Avoiding Paternalism

Abstract

Many people think paternalism is wrong: that it is wrong for the government to limit our liberty for our own good. Some government policies that seem paternalistic, however, also seem morally permissible, at least to many of us, seatbelt laws, for example. For this reason some philosophical opponents of paternalism seek to show that no widely supported policy is really paternalistic, by identifying a good nonpaternalistic justification for the seemingly paternalistic policies that most of us endorse. In this way they aim, as Joel Feinberg puts it, “to reconcile our general repugnance for paternalism with the seeming reasonableness of some apparently paternalistic regulations.”1 This project of reconciliation would be important if paternalism were always wrong, but I see no good reason to think so. Paternalism seems repugnant because it seems infantilizing. In limiting our liberty for our own good, it seems that the government treats us like children or that it impedes our development into fully mature adults, but there is no reason to think this is true of every paternalistic policy. Some liberties have a special value in symbolizing the status of adulthood within our society, the freedom to marry, for example. Having this freedom also provides an important kind of control over the shape and direction of our lives, and creates important opportunities for deliberation and choice. So it makes sense to think the government would treat us like children if it were to make our marital decisions for us. Not every liberty, though, has this kind of significance: the freedom to drive without a seatbelt does not. So there is little reason to think that every paternalistic policy is infantilizing in this way. Paternalism may be disrespectful on some other ground, but I do not think it always is, and in what follows I explain why. My goal here is not to defend any particular paternalistic policy, but to challenge the general presumption against paternalism that I believe distorts our thinking. Some government policies, such as drug and prostitution laws, strike us as paternalistic, I think, because the paternalistic reasons for these policies strike us as the most compelling, at least when our antipaternalistic guard is down. If, however, a general principle of antipaternalism is valid, then we should evaluate these policies by evaluating whether or not there is sufficient nonpaternalistic reason for them. In this way a principled antipaternalism requires us to ignore the reasons that initially strike us as the most compelling. Of course these reasons may not be the strongest, and they may in any case be too weak. Nonetheless I believe that our evaluation of such policies will be more sensitive to the relevant considerations if we freely consider the paternalistic reasons in their favor and conclude that they must be justified nonpaternalistically only after we conclude that the paternalistic reasons are too weak to justify the specific burdens these policies impose. I should say at the outset that it is wrong in my opinion for the government to limit certain basic liberties such as freedom of political expression for paternalistic reasons. My claim here is only that paternalistic interference with some liberties may be justifiable. I will not attempt to defend a general theoretical distinction between those liberties that may be limited for paternalistic reasons and those that may not be, but I assume that a nonarbitrary distinction can be defended. I also assume that the government has a general moral obligation to recognize and protect against paternalistic interference a sufficiently wide range of important life-shaping decisions to ensure that we have adequate control over our lives, enough to achieve genuine autonomy and independence. My position is that since some paternalistic policies are compatible with this general obligation, there is no compelling reason to think that paternalism is always wrong. The project of reconciliation is to show that there are good nonpaternalistic justifications for the seemingly paternalistic policies that many of us endorse. A recent example is Seana Shiffrin's nonpaternalistic defense of the unconscionability doctrine in contract law, according to which the courts are justified in refusing to enforce unconscionable contracts by the fact that a government has no general obligation to facilitate agreements that its citizens regard as exploitive or seriously immoral.2 Another recent example is Elizabeth Anderson's nonpaternalistic defense of compulsory contributions to health insurance and retirement programs, according to which the government is justified in taxing us to fund these programs by the fact that we have a general obligation to secure the conditions of everyone's freedom as a democratic citizen and a person who is sick or impoverished in old age cannot participate fully in the democratic process.3 Whether these arguments can establish, however, that the relevant policies are not paternalistic depends partly on what paternalism is. We commonly say that a policy is paternalistic if it limits a person's choices “for her own good,” where this is to limit her choices for a certain kind of reason: that this policy will promote her welfare or improve her situation in some other way. What is it, though, for the government to limit a person's choices “for” a reason? A natural answer is that the government limits a person's choices for a reason if and only if it adopts this policy because (and only because) those in the relevant political process (legislative, judicial, or administrative) count this reason in its favor. This suggests a motivational interpretation of paternalism: a policy that limits a person's choices is paternalistic toward that person if and only if the government adopts this policy because those in the relevant political process count the fact that it will benefit this person as a reason in its favor.4 This interpretation of paternalism fits poorly, however, with the project of reconciliation. To see why, consider a tension in Shiffrin's position. Her article aims to show that the unconscionability doctrine need not be “defended on paternalist grounds or applied in a paternalist way”5 by presenting a plausible nonpaternalistic rationale for this policy.6 If, however, Shiffrin's “motive-based characterization of paternalism”7 is correct, it is, strictly speaking, not necessary to provide a plausible nonpaternalistic rationale for this policy in order to show that it need not be paternalist, since the mere psychological possibility of a nonpaternalistic motive, no matter how feeble a justification it provides, suffices to show this.8 Furthermore, if Shiffrin's motive-based characterization of paternalism is correct, then the nonpaternalistic rationale for the unconscionability doctrine that she proposes does not actually show that this policy is not paternalist, even if it provides a fully adequate justification, since the availability of a fully adequate nonpaternalistic justification for a policy does not entail that government officials are not paternalistically motivated in adopting and applying this policy.9 It is tempting in view of this tension to characterize paternalism without reference to motive, but characterizations of this kind have odd consequences. Suppose we say that a policy is paternalistic toward A if and only if it can be justified only by paternalistic reasons (understanding paternalistic reasons as those that cite some benefit to A that A does not want).10 It follows from this analysis that every paternalistic policy is justifiable (since if a policy can be justified only by paternalistic reasons, then it can be justified by these reasons, from which it follows that it is justifiable), and no one wants to claim that every paternalistic policy is justifiable. Suppose we say instead that a policy is paternalistic toward A if and only if it cannot be fully justified unless paternalistic reasons are counted in its favor. From this it follows that every unjustifiable policy is paternalistic (since if a policy cannot be justified, then it cannot be justified unless paternalistic reasons are counted in its favor),11 and no one wants to claim that every unjustifiable policy is paternalistic. These problems can be avoided by holding that a policy is paternalistic toward A if and only if (a) this policy cannot be fully justified without counting any paternalistic reason in its favor, and (b) this policy would be fully justified if paternalistic reasons were allowed to count in its favor (which they might not be).12 But this interpretation also has odd consequences. It implies, for example, that, if the benefits of drug prohibition to potential drug abusers are too small to justify this policy, drug prohibition is not paternalistic even if it is adopted in order to protect potential drug abusers who do not want to be protected in this way and even if this policy cannot be justified in any other way.13 More generally, this interpretation entails that no policy is paternalistic if, as a matter of fact, paternalistic reasons are never strong enough to justify government policies, as some opponents of paternalism might argue.14 In light of these difficulties I believe the project of reconciliation is best understood to presuppose a hybrid characterization of paternalism that combines both motivational and justificatory elements. An example of this approach is that a government policy is paternalistic toward A if and only if (a) it limits A's choices by deterring A from choosing to perform an action or by making it more difficult for A to perform it; (b) A prefers A's own situation when A's choices are not limited in this way;15 (c) the government has this policy only because those in the relevant political process believe or once believed that this policy will benefit A in some way;16 and (d) this policy cannot be fully justified without counting its benefits to A in its favor.17 The project of reconciliation makes sense, then, because if there is sufficient nonpaternalistic reason for a policy,18 then it is not paternalistic. But this proposal also avoids the undesirable implications that every paternalistic policy is justifiable and that every unjustifiable policy is paternalistic. Finally, this proposal leaves open the possibility that some policies are genuinely paternalistic even if paternalistic reasons are, as a matter of fact, always too weak to justify the government in limiting a person's choices in ways that he or she does not want. Not everyone will be satisfied with this particular characterization of paternalism. Imagine the government makes free psychotherapy available to gays and lesbians to help them change their sexual orientation, and that it does so with the aim of helping them to lead happier, more satisfying lives, and that this policy cannot be justified in any other way. Some may think that if the intended beneficiaries do not want this policy, then it is objectionably paternalistic even if it does not, strictly speaking, limit anyone's choices. Or imagine someone who wants the government to make his marital decisions for him because he thinks he is bad at choosing a suitable mate. Some may think this policy is paternalistic, even though this person wants the government to adopt it, in virtue of the role the government would thereby play in guiding this person's life. Each of these objections might be addressed by characterizing paternalism in a different way, but I doubt that any characterization will match all our intuitions since I suspect we have different and conflicting intuitions about what policies are paternalistic. So I will simply assume that the four-part characterization stated in the previous paragraph is correct in order to focus the discussion that follows. This characterization identifies fewer policies as paternalistic than other possibilities do, but policies that meet this characterization are among those that critics of paternalism have found most troubling.19 It thus makes sense to focus on these policies here given that my aim is to challenge the view that paternalism is always wrong.20 Because our intuitions about paternalism are unsettled Shiffrin aptly observes that “clarifying what paternalism is involves considering why paternalism matters.”21 Since she believes that paternalism matters primarily because paternalistic motives are insulting, she is drawn to a motive-based characterization of paternalism.22 But one might think that paternalism matters primarily for a different reason. Government policies that limit our choices in ways that we do not want stand in need of justification. So it is important to identify the kinds of reason that can justify the government in adopting policies of this kind. If the fact that a person is benefited by a policy cannot justify the government in limiting her choices in a way that he or she does not want, the justifiable exercise of government power is to this extent limited. Paternalism matters, then, because the moral limits to government authority over our choices matter. If this is right, then reflection on why paternalism matters will support a hybrid characterization of paternalism at least as well as it will support any purely motivational characterization. Assuming, then, that some hybrid characterization is sound, government paternalism is always wrong if and only if at least one of two other principles is valid, a justificatory principle of antipaternalism and a motivational principle of antipaternalism. The precise content of these principles will depend on which version of the hybrid approach is best. Assuming, though, for the sake of argument that my four-part characterization is correct, these principles can be stated as follows. The justificatory principle of antipaternalism: if there is a policy that limits A's choices in some way and A would prefer her situation without this policy, then it is wrong for the government to adopt this policy unless it can be fully justified without counting any benefit to A in its favor. The motivational principle of antipaternalism: if there is a policy that limits A's choices and A would prefer her situation without this policy, then it is wrong for government officials to adopt this policy with the aim of benefiting A. If neither of these principles is valid, then “our general repugnance” toward paternalism is unwarranted. So my aim in this article is to challenge the validity of these two principles.23 Recent endeavors to reconcile widely accepted government policies with the principle of antipaternalism presuppose that paternalism is insulting, demeaning, or disrespectful in some way.24 If paternalism is insulting in virtue of its motive, as Shiffrin contends,25 then success in the project of reconciliation does not directly address what is objectionable about it, since government officials might have paternalistic motives even if there is a good nonpaternalistic justification for the policies they endorse. Success at reconciliation would still be of indirect value, though, since it would show how government officials might justifiably support the same policies they now support without having the insulting attitudes toward other citizens that paternalistic motives are thought to involve. What I question in this section is whether there is any good reason to think that paternalistic motives are inherently insulting, in order to challenge the validity of the motivational principle of antipaternalism. When a government official adopts a policy that someone does not want in order to benefit this same person, she assumes this person has an interest in this policy, or that there is good reason for this person to prefer his own situation under this policy. In supporting an unwanted policy for a paternalistic reason, government officials therefore support this policy from the same general concern to protect or advance individuals’ interests that might properly motivate them to support any government policy. What is insulting about this? Shiffrin suggests that paternalism is insulting because in limiting a person's liberty against his will for his own good, the government “substitutes its judgment” about what is good for this person for his own judgment about what is good for him.26 The government, however, commonly “substitutes its judgment” for the judgment of others when it adopts an unwanted policy, whether this policy is paternalistic or not. When the government imposes speed limits, for example, it limits the liberty of those who think they should drive faster than the speed limit allows, and so limits their liberty against their will. In so doing, the government substitutes its judgment that motorists should not drive faster for the judgment of those motorists who think they should drive faster. When the government imposes a speed limit paternalistically it acts on the judgment that the reasons of safety for a motorist to drive slower outweigh the reasons of efficiency and amusement for this same motorist to drive faster, and so substitutes its judgment about the relative weight of these reasons for this motorist's own judgment that the reasons of efficiency or amusement for her to drive faster outweigh the reasons of safety for her to drive slower. But when the government imposes a speed limit nonpaternalistically, it also substitutes its judgment for a motorist's own. It acts on the judgment that the reasons of safety for other motorists to want this person to drive slower outweigh the reasons of efficiency and amusement for her to drive faster, and so substitutes its judgment about the relative weight of these reasons for her own judgment that the reasons to drive faster outweigh others’ reasons against it. In both cases, the government limits a person's liberty on the basis of a judgment about the relative weight of reasons that she rejects and so “substitutes its judgment” for hers. Why, then, is paternalistic substitution in judgment insulting assuming that nonpaternalistic substitution is not? Some may suggest that it is inaccurate to claim that the government substitutes its judgment for the judgment of motorists in adopting speed limits for nonpaternalistic reasons. They may think that a driver in deciding to drive fast is simply making a judgment about what is best for him, and it is therefore a misdescription to say that the government is substituting its judgment about what is best all things considered for his judgment about what is best all things considered. One might say the same thing, however, about speed limits that are adopted for paternalistic reasons. In deciding to drive fast a driver is not making a considered judgment about what is best for him, one might say; he is simply attending to the advantages of driving fast, and not really thinking about the possible risks. If so, then the government in adopting speed limits for paternalistic reasons is not substituting its considered judgment about what is best for this driver for his own considered judgment about what is best for him. It is true that this driver might be acting on the basis of a considered judgment about what is best for him, in which case the paternalistic regulation would involve a substitution in judgment. But we are justified in believing this only on the assumption that his decision to drive fast has resulted from a deliberative process in which the most relevant prudential reasons have been given some weight (perhaps unconsciously). Since in the abstract we are no more justified in assuming this than we are in assuming that a driver's decision to drive fast has resulted from a deliberative process in which the most relevant nonprudential reasons have been given some weight (perhaps unconsciously), it follows that if we are justified in holding that paternalistic speed limits generally involve a substitution in judgment, we are also

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 73 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.