The Routledge Handbook of the Philosophy of Paternalism

Kalle Grill, Jason Hanna, Kalle Grill, Jason Hanna

• Publisher: Routledge
• Year: 2018
• ISBN:
  9781317326984
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

While paternalism has been a long-standing philosophical issue, it has recently received renewed attention among scholars and the general public. The Routledge Handbook of the Philosophy of Paternalism is an outstanding reference source to the key topics, problems and debates in this exciting subject and is the first collection of its kind. Comprising twenty-seven chapters by a team of international contributors the handbook is divided into five parts:

• What is Paternalism?

• Paternalism and Ethical Theory

• Paternalism and Political Philosophy

• Paternalism without Coercion

• Paternalism in Practice

Within these sections central debates, issues and questions are examined, including: how should paternalism be defined or characterized? How is paternalism related to such moral notions as rights, well-being, and autonomy? When is paternalism morally objectionable? What are the legitimate limits of government benevolence? To what extent should medical practice be paternalistic?

The Routledge Handbook of the Philosophy of Paternalism is essential reading for students and researchers in applied ethics and political philosophy. The handbook will also be very useful for those in related fields, such as law, medicine, sociology and political science.

Part I

What is paternalism?

1
The concept of paternalism

Danny Scoccia

Before we can determine under what circumstances, if any, paternalism is morally justified, we need to have some sense of what it is. We can answer the normative question only after answering (or assuming an answer to) the conceptual one.

1 Normative concepts; the standard definition of paternalism; motivational vs. justificatory reasons

One of the key questions about the concept of paternalism is whether it is normative.¹ A concept is normative just in case some basic or non-basic normative term (e.g., good/bad, right/wrong, justified/unjustified) is needed to state its definition. “Paternalistic” is often used pejoratively, as part of an accusation (“you’re being paternalistic!”). This might lead one to think that a negative evaluation is built into the concept, as one is in the cases of “thief,” “cruel,” and many other “thick” evaluative terms.

A large number – perhaps the majority – of moral and political theorists assume a definition of paternalism according to which:

P acts paternalistically toward Q (her “target”) just in case:

1. 1 P limits Q’s liberty or interferes with Q’s decision-making;
2. 2 against Q’s will, without his consent, or contrary to his preferences; and
3. 3 for Q’s own good.

Let’s call this the “standard” definition.² If it included the explicit requirement that the means by which the paternalist interferes with choice must violate some prima facie moral rule, then it would certainly be normative (Gert and Culver 1976). But the standard definition does not include such a condition.

There is no question that some paternalism is all things considered justified. A mother’s forbidding her small children to play with matches, sharp knives, etc. for their own good is the very paradigm of a paternalistic action, and it is clearly morally permissible.³ Paternalism is prima facie wrong, however, if all of it stands in need of justification. Deception too seems to be prima facie wrong, but it consists in (something like) intentionally causing what one thinks are false beliefs in others, making the concept/definition of it non-normative. The same may be true of paternalism on the standard definition. On the assumption that it is prima facie wrong to “interfere” with other’s choices without their consent, then conditions (1) and (2) imply that paternalism is prima facie wrong. Paternalism can be prima facie wrong without that wrongness being guaranteed by the concept itself.

Still, the standard definition might yield a normative concept because of (3). (3) implies that paternalism is undertaken for a certain reason, but “reason” here is ambiguous, referring either to a motivation or a justification. This ambiguity gives rise to two competing conceptions of what paternalism is. On an act/motivating reason conception of it, paternalism is a type of action (or practice, law, policy, etc.) that is motivated by altruism, friendship, belief in a duty of beneficence, or the like. The paternalist believes that her interference benefits her target and that this benefit justifies the interference. She need not believe that it by itself justifies it, or even that it is the most important or weightiest reason in support of it. She need only believe that her interference would not be justified if it didn’t prevent harm-to-self. Since it is a non-normative question whether this motivation is present in the paternalist, the motivational reading of (3) yields a non-normative version of the standard definition. On the reading of (3) where a reason is a justification, one needs to know whether an interference really would benefit its target to know whether there are good paternalistic reasons that support it, and thus, whether it is paternalism. Since it is a normative question whether interference would benefit or harm someone, the justifying reason conception makes paternalism a normative concept.

Whether interference with choice that prevents catastrophic harm from befalling the target is for that reason all things considered justified is a normative question that this essay does not address. It addresses only the conceptual question and proceeds by offering clarifications and criticisms of the standard definition. Toward the end it proposes a broader definition.

2 Paternalistic reasons; paternalism vs. moralism; de Marneffe’s “hybrid” characterization

On the justificatory reading of (3), paternalism is, first and foremost, a normative view like retributivism (“desert is a good reason to impose suffering or deprivation on another”). “Antipaternalism” is the view that promoting the good of another person is not a good reason to interfere with his acts or choices, so paternalism, being the opposite of this, must be the view that promoting another’s good is a good and sometimes by itself sufficient reason to do so. But on this reading of (3) paternalism may also refer to certain types of act (law, policy, etc.) whose justification requires an appeal to paternalistic reasons. If thirsty Kurt asks me to share my canteen of water with him and I do so, the fact that he benefits is a good justifying reason for my action, but it is not a paternalistic one. My action doesn’t “interfere” with any of his choices, so there can’t be paternalistic reasons for it. The justifying reason conception should make paternalism a “compound” of an act type (an “interference”) and a justification, as Kalle Grill (2007) has argued.

Justification refers to the good of the person whose choices are interfered with rather than the good of other persons. This distinguishes a paternalistic reason from a “prevention of harm to others” rationale for interference. While that distinction is well understood, the distinction between paternalistic reasons and moralistic ones is often blurred.⁴ Consider laws forbidding prostitution, “dwarf tossing,” or “same-sex sodomy,” defended on the grounds that while those who freely engage in the activities may do no harm to themselves or others, they degrade themselves, thus violating a moral duty to self to avoid self-degradation. This defense of the laws appeals to moralistic rather than paternalistic reasons.⁵

The main difference between the act/motivating reason and the justifying reason conceptions of paternalism becomes apparent if we suppose, first, that a law banning meth use was enacted by legislators all of whom believed that the ban is needed to prevent many people from ruining their own lives, and second, that there is a cogent “prevention of harm to others” rationale for the ban that the legislators wrongly rejected. The act/motivating reason conception implies that this ban is paternalistic, while the justifying reason conception implies that it isn’t.

Peter de Marneffe (2006) has noted that the act/motivating reason conception does not sit well with the project of “reconciliation” that many antipaternalists pursue. “Reconciliation” consists in identifying a nonpaternalistic justification for a policy that one wishes to defend, but is usually defended by others on paternalistic grounds. Richard Arneson (1980: 271–272) pursues this reconciliation strategy with anti-dueling laws, Seana Shiffrin (2000) with the unconscionability doctrine in contract law, and Elizabeth Anderson (1999: 300–301) with the welfare state’s requirement that everyone contribute to social insurance schemes like Medicare and Social Security. Upon identifying a plausible nonpaternalistic rationale for such policies, the “reconciler” claims that the policies are “not really” paternalistic no matter what motivated those who enacted them. De Marneffe suggests that reconcilers operate with a conception of paternalism based on motivating and justifying reasons. According to this “hybrid characterization,” P acts paternalistically toward Q only if P intends to promote Q’s good and there is no good nonpaternalistic rationale sufficient to justify P’s action. To see the attraction of this view, suppose that the members of a religious cult believe (perhaps somewhat inconsistently) both that murder is a sin the commission of which dooms one to eternal damnation and that all murder victims go straight to heaven and suffer no net harm from the premature ending of their earthly lives. The cult establishes a community on a remote island and enforces a criminal ban on murder; the motivation for the ban is to save the souls of would-be murderers. Should we think that this ban on murder is paternalistic? The “hybrid” view explains the appeal of a “no” answer here.

3 “Mixed” paternalism

Let’s say that an act, law, or policy is “mixed” paternalism on the justifying reason conception if paternalistic reasons are necessary but not by themselves sufficient to justify it; they are sufficient only in conjunction with other, nonpaternalistic reasons. An act, law, or policy is “mixed” paternalism on the act/motivating reason conception if the actor in question believes that paternalistic reasons are necessary but not sufficient to justify it. What are some plausible examples of paternalism that’s “mixed” in this sense?

De Marneffe complains that the antipaternalism that motivates “reconcilers” often distorts their thinking about how the laws they support are best justified. He claims that this occurs with the defense of anti-prostitution laws given by some feminists. Instead of defending such laws on the grounds that prostitution is prudentially bad for most prostitutes and johns, these feminists defend them on the grounds that they help prevent the rights violations of women coerced into prostitution. De Marneffe objects that this flies in the face of the evidence, that “the testimony of prostitutes suggests that prostitution is generally voluntary” (2006: 93). Note that even if de Marneffe is right, “generally” is compatible with a sizable minority of prostitutes being coercively exploited. Those of us who think that anti-prostitution legislation of some sort is justified might “split the difference” between de Marneffe and these feminists. Perhaps the paternalistic and prevention of harm to others rationales are each necessary but only jointly sufficient to justify it. If so, then this is an example of “mixed” paternalism on the justifying reason conception.

Consider next Prohibition enacted in the United States in the 1920s. What motivated Congress and state legislatures that enacted it to do so? Surely some legislators were motivated primarily by concern for the welfare of heavy drinkers, some by prevention of the harm to others that such drinkers cause, others by moralism, and still others by a combination of these rationales. Paternalistic motivation may well have been causally necessary but not sufficient to produce this ban on liquor. If so, then it’s an example of “mixed” paternalism on the act/motivating reason conception of what paternalism is.

4 Broad vs. narrow conceptions of paternalism; the soft/hard distinction

Suppose that while very drunk, you accept a dare to perform an exceedingly dangerous stunt, but I intercede and restrain you for your own good. Or suppose a paranoid schizophrenic during a psychotic episode claims that small extraterrestrials have invaded his body and to kill them he must ingest a can of drain cleaner; relatives restrain him and have him committed to a psychiatric hospital for treatment. Are these examples of paternalism? On a broad conception of paternalism, like the standard one with which we began, the answer is “yes.” The interference in these two cases is clearly morally justified, and on a broad conception of paternalism there are many circumstances in which it is justified. On this conception paternalism is more like homicide than murder. Those who operate with a narrow conception of paternalism treat it as more analogous to murder (i.e., as seldom if ever justified). They will object that the standard definition is overinclusive: the range of Q ought to be limited to competent adults, and the range of acts and choices in condition (1) ought to be limited to unimpaired and well-informed ones. Interfering with the choices of drunks or psychotics for their own good isn’t “really” paternalism at all. “Real” paternalism, on the narrow conception, is limited to cases like forcing a life-saving blood transfusion on an adult Jehovah’s Witness who has refused to consent to one, or the state’s requiring that all motorcyclists, including those well informed of the risks of helmetless riding, wear a helmet.

Joel Feinberg’s discussion of paternalism shows ambivalence about which of these conceptions of paternalism to adopt. On the one hand he draws a distinction between “hard” and “soft” paternalism – the former targets the substantially voluntary choices of competent adults, while the latter targets substantially nonvoluntary choices (Feinberg 1986: 12).⁶ Thus, interference with the drunken daredevil or the psychotic is soft paternalism, which implies Feinberg’s acceptance of the broad conception. On the other hand Feinberg admits to not being entirely comfortable with that conception. “It is not clear that ‘soft paternalism’ is ‘paternalistic’ at all, in any clear sense,” he says (Feinberg 1986: 12). A reply to Feinberg is that the narrow conception of paternalism is clearly underinclusive. Again, a mother’s forbidding her small children to play with matches, sharp knives, etc. for their own good is the very paradigm of paternalism.

Soft paternalism includes not just cases involving children, drunks, or other incompetents, but ones involving ignorance or mistaken belief about the consequences of one’s choice, such as John Stuart Mill’s bridge-crosser, who is about to cross a bridge that unbeknownst to him is dangerously unstable, and there is no time to warn him (1977 [1859]: 294). It might be objected that intervention in such cases isn’t paternalism even on the standard definition, because condition (2) is not satisfied. After all, the bridge-crosser has no will or wish to commit suicide or run dangerous risks.⁷

To see why condition (2) (or something like it) is necessary, consider a couple of examples. Dworkin cited a law prohibiting dueling as an example of paternalism, but suppose, as does Arneson in his discussion of this case, that while the vast majority of the upper-class males in the society to which the law applies prefer to accept a duel challenge if one is made over losing honor by declining one, they also prefer even more never receiving such challenges in the first place. The law solves a collective action problem and, under these assumptions, gives its targets the outcome that they themselves judge best. Another relevant example is:

Extortionate Demand: Samantha is starving, but she has $50. The only grocer in town would happily sell her a loaf of bread for $5, but knowing that Samantha is desperate, the greedy grocer is ready to exploit her and demand $50, which Samantha would have to pay. However, the state intervenes, blocking the sale at that price with laws that forbid “price gouging.”

Neither of these examples involves paternalism, because in both the interference enables its targets to realize their own present actual goals or wishes. For interference with someone’s choices to count as paternalism, the paternalist must override or ignore those goals or wishes.⁸

Mill’s bridge-crosser is different from these cases, however, because he does will or wish to cross the bridge. That is, his actual de re intention is to cross this bridge now. The paternalist who detains him thwarts that intention. It’s true that the bridge-crosser would not have the intention if he knew how dangerous it would be to try crossing the bridge. But that’s irrelevant. There remains a clear sense in which interference overrides his will but not Samantha’s or the upper-class males assumed to have the preferences stipulated by Arneson. It satisfies all of the conditions laid down by the standard definition and thus counts as an example of paternalism under that definition – albeit of the softest, most easily justified kind. Since cases like it count as paternalism only on a broader conception of paternalism, they give us another reason to reject the narrow conception for being underinclusive.

5 Different senses of “his own good” and different types of paternalism

The requirement that paternalistic interference with another’s actions or choices be for that very person’...

Table of contents

• Cover
• Title
• Copyright
• Contents
• Notes on contributors
• Introduction
• Part I What is paternalism?
• Part II Paternalism and ethical theory
• Part III Paternalism and political philosophy
• Part IV Paternalism without coercion
• Part V Paternalism in practice
• Index