judgments are most likely to be made in the cases in which the evidentiary standard tends to be less strict-cases in which the patient is terminally ill or in a persistent vegetative state. Thus avoiding the evidentiary standard does not explain resort to futility. 48 Orentlicher's hypothesis about when futility judgments are invoked may be accurate, but it is interesting to speculate whether it will continue to hold in the aftermath of several major states' adoptions ofthe clear and convincing evidence standard for patients who are not terminally ill or in persistent vegetative states. 49 A second explanation for disguising rationing as futility may be the need to avoid costly reconsideration of rationing decisions on a case-by-case basis. Orentlicher believes there are examples of the use of bright-line rules to avoid case-by-case reconsideration, but he does not believe they have typically invoked futility. His examples are the variety of brain death statutes and the Oregon rationing scheme. 50 44. Orentlicher, supra note 1, at 150-52. 45. Id. at 149. 46. Id. at 150. 47. Id. at 152. 48. Id. at 154. 49. See Conservatorship of Wendland, 28 P.3d 151 (Cal. 2001); In re Edna M.F., 563 N.W.2d 485 (Wis. 1997). 50. Orentlicher, supra note 1, at 157. 1000 INDIANA LAW REVIEW [Vol. 35:989 The need to hide tragic choices, in Orentlicher's judgment, is a third explanation that does justify disguising rationingjudgments as judgments about futility. Because we cannot say to individual patients that a choice has been made between preserving their lives and providing care to others, we characterize the care as not indicated for them in the first place. The apparent expert judgment that they are "not qualified" replaces the judgment that care for them is worth less than care for others. That burying tragic choices explains the subterfuge does not, however, justify it. One concern, that growing public understanding may undermine the subterfuge's effectiveness, is omnipresent.51 From a moral point of view, an even more important concern is the dishonesty itself. When physicians disguise the bases for their recommendations about care, they violate their patients' trust, albeit for more general social ends. Like Sissela Bok, Orentlicher believes that there may be circumstances in which society has implicitly authorized the deception. 52 Rationing medical care may be an example of such implicit authorization. 53 As a justification for a strategy for translating moral principle into moral practice, this is surely inadequate. Orentlicher recognizes its weakness, but seems to believe that all that can be said here is that we are weighing two values, deception and rationing, and simply need to decide which is the stronger. 54 The perspective of political philosophy, however, would have us ask when it is appropriate for society to disguise decisions rather than to subject them to open scrutiny. Britain, for example, came under intense criticism for disguising rationing decisions as decisions about medical appropriateness without open dialogue. 55 Britain was also criticized for maintaining the reasonable physician, rather than the reasonable patient, standard for informed consent as an implicit rationing strategy. 56 Surely there is more to be said about when it is appropriate to avoid open dialogue, and when deliberative democracy should hold sway. 57 In Bok's example of implicit authorization, undercover policing, there are arguments to be made that the protection is for the benefit of everyone and that the deception is constrained by stringent monitoring. There are also arguments that those ensnared by the deception have themselves acted wrongly and that their treatment is constrained by strict constitutional and other limits. None of these features would appear to characterize the rationing of health care-even if we think the underlying decision to ration care is itselfjust. Instead, we might argue, public scrutiny is exactly what is needed to reassure us that rationing 51. Id. at 162. 52. Sissela Bok, Lying: Moral Choice in Public and Private Life 165-81 (Vintage Books 1989) (1978). Bok's example is undercover policing. 53. Orentlicher, supra note 1, at 163. 54. Id. 55. E.g. , Henry J. Aaron & William B. Schwartz, The Painful Prescription ( 1 984). 56. Robert Schwartz & Andrew Grubb, Why Britain Cannot Afford Informed Consent, 1 5 Hastings Center Rep., Aug. 1985, at 19. 57. The term is from Amy Gutmann & Dennis Thompson, Democracy and Disagreement (1996). 2002] BOOK REVIEW 1001 decisions are as just as they can be. Without such scrutiny, we cannot be sure that hidden rationing does not add the immorality of deceit to the injustice of rationing. II. Ideal Moral Rules and Moral Practice Under Circumstances of Injustice When moral rules are translated into moral practice in ordinary life, they may well confront circumstances ofinjustice. In coining the term "partial compliance theory," John Rawls suggested that the translation rules governing circumstances of injustice might be different from those governing translation into a just world. 58 At the very least, whether the rules are the same requires investigation and argument. Perhaps the translation is merely linear, so that in applying moral rules to circumstances of injustice we should simply try to further progress towards justice, bearing in mind the need to avoid problems like perverse incentives which might impede progress. Or, perhaps, more complex moral considerations come into play, including consideration of what is involved in respect for persons under circumstances of injustice. Each of the translation problems Orentlicher considers in detail arguably involve translation into circumstances of at least moderate injustice. In the United States, access to health care is impressively variable. 59 At any given time, over forty million Americans lack any form ofhealth insurance, and many others are under-insured. 60 Many recent articles have highlighted significant differences in effective access to care by race. 61 Injustice in the distribution of health care may reflect and complicate other persistent forms of injustice in American society. Given such pervasive concerns about justice and health care, it seems reasonable to at least consider how problems ofjustice might affect the morality of translation practices in each of Orentlicher's examples. With end of life decision-making, including aid-in-dying, issues ofjustice 58. See Rawls, supra note 3, at 2 1 5- 1 8. 59. See, e.g. , R. Adams Dudley & Harold S. Luft, Managed Care in Transition, 344 N. Eng. J.Med. 1087(2001). 60. See, e.g., John Holahan & Brenda Spillman, Health Care Accessfor Uninsured Adults: A Strong Safety Net Is Not the Same as Insurance, URBAN INSTITUTE (2002), at www.urban.org (2002). 61. E.g., John Z. Ayanian et al., The Effect ofPatients ' Preferences on Racial Differences in Access to Renal Transplantation, 341 New ENG. J. MED. 1661 (1999); Peter B. Bach et al., Racial Differences in the Treatment ofEarly-Stage Lung Cancer, 341 NEW ENG. J. MED. 1 198 (1999); J. Marie Barnhart & Sylvia Wosser Meil-Smoller et al., Racial Variation in the Use of Coronary Revascularization Procedures, 339 NEW ENG. J. MED. 131(1 997); Risa B. Burns et al., Black Women Receive Less Mammography Even with Similar Use ofPrimary Care, 1 25 Annals INTERNAL MED. 173 (1996); Linda C. Harlan et al., Factors Associated with Initial Therapyfor Clinically LocalizedProstate Cancer: Prostate Cancer Outcomes Study, 93 J. Nat'lCancer Inst. 1 864 (200 1 ); Sidney D. Watson, Race, Ethnicity and Quality ofCare: Inequalities and Incentives, 27 AM. J.L. & MED. 203 (2001). 1002 INDIANA LAW REVIEW [Vol. 35:989 include prior access to care and access to end of life care, as well as background inequalities in distribution of wealth, education, employment, and other social goods. Suppose we believed that in translating moral principles into moral practice, we should take particular care not to magnify unjust treatment ofthose who have already been victimized by injustice. We might then be particularly concerned that end of life decision making not deepen existing injustice. Inadequate access to palliative care or to hospice, for example, might deepen the injustice already experienced by those who, with limited access to care, have had illness diagnosed at later, less treatable stages, or who have received less effective forms oftherapy. There is evidence that African-Americans have more difficulties in access to palliative care than other Americans. 62 There is also evidence that African-Americans are less likely to favor withdrawing or withholding care, as well as physician assisted suicide, possibly in part because of failures of trust and concerns about discrimination. 63 Concerns about deepening already-existing injustice might generate support for adopting bright- line tests that would not draw support in a world of more perfect justice. The American Bar Association Commission on the Legal Problems ofthe Elderly, for example, opposed the legalization ofaid-in-dying absent universal health care in the United States. 64 On the other side, there is longstanding suspicion that women are not viewed as fully autonomous agents in end-of-life decision making. 65 Such suspicion might suggest greater attention to patient choice in decisions whether to withdraw, withhold, or continue care, especially when the patient is a woman. For the example ofcompelled treatment of fetuses, the background concern is discrimination based on race and sex. When the intervention was compelled caesarean sections, data indicated that a high proportion of the patients were women of color. Many also did not speak English as a first language.66 Critics 62. See Vence L. Bonham, Race, Ethnicity, and Pain Treatment: Striving to Understand the Causes and Solutions to the Disparities in Pain Treatment, 29 J. Law Med & ETHICS 52 (2001); LaVera Crawley et al., Palliative and End-ofLife Care in the African-American Community, 284 JAMA 25 1 8 (2000); Pain Drugs: Access Difficult in Minority Neighborhoods, AM. HEALTH LINE, Apr. 6, 2000; Palliative Care: Bridging the Race Divide at Death, AM. HEALTH LINE, Feb. 16, 2000. 63 . Researchers MeasureAttitudes about DeathAmong Terminally 111 Cancer Patients, PAIN & Cent. Nervous Sys. Wk., July 15, 2000. 64. Leslie Pickering Francis, AssistedSuicide: Are the Elderlya Special Case?, mPHYSIClAN Assisted Suicide (Battin, Rhodes, & Silvers, eds. 1998). 65. See Steven H. Miles & Allison August, Courts, Gender and the "Right to Die, " 18 L., Med. &Healthcare 85 (1990). Miles and August discussed reported cases in which withholding or withdrawing care was sought on behalf on incompetent patients. It is perhaps worth noting anecdotally that the most prominent effort to insist on continued care, over physicians' objections that the care was futile and that the patient's preferences were unclear, also involved a woman patient. Miles, supra note 42. 66. Veronika E.B. Kolder et al., Court-Ordered Obstetrical Interventions, 3 1 6NEW ENG. J. Med. 1192(1987). 2002] BOOK REVIEW 1003 objected that court-ordered maternal intervention was both sexist and racist. If we agree that as a matter of partial compliance theory we should avoid perverse incentives, we might in addition be especially concerned about incentives that operate against already-disadvantaged groups. We might also be concerned about actions that risk new forms of injustice. Nancy Rhoden argued a number ofyears ago that court-ordered caesarean sections, even when they might benefit both fetus and mother, risk creating precedent that can be used in troubling ways tojustify intervention with reproductive liberty. 67 More recently, Laura Shanner has argued in the context of Canadian law that legally compelling pregnant women to seek medical care may have "dangerous" implications for women, even when good moral arguments support the intervention.68 Underlying the disguise of decisions to ration as judgments of medical futility is the problem of macroallocation of social resources. Do we, as a society, spend sufficient dollars on health care, and do we direct the dollars we spend appropriately? This is not a debate that American society has handled particularly well in recent years. At least part of the problem is that we have created some entitlements that are very costly without full examination, and then recoiled from creating other entitlements as a result. The United States offers more renal dialysis than other advanced industrialized countries, at public expense, yet fails to offer universal access to health care. A plausible maxim of partial compliance theory is that when we act to increase justice, we should take care generally not to create new moral roadblocks. Hasty adoption of the end stage renal disease program in Medicare has perhaps operated in just this way. Orentlicher's argument about tragic choices is that we may sometimes find it too painful to explain the application of a rationing decision in the context of an individual case. To be sure, but it is important also to recognize how our silence may contribute to continued injustice in American health care overall. First of all, when decisions to ration are disguised, individual patients may lose the opportunity for dialogue about care options and their desirability. Discrimination against patients who are disabled, elderly, or disadvantaged, may go unexamined. On the other side, we lose the opportunity for dialogue about when care may be undesirable. Hidingjudgments about rationing as judgments about futility may encourage people to think that all care is desirable until it is labeled as futile. An apt example is patient preferences for cardiopulmonary resuscitation. Patients are less likely to choose resuscitation when they have a clear idea of what it involves and what their prognoses with it may be.69 Our acquiescence in disguise on an individual level may contribute to ill-informed or unjust decisions in individual cases. It may also compound our inability, on a social level, to further justice in health care. Without the realization that we, as 67. E.g., Nancy K. Rhoden, The Judge in the Delivery Room: The Emergence of Court- Ordered Caesarians, 74 Cal. L. REV. 1951 (1986). 68. Laura Shanner, Pregnancy Intervention and Models of Maternal-Fetal Relationship: Philosophical Reflections on the Winnipeg C.F.S. Dissent, 36 ALBERTA L. REV. 751 (1998). 69. Donald J. Murphy et al., The Influence of the Probability of Survival on Patients, Preferences Regarding Cardiopulmonary Resuscitation, 330 NEW ENG. J. MED. 545 (1994). 1004 INDIANA LAW REVIEW [Vol. 35:989 individuals, sometimes find care no longer worthwhile we may also lose the opportunity, as a society, to discuss when we believe care is worthwhile and when we believe it is not. In short, the failure to face tragic choices is not always benign or limited to the circumstances of the individual case. If problems of cost-control and access are inter-related, as many commentators contend,70 the tragic choices strategy may play a role in entrenching the injustices in American health care today. Conclusion Matters of Life and Death makes important contributions to our understanding of how we apply moral principles in the complex circumstances of moral life. Orentlicher explores three concerns which arise in translating moral principles into practice: our need for principles; our need to be sure that our principles, when we have them, do not create perverse incentives; and our occasional need to disguise from ourselves the principles we are using, when we use them. Orentlicher shows us how each of these translation strategies has shaped our approach to moral problems in contemporary health care, from end of life decision making, to public health, to decisions to ration health care. That there is more to be said about how the translation strategies might interrelate or apply against the background of existing injustice by no means diminishes the insightfulness of Orentlicher' s work. Indeed, it only suggests how a good book prompts us to think harder and to do better. 70. See, e.g., Norman Daniels & James Sabin, Last Chance Therapies and Managed Care: Pluralism, Fair Procedures and Legitimacy, 28 HASTINGS CENTER REP. 72 (1998).