When, if ever, may or should a healthcare professional withhold a diagnosis of a serious disease from a patient? Or downplay the seriousness of that disease? Or refrain from mentioning the likelihood of imminent death? Is it justifiable for a healthcare team to continue treating a severely burned young man against his wishes in order to prevent his death, in the belief that he will later be glad to be alive? Under what conditions may and should a clinician seek to have a patient involuntarily hospitalized because of risks of self-harm or suicide? When may or should policy-makers seek to develop laws to control individuals’ risky actions even though those actions do not impose risks, costs, or burdens on other individuals or on the society? In short, and more abstractly, when, if ever, are paternalistic actions and policies ethically justified?
Controversies surround the meaning and justification of paternalistic actions in healthcare and of health-related paternalistic public policies. Despite extensive and intensive critiques of paternalism, particularly from the standpoint of respect for personal autonomy, it persists and remains common and important in both contexts. Indeed, it has even gained new appreciation and new momentum over the last fifteen years or so. For instance, in health-related policy, “neopaternalists” have offered arguments for government policies that, at a minimum, seek to protect or benefit individuals through shaping or steering their choices without, in fact, limiting or coercing those choices.1 Similarly, in healthcare, arguments have emerged that the physician should take a leading role in helping patients select the goals of care, instead of merely presenting in a neutral way “a laundry list of means and insist[ing] that patients choose for themselves.”2
This chapter will examine the nature of paternalism, particularly by discussing its moral foundations and limits and drawing several distinctions that bear on its interpretation and justification, consider circumstances in which it may be justified, and, finally, explore the new paternalism, based on recent studies in behavioral economics and psychology, that plays a role in a variety of proposals for health-related public policies.
What Is Paternalism?
Paternalistic actions display at least two features: they aim at protecting or promoting the welfare of individuals themselves, and they do so by not acquiescing in the preferences, choices, or actions of those individuals. Paternalism invokes a metaphor of the relationship between father and child, as depicted in the late nineteenth century when the term emerged. Even before the term itself appeared, the metaphor functioned in the language, for instance, of “paternal government” that John Stuart Mill and others used to criticize governmental policies.3 The metaphor is problematic because it expresses gender roles. However, there are reasons, some feminists argue, for retaining the gendered language of paternalism because it highlights the link between a father’s privileges in a patriarchal family and a physician’s privileges in a hierarchical medical system.4 Nevertheless, both “parentalism”5 and “maternalism—as well as the “nanny state” and various other terms—have been used to cover beneficence-based actions that do not acquiesce in the preferences, choices, and actions of individuals for their own benefit. Each of these terms also has social and cultural baggage. In the final analysis, the term “paternalism” still remains the most appropriate, because of common usage, tradition, and philosophical debates.
Both medical paternalism and governmental paternalism have a moral foundation in the principle of beneficence and/or the virtue of benevolence.6 In pure paternalistic actions, the intended beneficiary must be an individual whose own good is sought (or, for health-related policies, classes of individuals whose own good is sought). The philosopher Gerald Dworkin has defined paternalism as “the interference of a state or an individual with another person, against their will, and justified by a claim that the person interfered with be better off or protected from harm.”7 (The agent need not be limited to the “state” or an “individual” but can include an institution or a group of individuals in specific roles.)
Gerald Dworkin’s definition captures the beneficence/benevolence foundation of paternalism, but there is more debate about how to understand the means employed by paternalists. Specifically, how should we understand “interference” and is it the best term? A broad interpretation of interference is nonacquiescence in the preferences, choices, and actions of others.8 More limited interpretations focus on interference with liberty. Still others focus on infringement of certain moral rules. And so forth. What is involved in “interference” is important for determining whether and why paternalism is ethically problematic and when, if ever, it can be ethically justified. For instance, efforts to persuade an individual to act in certain ways for his or her own welfare are rarely problematic from an ethical standpoint, while coercion always stands in need of ethical justification. Interferences are most problematic and even prima facie wrong when they infringe the principle of respect for autonomy and/or specific rules such as liberty, privacy, or confidentiality.
Still another distinction may illuminate paternalistic acts in healthcare. Gerald Dworkin’s term “interference” captures much active paternalism, which occurs, for example, when a clinician intervenes, perhaps by providing information or treatment against a patient’s request. But “interference” may be too interventionist and too strong to encompass passive paternalism. Consider a healthcare professional’s refusal to perform an action requested by a patient, on the grounds that it would not benefit and would even be harmful to that patient. For example, a physician refuses to perform a requested permanent sterilization because of his or her judgment that it would not be in the sexually active, twenty-year-old requester’s best interests in the long run. Or a physician declines to participate in assisted suicide in a jurisdiction where it is legal because he or she does not believe it would be in that patient’s best interest. Or a physician elects not to provide a futile treatment requested by the patient. In passive paternalism, then, a person refuses to be an agent for the requester, while leaving open the possibility that the requester could find someone else to carry out the request. Other things being equal, it is easier to justify passive paternalism than active paternalism, and the remainder of this chapter will focus largely on active paternalism.
Weak and Strong Paternalism
In the moral analysis and assessment of paternalistic actions or policies, one of the most prevalent and important distinctions is between weak and strong paternalism, first formulated by Joel Feinberg and subsequently elaborated by many others.9 In strong paternalism, the intended beneficiary is deemed to be an autonomous or substantially autonomous person whose preferences, choices, and actions appear to damage or threaten to damage his or her best interests. A strong paternalistic action will, at a minimum, infringe the intended beneficiary’s autonomous choices and actions and thus infringe an important moral principle of respect for autonomy. It may also infringe other moral principles and rules. The burden of justification for strong paternalism is heavy, though not impossible to meet.
By contrast, in weak paternalism, the intended beneficiary’s choices or actions are deemed to be nonautonomous or substantially nonautonomous. As a result, those choices or actions do not carry the same weight as autonomous ones. Indeed, in the absence of patient autonomy, the principle of beneficence encounters no tension with and no resistance from the principle of respect for autonomy. For instance, an incompetent patient—that is, one who lacks the mental capacity to provide informed consent or refusal—may still oppose a particular treatment the clinician believes is necessary. In such a case, the principle of beneficence toward the patient does not conflict with the principle of respect for autonomy because the patient lacks substantial autonomy. Hence, there would be a clear and easy justification for overriding the patient’s opposition to the treatment in question, especially if it is necessary to produce a major health benefit. Various means that would otherwise be morally problematic could also be justifiable, such as nondisclosure of information. An example might be nondisclosure of the diagnosis of Alzheimer’s disease to a patient who already has advanced symptoms.
Is weak paternalism even a form of paternalism? One important conceptual question is just this: Is weak paternalism even paternalism at all, at least in any morally interesting and significant sense, if the intended beneficiary of a beneficent/benevolent deed lacks substantial autonomy?10 Answers to this question clearly affect the pattern of justification for nonacquiescence or interference in a patient’s preferences, choices, and actions. However, it is helpful to use the term “paternalism” for beneficence-based actions targeting the patient’s own best interests when they are in apparent conflict with the patient’s preferences, choices, and actions, even if the patient is substantially nonautonomous.
There is a presumption that adult persons can form their own preferences and choose and act with substantial autonomy. In healthcare that presumption provides a starting point from which clinicians address any conflict that emerges. In doing so they may determine that it is justifiable to disqualify the patient as a decision-maker, for example, because of his or her incompetence and hence lack of autonomy. Caution is needed because judgments of incompetence—and of other failures of autonomy—are not merely scientific and technical; they sometimes mask value conflicts. Hence, it is more illuminating and useful to screen actions, including possibly paternalistic actions, in light of the relevant moral principles and the available evidence about the relevant benefits and harms to the patient and about his or her autonomy. Nevertheless, if the patient is not substantially autonomous, beneficence triumphs easily because the principle of respect for autonomy offers no resistance in those circumstances. Weak paternalistic actions are thus easily justified but are not uncontested instances of paternalism.
By contrast, strong paternalistic actions are clearly instances of paternalism, but are not so easily justified, if, indeed, justifiable at all, in part because, as Ronald Dworkin notes, they are disrespectful, demeaning, and insulting to the intended autonomous beneficiary.11 The most plausible approach, in my judgment, is to keep the door slightly ajar to the justification of some strong paternalistic acts, but to recognize that we should rarely open that door all the way. Beyond the principle of respect for autonomy and related conceptions, John Stuart Mill and others have provided utilitarian reasons for deep suspicion of paternalistic interventions, stressing the odds that the paternalist will be mistaken. According to Mill, “[t]he strongest of all the arguments against the interference of the public with purely personal conduct, is that when it does interfere, the odds are that it interferes wrongly, and in the wrong place.”12 Here Mill offers a more specific utilitarian argument than his broadly utilitarian argument that rules of liberty, where the agent’s actions do not put others at risk without their consent, contribute to human fulfillment.13
Nevertheless, strong paternalistic actions may sometimes be ethically justifiable. By and large, such actions have a significant moral presumption against them, particularly in a liberal society that emphasizes the principle of respect for autonomy and related principles and rules. Nevertheless, theoretically, both beneficence and respect for autonomy are prima facie binding and their respective weights can only be determined in specific situations. What conditions need to be met in justifying strong paternalistic actions that infringe the principle of respect for autonomy because the intended beneficiary is substantially autonomous? Following are several important conditions for justifying strong paternalistic acts in healthcare; these conditions both modify and expand the conditions presented by Beauchamp and Childress:14
1. A patient is at risk of a significant, preventable harm.
3. The paternalistic action is necessary to prevent the harm.
4. The projected benefits of the harm prevention to the beneficiary outweigh its risks to the beneficiary.
5. The paternalistic action involves the alternative that least restricts the beneficiary’s autonomy while still securing the benefits for him or her.
These conditions specify for strong paternalistic actions the broader, more general conditions for overriding any prima facie principle in order to maintain another one.15
Earlier I suggested that physicians might decline, on grounds of weak paternalism, to disclose a diagnosis of Alzheimer’s disease to a patient with advanced symptoms. A more difficult ethical question is whether physicians should tell patients in the early stages of Alzheimer’s disease their diagnosis. Consider the following case: A man in his sixties is brought to the physician by his son, who suspects that his father has Alzheimer’s disease because of his apparent problems in interpreting and handling what used to be normal day-to-day activities, in part because of his lapses of memory.16 The son also asks the physician not to tell the father if she reaches a diagnosis of Alzheimer’s disease. After the appropriate tests, the physician has a diagnosis of “probable Alzheimer’s disease” and discusses with a nurse and a social worker the son’s “impassioned plea” not to tell his father.17
The nurse notes that there is now a settled and fairly strong consensus that patients who have cancer should be told their diagnosis—truthful disclosure is required, because beneficence usually does not outweigh respecting the patient’s autonomy in these cases. (Of course, beneficence should structure how the information is disclosed—a point that is often overlooked.) Nevertheless, the physician wonders just whether and how the principles and precedents regarding the disclosure of cancer illuminate the case of the patient diagnosed with early Alzheimer’s disease. After all, there are some important differences between patients diagnosed with cancer and those diagnosed with Alzheimer’s disease. First, in contrast to most cancers, the diagnosis of Alzheimer’s disease lacks certainty (though over 90% of these diagnoses are confirmed by autopsy). Second, in contrast to many cancers, the course of Alzheimer’s disease is unclear for patents. Third, while therapeutic options have greatly increased for patients with many kinds of cancer, the therapeutic options for patients with Alzheimer’s disease are still quite limited even though they are improving. Fourth, patients with Alzheimer’s disease face an inevitable erosion of their decision-making capacity, which may not occur for most patients with cancer.
The physician in this case also has a specifically paternalistic concern. She worries that telling the patient would harm him, because patients with Alzheimer’s disease tend to have limited coping mechanisms, perhaps because of the neurobiological effects of the disease, and disclosure could lead to functional decline, depression, agitation, or paranoia—perhaps even suicide. As the care of patients with Alzheimer’s disease has evolved, investigators have noted the potential positive benefits of disclosure to patients with early Alzheimer’s disease since, for example, they can usually still participate in important decisions about their present care and in advance about their future care. Although the available evidence indicates that most patients with cancer want to know their diagnosis, the data are less clear for patients with Alzheimer’s disease, even though the evidence indicates that more and more patients now want to know their diagnosis particularly in the early stages.
In view of this range of arguments, clinicians (and family members) of good will may be quite uncertain about the right course of action in these circumstances—as they affirm beneficence, and nonmaleficence, as well as respect for autonomy. There is no reason to believe that this patient diagnosed with early Alzheimer’s disease is substantially nonautonomous. Hence, if the physician determines that disclosure would not be in the patient’s best interest, the rationale would be strong paternalism. In view of the range of arguments just summarized about the disclosure of a diagnosis of early Alzheimer’s disease and the facts of this case, it would be difficult, in my judgment, to justify nondisclosure as an act of strong paternalism.
Pure and Mixed Paternalism
Suppose the professional’s or policy-maker’s motives are mixed—suppose he or she seeks to benefit particular individuals, to protect the public health, to avoid burdens to third parties, and so forth. How should we characterize actions that express such a variety of motives? Debates about pure and impure or mixed paternalism surface in deliberations about both clinical actions and health-related policies. Consider a clinical case in which a cognitively impaired teenager’s parents and healthcare professionals are trying to determine whether it would be appropriate to have her sterilized because she has become sexually active. They concentrate on the young woman’s best interests but also consider the impact a pregnancy might have on several other parties, including potential offspring, the family, and the society. If they elect sterilization, their primary rationale might be paternalistic (the young woman’s best interests), but the other concerns may also be significant in their decision. Hence, their decision might be characterized as impure or mixed rather than pure paternalism.
Rarely are governmental interventions in a liberal society, which professes the value of personal liberty, defended as purely paternalistic. Most often their supporters appeal to the protection of other individuals or of the society, sometimes because of threats to public resources. In contrast to the mid-nineteenth century, when Mill was writing On Liberty, a justification based on public resources is now much more plausible. In the context of public expenditures on healthcare and other goods and services, claims arise about the societal impact of individuals’ actions that may initially appear to harm only themselves. Hence, relevant third-party effects that justify nonpaternalistic or mixed paternalistic actions include not only traditional public health threats to others through infectious diseases, violence, and so forth, but also excessive burdens on public resources.
A pure paternalistic justification would focus solely on the harms prevented or reduced or on the benefits provided to the affected individuals. For instance, a paternalistic campaign targeting obesity would feature the welfare of potentially or currently obese individuals. However, most campaigns, whether against obesity or cigarette smoking, to take two examples, explicitly invoke both paternalistic and nonpaternalistic justifications, often with an emphasis on the latter because they are more palatable to the public in an individualistic society. It is appropriate to describe such campaigns as impure paternalism or mixed paternalism, because the warrant is directed both at the individuals affected and at the impact their actions have on other individuals or on the society.
In debates about legislation to require motorcycle helmets, some arguments in support of mandatory helmet laws are clearly paternalistic—the legislation is intended, at least in part, to protect the motorcyclists themselves. Other arguments strain to show that nonhelmeted motorcyclists increase the risks of harm to others, for instance, in creating hazards to passing vehicles and in imposing unfair burdens on ambulance drivers, emergency teams, nurses, and neurosurgeons, as well as on the public who, in an interdependent society, may pick up part or all of the costs of the motorcyclist’s care. Exempting motorcyclists who internalize the financial costs, perhaps through mandatory health insurance, could reduce some of the externalities. But even so, as in debates about governmental policies to reduce obesity, a contemporary source of major health problems for so many people, the line between what Mill called self-regarding and other-regarding acts and effects may be unclear.18
Paternalistic actions often masquerade, at least in part, as protection of others or of the society, rather than the individual himself or herself, again because these justifications can be more easily accommodated within a liberal framework. Hence, it is important to consider, in light of the best available evidence, the respective roles and merits of different arguments focused on protecting individuals’ best interests, the public health, and the public treasury. For example, as public health returned to center stage in recent years because of newly emergent infections, the possibility of pandemic influenza, and the threat of bioterrorist attacks, the temptation has been to highjack the language and norms of public health to cover private harms and thus to invoke public health rather than paternalism to justify proposed interventions.
Soft and Hard Paternalism
A similar issue arises with respect to soft and hard paternalistic interventions by the government, what Mill called a “paternal government” or what in Britain has been called “the nanny state, a protective but intrusive matriarch, coddling citizens for their own good.”19 In debates about paternalism, the terms “hard” and “soft” have been used in different but overlapping ways. On one interpretation, soft paternalism appeals to values that the beneficiary actually holds but cannot realize because of problems of limited rationality or limited self-control.20 The individual’s preferences, choices, and actions are unwise even by his or her own standards. By contrast, in hard paternalism, the intended beneficiary does not accept the values that the paternalist uses to define the intended beneficiary’s own best interests. While soft paternalism reflects the intended beneficiary’s conception of his/her best interests, hard paternalism reflects the benefactor’s conception of the beneficiary’s unrecognized best interests. According to Glen Whitman, hard paternalists, as representatives of the old paternalism, say: “We know what’s best for you, and we’ll make you do it,” while soft paternalists, as representatives of the new paternalism, say: “You know what is best for you, and we’ll make you do it.”21
The relations between two sets of distinctions are complex—hard and soft paternalism, on the one hand, and strong and weak paternalism, on the other hand. The soft paternalist usually claims that limited rationality and/or limited self-control compromise the intended beneficiaries’ autonomy and thus prevent them from realizing their own recognized best interests. By contrast, the hard paternalist may hold both weak and strong conceptions of paternalism: the intended beneficiary may be viewed as a substantially autonomous agent who has selected and acts on the wrong values or as a substantially nonautonomous person who suffers from encumbrances in reasoning, willing, or acting.
A related but distinguishable conception of hard and soft paternalism focuses instead on the means used to achieve the paternalistic goal. These distinctions are connected because the kinds of means used in soft paternalism generally presuppose greater compatibility with the beneficiary’s own beliefs and values, not yet realized or implemented because of inadequate rationality or inadequate self-control. Hard paternalistic interventions generally ban or prescribe or regulate conduct in ways that coerce individuals’ actions to secure the desired result. These often involve clear trade-offs, as in “sin taxes” directed at harmful conduct such as smoking cigarettes. By contrast, soft paternalistic means tend to influence, shape, or steer individuals’ choices without undermining their freedom to choose—for instance, they often frame information in certain ways while still being truthful and honest. Soft paternalists characterize their means as relatively weak and nonintrusive.22
In light of this second version of the distinction between hard and soft paternalism, with particular attention to means, we can now qualify Glen Whitman’s striking characterization of soft paternalism. Rather than saying, “You know what is best for you, and we’ll make you do it,” the new soft paternalist says, “You know what best for you, and we’ll enable you to do it without curtailing your liberty.” Along these lines, some proponents of soft paternalism even label their position “libertarian paternalism” in order to underline the compatibility between liberty and this kind of paternalism. For instance, drawing on literature in psychology and behavioral economics, Sunstein and Thaler write, “The idea of libertarian paternalism might seem to be an oxymoron, but it is both possible and desirable for private and public institutions to influence behavior while also respecting freedom of choice.”23 As soft or as minimal—another term is “minimal paternalism”—as it is, what they propose is still paternalistic because of the “claim that it is legitimate for private and public institutions to attempt to influence people’s behavior even when third-party effects are absent.”24
The new soft paternalists focus on two major limitations on the intended beneficiary’s preferences, choices, and actions that may justify interventions to benefit him or her. These two limitations also show the convergence between versions of soft paternalism (in both senses) and weak paternalism. The limitations are what Sunstein and Thaler call “bounded rationality” and “bounded self-control.”25 The conception of bounded rationality focuses on the bounds of rational decisionmaking that lead to departures from the “economic assumption of unbounded rationality.”26 For instance, different default rules on the same action will lead many people to make different choices, or techniques of debiasing may correct biases such as the optimism bias that leads many people to underestimate the risks of some actions for themselves, even if they accurately estimate the risks to the population in general. In the presence of bounded rationality, the state may intervene, softly, in several ways, on paternalistic grounds, that is, to enable individuals to choose and act in accord with their overall best interests over time. As Jolls and Sunstein note, if the available evidence were to establish that smokers discounted the risks of smoking because of an “optimism bias,” among other factors, “it is hardly obvious that government would violate their autonomy by giving a more accurate sense of those risks, even if the best way of giving that accurate sense were through concrete accounts of suffering.”27
If private individuals are limited in their decision-making by bounded rationality, is there any reason to believe that governmental officials are exempt from this limitation? And if they are not exempt, then their policies may be unwise, ineffective, or counterproductive. Recognized flaws in human cognition, Glaeser contends, “should make us more, not less, wary about trusting government decisionmaking,” which may even be more flawed than private decision-making.28
Some proponents of soft paternalism argue that, in any event, the government cannot avoid using soft paternalism, for instance, in framing the information it presents. After all, any presentation of information involves some standpoint or perspective and hence some framing because pure neutrality is impossible. However, in my judgment, neutrality should remain an important critical (even if ultimately unrealizable) ideal for many, though not all, situations in which the government discloses information about health-related risks. Others contend that soft paternalism is not only unavoidable but also justifiable, at least in some circumstances. Even though it can sometimes be justified, there are good reasons for sounding a cautionary note about the wave of support for soft paternalistic policies.
Several arguments undergird this suspicion,29 but I will here focus only on a few important points. One apparent advantage of soft paternalism may turn out to be a serious ethical disadvantage from the perspective of social and political philosophy. Recall that this paternalism is soft—that is, it reflects many values that individuals would realize or implement themselves if they did not encounter internal limits of rationality and of control, and the means it employs shape and steer, without thwarting, free choice. As a result, specific paternalistic policies may face little opposition and resistance and provoke few calls for or efforts at monitoring their implementation and effects. Indeed, these policies may even generate their own social and political support. All of this may happen without the transparency and publicity needed for public assessment, in part because some forms of soft paternalistic actions may be incompatible with transparency and publicity—once they are disclosed, explained, and justified, they may be rendered ineffective.
At least, hard paternalistic interventions that involve coercion or serious and explicit trade-offs will be transparent and public, and opponents can mount counterarguments and even resistance. Hence, it is reasonable to suspect that soft paternalistic governmental policies may also be susceptible to abuse in part because of their lack of transparency. Furthermore, government decision-makers may single out some conduct for correction not only or primarily because it involves self-harm over time but also because it displays some moral flaws (e.g., lack of self-control) or is morally distasteful. Not only may such moralistic judgments single out some conduct for correction, among the wide range of acts that involve self-harm, but they may also intensify efforts to censure the conduct and ensure the correction.
Other related ethical concerns also arise. One focuses on stigmatization of conduct that breaches the social norms invoked in soft paternalistic policies. While there is evidence that stigmatization can change behavior, there are also concerns about its psychosocial costs. Proponents of stigmatizing policies usually insist that they target acts, not persons. However, in practice, it is easy to slide from stigmatizing certain conduct to stigmatizing the people who engage in that conduct. For example, this has happened in the United States where, over time, stigmatization has played an increasingly explicit and important role in private and public efforts to curtail smoking: “the anti-tobacco movement has fostered a social transformation that involves the stigmatization of smokers”30 The slide from stigmatization of acts to stigmatization of the people who engage in those acts can lead, as Glaeser31 reminds us, to hostility and even hatred for population subgroups. John Stuart Mill cautioned against treating individuals engaged in harmful self-regarding conduct “like an enemy of society.”32 Again, cigarette smoking may be an example, especially as smoking has now become more common among lower socioeconomic groups in the United States. The ethical concerns, as Bayer and Stuber stress, become “all the more pressing as stigmatization falls on the most socially vulnerable—the poor who continue to smoke.”33
Another possible or even probable slippage provides yet another reason for suspicion of soft paternalism. The acceptance of soft paternalistic interventions, as Glaeser34 suggests, can prepare the way for and even lead to hard paternalistic interventions. One reason is that soft paternalistic interventions succeed in part by increasing support for the social values and norms that undergird the use of those interventions. The campaign against cigarette smoking again provides an instructive example—the movement from disclosure of information to sharper warnings to hard paternalistic measures such as ever-increasing taxation of cigarettes.35
Paternalism is here to stay, in both healthcare and health-related public policies. Yet, that statement is uninformative without a more nuanced analysis of the many different kinds of paternalism this chapter has attempted to identify. Not all paternalistic acts or policies are the same, and they need and receive different justifications and face different limits and constraints. Not all of them, for instance, infringe the principle of respect for personal autonomy. Hence, weak paternalistic acts and policies are more readily justified than strong paternalistic ones. Other distinctions are also important: active and passive paternalism; pure and impure or mixed paternalism; and hard and soft paternalism. While each type of paternalistic act or policy can be justified under some circumstances, the burden of proof is the heaviest for pure, strong, active, and hard paternalism, but other forms, such as soft paternalism, also need close ethical scrutiny for the reasons presented in this chapter.
1. Cass R. Sunstein and Richard H. Thaler, “Libertarian Paternalism Is Not an Oxymoron,” University of Chicago Law Review 70 (Fall 2003): 1159–1202; Thaler and Sunstein, “Libertarian Paternalism,” American Economics Review 93 (2003): 175–179. See also Thaler and Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (New Haven, CT: Yale University Press, 2008).
2. Erich H. Loewy, “In Defense of Paternalism,” Theoretical Medicine and Bioethics 26 (2005): 445–468.
4. Susan Sherwin, No Longer Patient: Feminist Ethics and Health Care (Philadelphia: Temple University Press, 1992).
5. John Kultgen, Autonomy and Intervention: Parentalism in the Caring Life (New York: Oxford University Press, 1995).
6. James F. Childress, Who Should Decide? Paternalism in Health Care (New York: Oxford University Press, 1982); Tom L. Beauchamp and James F. Childress, Principles of Biomedical Ethics, 6th edition (New York: Oxford University Press, 2009).
9. Joel Feinberg, “Legal Paternalism,” Canadian Journal of Philosophy 1 (1971): 105–124.
10. Tom L. Beauchamp and Laurence B. McCullough, Medical Ethics: The Moral Responsibilities of Physicians (Englewood Cliffs, NJ: Prentice-Hall, 1984); see also the discussion in Beauchamp and Childress, Principles of Biomedical Ethics, 6th ed., pp. 208–211, where the terms “soft” and “hard” are used as equivalent to and in place of “weak” and “strong.” I am using “weak” and “strong” in the conventional sense in this chapter while reserving “soft” and “hard” for another distinction regarding paternalism.
11. Ronald Dworkin, Taking Rights Seriously (Cambridge, MA: Harvard University Press, 1977), pp. 262–263.
14. Beauchamp and Childress, Principles of Biomedical Ethics, 6th ed., pp. 215–216; this list is slightly revised from the 5th edition (2001).
16. This fictional case was suggested by the discussion in Margaret A. Drickamer and Mark S. Lachs, “Should Patients with Alzheimer’s Disease Be Told Their Diagnosis?” New England Journal of Medicine 326 (April 2, 1992): 947–951. The presentation of different pro-con arguments in this section, attributed to different characters, draws heavily on this article. However, the case itself is fictional.
17. Years after this chapter was written, diagnostic tests for Alzheimer’s disease are still quite limited. See National Institute of Aging, “Alzheimer’s Disease Fact Sheet,” at https://www.nia.nih.gov/health/alzheimers-disease-fact-sheet (last accessed May 25, 2019). Even in 2019, the fact sheet notes, “Alzheimer’s disease can be definitely diagnosed only after death, by linking clinical measures with an examination of brain tissue in an autopsy.” The available diagnostic methods and tools for use with a living patient can determine with a fair degree of accuracy whether a patient with memory problems has “possible Alzheimer’s dementia” (i.e., the dementia may have another cause) or “probable Alzheimer’s dementia” (i.e., no other cause can be identified).
18. See Mill, On Liberty, pp. 10–11, et passim. For example, Mill distinguishes “self-regarding conduct” from conduct that affects others (pp. 76, 78, passim), discusses “self-regarding” virtues, qualities, deficiencies, and faults (pp. 71–73), and notes that “the engines of moral repression have been wielded more strenuously against divergence from the reigning opinion in self-regarding, than even in social matters” (pp. 14–15). See further the discussion of Mill’s views in chap. 13 in this volume.
19. “The State Is Looking after You,” The Economist, April 6, 2006, http://www.economist.com/opinion/displaystory.cfm?story_id=6772346 (last accessed June 23, 2006).
21. Glen Whitman, “Against the New Paternalism: Internalities and the Economic of self-Control,” Policy Analysis, No. 563 (February 22, 2006).
26. Christine Jolls and Cass R. Sunstein, “Debiasing through Law,” Journal of Legal Studies 33 (January 2006): 199–237.
28. Edward L. Glaeser, “Symposium: Homo Economicus, Homo Myopicus, and the Law and Economics of Consumer Choice: Paternalism and Psychology,” University of Chicago Law Review 73 (Winter 2006): 133–157.
30. Ronald Bayer and Jennifer Stuber, “Tobacco Control, Stigma, and Public Health: Rethinking the Relations,” American Journal of Public Health 96, No. 1 (January 2006): 47–50.
35. W. Kip Viscusi, “The New Cigarette Paternalism,” Regulation (Winter 2002–2003): 58–64.