Kalle Grill

I first distinguish four types of objection to paternalism and argue that only one – the principled objection – amounts to a substantive and distinct normative doctrine. I then argue that this doctrine should be understood as preventing certain facts from playing the role of reasons they would otherwise play. I explain how this filter approach makes antipaternalism independent of several philosophical controversies: On the role reasons play, on what reasons there are, and on how reasons are related to values. I go on to contrast the filter approach with the competing and dominant action-focused approach, which understands objections to paternalism in terms of paternalistic action, behavior, law, policy and the like. Seana Shiffrin and Peter de Marneffe are singled out as prominent recent proponents of this approach. By engaging with their definitions of paternalism, I explain how the action-focused approach makes antipaternalism dependent on the sorting of actions into paternalistic and nonpaternalistic according to what reasons support them. Because one and the same action can be supported by many different reasons, and by different sorts of reasons, such sorting is very difficult. The upshot is that antipaternalism on the action-focused account fails to provide the precise normative implications of the filter approach that I favor.

I first provide an analysis of Joel Feinberg’s anti-paternalism in terms of invalidation of reasons. Invalidation is the blocking of reasons from influencing the moral status of actions, in this case the blocking of personal good reasons from supporting liberty-limiting actions. Invalidation is shown to be distinct from moral side constraints and lexical ordering of values and reasons. I then go on to argue that anti-paternalism as invalidation is morally unreasonable on at least four grounds, none of which presuppose that people can be mistaken about their own good: First, the doctrine entails that we should sometimes allow people to unintentionally severely harm or kill themselves though we could easily stop them. Second, it entails that we should sometimes allow perfectly informed and rational people to risk the lives of themselves and others, though they are in perfect agreement with us on what reasons we have to stop them for their own good. Third, the doctrine leaves unexplained why we may benevolently coerce less competent but substantially autonomous people, such as young teens, but not adults. Last, it entails that there are peculiar jumps in justifiability between very similar actions. I conclude that as liberals we should reject anti-paternalism and focus our efforts on explicating important liberal values, thereby showing why liberty reasons sometimes override strong personal good reasons, though never by making them invalid.

The philosophical debate on paternalism is conducted as if the property of being paternalistic should be attributed to actions. Actions are typically deemed to be paternalistic if they amount to some kind of interference with a person and if the rationale for the action is the good of the person interfered with. This focus on actions obscures the normative issues involved. In particular, it makes it hard to provide an analysis of the traditional liberal resistance to paternalism. Given the fact that actions most often have mixed rationales, it is not clear how we should categorize and evaluate interfering actions for which only part of the rationale is the good of the person. The preferable solution is to attribute the property of being paternalistic not to actions, but to compounds of reasons and actions. The framework of action–reasons provides the tools for distinguishing where exactly paternalism lies in the complex web of reasons and actions

Introduction: Preference, Choice and (Libertarian) Paternalism Kalle Grill & Danny Scoccia This special issue originated in a workshop organized by one of the editors, Kalle Grill, at Umeå University in March 2014, with funding from The Swedish Foundation for Humanities and Social Sciences. The theme of the workshop was Respecting Context-Dependent Preferences. Contributors to this issue who were also speakers at the Umeå workshop are Richard Arneson, Kalle Grill, Jason Hanna, Sven Ove Hansson, Robert Sugden, and Torbjörn Tännsjö. The other speakers at the workshop were Luc Bovens, Sarah Conly, Fabienne Peter and Danny Scoccia. Preferences are context-dependent when they are determined by circumstances other than the contents of the alternatives that preferences range over, such as the way alternatives are presented. As behavioral research increasingly indicates, many of our preferences are context-dependent in this sense. Context-dependence raises difficult issues for many areas of moral and political philosophy. Importantly, the fact that preferences are not in any obvious way based on settled values or considered judgments may cast doubt on liberalism’s stance on when paternalism is morally objectionable: What exactly is liberal antipaternalism supposed to protect, if not settled preferences? What about paternalism that operates by influencing the formation of preferences? Should we, as proposed by so called "libertarian paternalists", shape the choice environment so as to promote choices that are conducive to health, wealth and happiness over the long run? More generally, we may inquire what respect for persons should mean given the context-dependence of their preferences. Several articles in this issue deal with the plausibility of libertarian paternalism and the nudging it endorses. Sugden argues that those who take context-dependence to justify paternalism often rely on ideas of latent preferences that lack a convincing psychological explanation. William Glod argues that, because of informational constraints, nudging often fails to track people's actual preferences. Hanna, in contrast, defends nudging against the objection that it is manipulative. Andres Moles surveys some legitimate aims for which nudging can be justified. The remaining articles either take a more general view of context-dependence, or investigates it in some specific context: Arneson considers, in light of the debate on libertarian paternalism, the general concept of paternalism, and proposes a new theory for when paternalism is justified, in terms of a limited prerogative to be imprudent. Grill argues that respect for persons should be pluralist and include respect for their choices as well as their preferences, even when these two come apart. Tännsjö re-considers, in the light of context-dependence, his earlier rejection of any form of coercive care. Hansson contributes a historical perspective with a discussion of John Stuart Mill's three principles regulating the extent of individual liberty. The guest editors wish to thank the journal editors for giving us the opportunity to produce this special issue. We are grateful also to the reviewers who have helped in the editorial process, and to Margaret Dancy, the Managing Editor, for her friendly support throughout.

Paternalism means, roughly, benevolent interference: benevolent because it aims at promoting or protecting a person’s good; interference because it restricts his liberty without his consent. The paternalist believes herself superior in that she can secure some benefit for the person that he himself will not secure. Paternalism is opposed by the liberal tradition, at least when it targets sufficiently voluntary behavior. In legal contexts, policies may be paternalistic for some and not for others, forcing trade-offs. In medical contexts, paternalism can be an open or hidden aspect of the relationship between caregiver and patient.