Paternalism at a Distance

Abstract

I argue that the distance between state and citizen gives state paternalism a pro tanto advantage over paternalism between individuals. Pace Jonathan Quong, the state neither denies nor diminishes my moral status by acting on a justified negative judgment about my rational or volitional capacities. Nor does its failure to paternalize on the basis of detailed information about individuals constitute a source of disrespect. Rather, the less discriminating nature of general legislation both reduces the risk of social stigmatization and avoids a problematic dynamic with the paternalizee. But paternalistic policies may give us reason to be concerned about superiority or contempt in policy-makers towards the citizens at whom they are directed. Governments must remain ‘faceless’ enough for paternalism to operate at a distance, but they must reassure the governed that the judgment that they can do better for them does not conceal the attitude that they are better than them.

INTRODUCTION

Objections to paternalism often focus on paternalism by the state. Instrumental objections may target the likely effectiveness of the state’s interventions or focus on the reduction in citizens’ autonomy that they involve. Non-instrumental objections, by contrast, may claim that the state lacks standing to make the requisite judgments about a person’s good or to intervene on the basis of those judgments. I make no comment here on the strength of those arguments. In this paper I look at two different objections to paternalism, arising from the work of Jonathan Quong, and argue that their implications for state paternalism are in fact either equivocal or positive. I argue that, provided that those who govern have and demonstrate the right kind of attitudes towards the people they govern, the distance between state and citizen presents instead a pro tanto advantage for state over individual paternalism. Although I present my view by contrast with Quong’s, the overall aim – which emerges from Section 4 onwards – is to put the case for a positive way of looking at state paternalism which transcends any particular disputes within the literature.

In Section 2, I examine the argument from moral status. I argue that the negative judgments that motivate paternalistic actions do not impugn the paternalizee’s moral status. On Quong’s presentation, the argument from moral status has non-comparative and comparative versions. In its non-comparative form, it claims that paternalism diminishes the paternalizee’s moral status. As against this, I argue in Section 3 that such concerns about moral status are more plausibly regarded as concerns about social status; and that they can, given plausible empirical assumptions, ground a presumption of sorts against paternalistic policies. But they do not show that paternalism is wrong in itself. In its comparative form, the argument claims that a paternalizer usually denies the equal moral status of the paternalizee. In Section 4 I argue that this version is not even relevant to the state because the state is the wrong kind of comparator. Here, our concern should instead be with the contingently held attitudes of policy-makers and implementers towards those who are subject to the policies. Acting paternalistically towards a person does not necessarily imply disrespect but may be accompanied by it.

In Section 5, I introduce Quong’s second argument against state paternalism, the argument from non-differentiation. Here, the worry is that general paternalistic policies fail adequately to discriminate between people on the basis of their individual characteristics and are disrespectful as a result. On Quong’s view, individual paternalism – where more fine-grained distinctions are possible – is therefore more likely to be justified, all things considered. In response, I argue that there is some reason to think that the state is better placed than are individuals to justify its paternalistic interventions. Non-differentiation is in fact the more respectful approach, and the distance at which the state acts makes an important difference to the nature of its paternalistic interventions.

In Section 6, I draw together the principal claims made here, concluding that the advantage for state over individual paternalism requires a delicate balance on the part of the government. Paternalism is not inherently disrespectful and in fact fits better into the distanced relationship between state and citizen than the close personal relationship between friends, but citizens will be disinclined to view state paternalism in its best light, and thus enable this advantage, unless governments communicate appropriate attitudes towards those whose lives they would seek to improve.

THE ARGUMENT FROM MORAL STATUS

Jonathan Quong claims that ‘paternalism is presumptively wrong because of the way it denies someone’s moral status as a free and equal citizen’.¹ It does this, he thinks, because paternalistic action is characteristically motivated by a judgement that B lacks the ‘capacity to effectively advance his or her own interests’.² The argument comes in two forms. According to the comparative version, setting oneself up as having superior judgment or willpower to another denies that the other has equal moral status. Even where this comparative element is absent, the non-comparative version claims that ‘treating an adult as if he or she (at least temporarily) lacks the ability to rationally pursue his or her own good’ ‘demean[s] or diminish[es]’ his or her moral status.³

The argument from moral status begins with the observation that a paternalizer necessarily makes a negative judgment about the paternalizee’s rational or volitional powers. The ‘core element of paternalism’, according to Quong, is A’s ‘holding a negative judgment about [B’s] capacity to effectively advance his or her own interests’.⁴ But, as several writers have pointed out, it is hard to see how acting on an epistemically justified judgment about a person’s capacities⁵ can be wrong simply in virtue of the content of that judgment.⁶ While a person’s motivation may reflect adversely on her character, we do not generally think that it makes a difference to the permissibility of the act itself.⁷ More tellingly still, this view of the wrong of paternalism requires us to hold that norms of belief-formation respond to moral as well as (or sometimes perhaps instead of) epistemic reasons. This seems implausible to many and at best requires a good deal of work in epistemology to make it convincing.⁸

Any attempt to build an argument against paternalism simply from the presence of a negative judgment about the paternalizee’s powers – call that the ‘argument from negative judgment’ – is threatened by what we can call the epistemic justification objection, that it cannot be wrong to hold an epistemically justified belief or to be motivated to act by an epistemically justified belief.⁹ The evident difficulties of any argument that seeks to build the wrongness of paternalism simply from the negative judgment that motivates the act suggest that something more needs to be said. Quong seeks to push the line of thought further with the argument from moral status:

Even if a paternalistic policy is justified, all things considered, it remains true that this sort of policy involves a prima facie wrongness. The wrongness is that it treats citizens as if they cannot make effective decisions about their own good, and thereby diminishes the moral status accorded to citizens. Even if we decide in the end that the benefits of the policy are sufficiently great that the policy should be implemented, it will nevertheless be true that we must weigh these benefits against the substantial cost of denigrating citizens’ moral status.¹⁰

The argument from negative judgment leaves many asking, ‘what could be wrong with that?’ A’s conduct is supposed to be made wrongful simply by her beliefs – beliefs, moreover, which we can suppose for the purpose of argument are justified. But the same response cannot be given to the argument from moral status. Contemporary political philosophy is egalitarian. Whatever interpretation of equality one subscribes to, no one will want to deny that persons have equal moral status. Even if we reject the political liberal school of thought to which Quong’s argument is a contribution, we can still endorse its egalitarian spirit. The idea that paternalism involves a denial of a person’s equal moral status, then, seems designed to show the sceptic that she is, by her concession that paternalism involves motivation by a negative assessment of B’s rational powers, thereby also committed to an implication whose normative valence she cannot so easily deny. If you accept that there’s a negative judgment, the argument goes, then you’re accepting that there’s a denial of equal moral status; and even if you remain relaxed about the former, you can’t endorse the latter while remaining (what we all profess to be) an egalitarian.

The argument, then, looks like this:

• (1) Paternalism by A towards B involves a negative judgment by A about B’s rationality, emotion management, or willpower.

• (2) If A makes a negative judgment about B’s rationality, emotion management, or willpower, then A either (i) denies that B has moral status equal to her own [comparative version], or (ii) diminishes B’s moral status [non-comparative version], or both.

• (3) It is wrong for a person either to deny that another has moral status equal to her own, or to diminish another’s moral status.

• (4) Therefore, paternalism is presumptively wrong.¹¹

In what follows, I grant that (1) is true. (I will call the relevant judgments simply ‘negative judgments’ for short from now on.) I also grant for the sake of argument that (3) is true, although – as will become clear – I doubt whether it is possible to diminish a person’s moral status. But (2), I claim, is false.

First, a clarification about premise (1).¹² Quong’s specific claim is that paternalism treats the paternalizee as if she ‘lacks the ability to rationally pursue his or her own good’. But it is not obvious that paternalism treats anyone as if they lack anything. To say that a person makes errors in the exercise of an ability is not the same as saying that she lacks that ability. Indeed, it is to say no more than everyone acknowledges, as de Marneffe observes:

Errors in practical judgment are normal, though, whether they are about what is best for oneself or about what is best all things considered. So the supposition that someone is wrong about what is best for him with respect to a particular decision does not imply that he is stupid. It implies only that his rationality is imperfect, and so is open to the kinds of errors we all are.¹³

Whether the paternalizer necessarily regards the paternalizee as manifesting only faults that she (the paternalizer) shares is a question I will address in Section 5. But the point here is that, insofar as the argument from moral status relies on a claim that there is a discrete rational (or moral) power that a paternalizer necessarily treats a paternalizee as altogether lacking, it is unconvincing. (Quong says a person may lack an ability ‘at least temporarily’, but I do not find this move persuasive.¹⁴) In any case, it is not clear to me why an accusation of temporarily lacking a power should be thought to carry any greater implication, in premise (2), for a person’s moral status than an accusation of failing adequately to exercise a power one has. So I stick with the less committal idea of a negative judgment.

The argument from moral status is dialectically awkward. Here is the problem. The argument claims that there is a relationship between my being motivated by a negative judgment about B and my denying B’s equal moral status. What is wrong is the denial, but – the argument goes – I cannot be motivated by the judgment without also making the denial. The argument assumes, reasonably, that no one will want to respond by saying that my denying B’s equal moral status¹⁵ is OK. But for the argument to work, there must be an interpretation of equal moral status on which denial of that status is something distinct from affirming or implying negative judgments. Otherwise, the fact that my paternalistic intervention implied such a judgment would not explain why that intervention denied B’s equal moral status; rather, my implied judgment would just constitute a denial of equal moral status. That being the case, the argument from moral status needs a substantive account of equal moral status such that an opponent will agree both:

• (i) that equal moral status is something that it is important not to deny someone, and

• (ii) that the denial of equal moral status is entailed by negative judgments.

Such an account is going to be hard to come by. Consider proposition (ii). Given that the paternalist thinks that making and acting on negative judgments is unproblematic,¹⁶ presumably she also thinks that such judgments do not deny A’s equal moral status. She rejects (ii). Why is that? What does the paternalist think equal moral status might consist in such that these judgments do not deny it? There are many possible answers to that question. But clearly equal moral status does not consist in (and does not entail) equal rational powers or equally sound judgment, or something like that. That would provide an easy link between negative judgments and the denial of equal moral status. But that is not controversial. No one thinks that for people to be equal in moral status they must also be equal in their ability to choose the prudent option, or equal in their ability to assess the relative merits of a range of choices on offer, or the like. It is equal moral status, not equal abilities.¹⁷

Perhaps the argument is that, although people don’t actually have an equal capacity for making rational choices, if we are to regard them as having equal moral status, we must treat them as if they had such equal capacities.¹⁸ But if that is the argument, then it is no longer clear that the anti-paternalist is going to be able to secure proposition (i). Because now the paternalist may say that, if we must understand a person’s equal moral status as being secured only if we treat her as if she had an equal capacity to make rational choices, then so much the worse for equal moral status. We ought to treat people as having the capacities they actually have – whether that is out of respect for their individuality, or with a view to best promoting their well-being, or whatever – and we should reject any normative political construct that would require us to do otherwise.

It is hard to believe that a commitment to equal moral status could give rise to a duty to engage in an obvious fiction as regards a person’s rational capacities. That is not to say that reasons to act as if something were the case cannot exist. They can, and they do. We may have reasons to act as if we had virtues we do not, or to act as if God existed.¹⁹ We might have had a reason in 1982 to party like it was 1999. How might acting as if a person had capacities she does not be necessary to treating her as having a certain moral status? I can think of two possibilities. The first is that treating a person as having the capacities she actually has in itself reduces her moral status. But that cannot be right. That citizens have equal moral status means that they are to be regarded as equally deserving of moral consideration independently of their personal qualities. The implication of this view would be that recognizing a person’s rational capacities for what they are and acting accordingly in itself reduces the degree of moral consideration that her personhood warrants. I cannot see any reason to draw this conclusion. A person has the moral status she has – that we all have – qua person regardless of the judgments we make about her.²⁰

The second possibility, however, is that treating a person as having impaired rational powers may lead people to think that she warrants a lesser degree of moral consideration. This seems much more plausible. That people may think this does not make it so, of course – but people’s judgments about us constitute our social status. I will have more to say about this in the next section. For now, I will note that, while we may have to pretend away some of a person’s unfortunate lapses in order to protect her social standing, we do not have to pretend that someone lives up to some arbitrary standard of rationality in order for us to be justified in granting her equal moral status – even as we acknowledge that such a failure may result in her social status being lowered. I cannot see, then, what justification there might be for treating a person as if her rational capacities lived up to such a standard in support of the attribution to her of equal moral status.

Let me now consider a different interpretation of the argument from moral status. Where, it might be asked, does freedom feature in the argument?²¹ After all, the moral status that is said to be demeaned or denied by the paternalist is a person’s status ‘as free and equal’. I have argued that a person’s status as equal is neither demeaned nor denied by another’s making negative judgments about her rational capacities. What about the conception of citizens as free? Does paternalism not call this into question?

It is important to recognize, first, that Quong does not locate the wrongness of paternalism in the limitations it places on freedom. He explicitly rejects what he calls the ‘liberty-limiting definition’ of paternalism.²² He also rejects what he calls the ‘autonomy argument’: paternalism, he thinks, is not wrong because it denies or threatens the inherent value of autonomy.²³ Nor does Quong take the moral status argument to be an objection to limitations on freedom as such. He wants his argument to be effective against those he labels ‘liberal perfectionists’, who ‘believe the state should (at least in most cases) not coerce me in order to make my own life worthwhile’.²⁴ Accordingly, the examples he focuses on are generally cases not of liberty-reducing but of choice-enhancing paternalism – because in these cases, as he sees it, the moral status of those whose capacities are negatively assessed is still impugned even though their liberty remains unaffected.

Perhaps, then, the claim is that, in making negative judgments about people’s rational capacities, we fail to respect persons as autonomous beings (that is, we demean their status as free and equal). This worry would apply whether or not the paternalistic action involved restrictions on freedom. This interpretation would bring greater unity to Quong’s anti-paternalistic arguments by making the argument from moral status and the argument from non-differentiation²⁵ both ultimately about respect.²⁶ Finally, Quong’s claim would be distinct from that of the autonomy argument that he explicitly rejects, viz. that ‘to be able to choose is a good that is independent of the wisdom of what is chosen’.²⁷ What is at stake is not the ability to choose but one’s status as a being capable of choice.²⁸

I have two responses to this argument. The first is that it is not clear it can avoid the epistemic justification objection. If what it means to demean a person’s status as an autonomous being is simply to disrespect that person (as autonomous), and if coming to a negative judgment about a person’s ability to pursue a conception of the good amounts to such disrespect, then it looks as though there are some epistemically justified judgments about people that we are barred from making on pain of disrespecting those people. This is hard to believe.

The second response challenges directly the claim that we fail to recognize persons as capable of making choices simply because we make negative judgments about their rational capacities. There is, after all, no such thing as perfect rationality. The recognition of a person’s status as an autonomous being can only ever be recognition of her imperfect capacity to pursue a conception of the good. As such, it is compatible with the judgment that that person’s rational capacities are sometimes subject to error.

SOCIAL STATUS AND A PRESUMPTION (OF SORTS) AGAINST PATERNALISM

One way to challenge the argument from moral status is to show how we can do justice to its concerns on a quite different understanding of paternalism. In this section, therefore, I offer some reasons to think that the argument either misdiagnoses what really concerns us, or identifies features of paternalism that are genuinely problematic but whose normative significance is best understood other than in terms of moral status. I am concerned in this section with the non-comparative version of the argument, that is, the version that relies on the idea of A’s denigrating or diminishing B’s moral status, rather than denying that it is equal to her own. I discuss the comparative version in the next section.

Quong recognizes that sometimes the negative judgment motivating the paternalistic policy will be epistemically justified, and that in such circumstances it may even justify the policy, all things considered. Nevertheless, he insists that ‘we must weigh these benefits against the substantial cost of denigrating citizens’ moral status’.²⁹ This is a general difficulty with the argument from moral status: why is it at all objectionable to treat someone as deficient in a power in which she is in fact deficient? The inference from an assessment that someone’s judgment or willpower is deficient to the denigration of her moral status is far from straightforward. Here is Quong’s reply to this query:

In order to rebut my position about the prima facie wrongness of paternalism, a critic would need to insist that paternalistic policies represent no threat at all to citizens’ moral status. But is this a plausible position?³⁰

Now if the question here really is whether it is plausible to think that A’s forming a negative opinion of B’s powers of judgment, willpower, and emotion management carries no implication that A denies or denigrates B’s moral status, then the answer seems to me to be, ‘Yes, it is entirely plausible’. And it is plausible because the alternative view rests on an account of the grounds of moral status that is itself highly implausible. On that account moral status is sensitive to differences in the kind of rational, affective, and volitional capacities that we know vary from person to person. The whole point of moral status – what explains why it has become such an indispensable tool for moral thinking at large, not just for academic philosophy – is that it is not like that.³¹

Social status, by contrast, is just like that. Human beings are all too ready to think a person less important, or less worthy of attention, or less deserving of consideration, as a result of her shortcoming, however minor, in some property that they regard, with or without justification, as valuable. And in fact Quong’s language at this point suggests that this is where his real concern lies as well. Notice that we are asked not about the necessary implications of being motivated by negative judgments, but rather about the potential threat posed by paternalistic policies. This looks like a different question altogether. This, it seems to me, is a worry about whether the state’s pursuit of paternalistic policies, motivated by negative judgments about a portion of its citizens, might lead to the status of those citizens being diminished in the eyes of their fellows.

We should distinguish, then, between two understandings of the ‘moral status’ terminology. Quong starts off using this to refer to a Rawlsian political assumption – what on a Kantian or Theory of Justice-era Rawlsian account we might think of as a metaphysical status. On the Rawlsian account that Quong favours it is a normative presupposition of liberal political discourse. On either view, that citizens have equal moral status means that they are to be regarded as equally deserving of moral consideration independently of their personal qualities. But there is another kind of status: our social status, as reflected in our concrete relations with others. This is about how others think of us and how we think of ourselves as a result.³²

Does social status have implications for paternalistic policies? I think it is clear that it does. Individuals and groups can easily become stigmatized as lesser citizens if they are widely regarded as needing help that others can do without. And depending on the extent to which pernicious social attitudes tend to be encouraged or reinforced by paternalistic policies, we may even be inclined to think that paternalism in a given society is prima facie wrong even though there may be nothing intrinsically objectionable about it. We may agree with Quong that in such a society paternalistic policies:

are always presumptively wrong: they always stand in need of special justification in the sense that the benefits of any such policy will have to be great enough to outweigh the substantial cost in terms of citizens’ moral status.³³

Having rejected the idea that citizens’ moral status is implicated in paternalistic interventions, I read the reference to ‘moral status’ here as a reference to social status. So read, I think the claim makes more sense. If the argument were really that paternalistic actions, in virtue of their being motivated by a negative judgment about a person’s rational powers, constituted a denigration of her moral status,³⁴ it would be difficult to know how to weigh the cost of this denigration against the benefits to her. The argument from moral status draws its appeal from its suggestion that paternalistic interventions disclose a wrong that is done to a person that is independent of any harm. But the weighing exercise suggested here makes most sense if the real concern is the possible diminution of a person’s status in the eyes of her peers, and in her own eyes, as a result of her being treated as incapable of directing her own life. If that is the exercise, we are being invited to assess the harm done to people in the matter of their social status (and possibly their self-respect) by being on the ‘wrong end’, so to speak, of paternalistic policies against the benefits to them of those policies. This is a more tractable and intuitive problem than weighing up the benefits to a person of a policy against the ‘cost’ of denigrating her moral status by pursuing it.³⁵ This view also makes more sense of Quong’s own wording of his concern about moral status. He challenges the paternalist to affirm that paternalistic policies constitute ‘no threat at all’ to citizens’ moral status. The implication is that citizens’ moral status must surely be undermined or diminished to some extent by the pursuit of the policy, not that it is denied or denigrated simply by the policy’s motivating negative judgment. A person who denies or denigrates something may fail to respect it or to appreciate its value, but she does not threaten its existence or integrity.³⁶

On this view, paternalistic policies do not deny or denigrate anyone’s moral status, but they might lead a person’s fellow citizens to look down on her or regard her as inferior. Given plausible assumptions about people’s tendencies towards hierarchical behaviour, it might seem reasonable to regard such consequences as justifying a presumption against paternalistic intervention. The presumption would arise from the contingently held attitudes that citizens have towards one another, and would require no reference to moral, as distinct from social, status.

What factors might influence the relevant attitudes here? One important factor is likely to be the extent of social and economic inequality that exists already. In societies that are already marked by hierarchies of class, and/or disparities of wealth and income, it may be that paternalistic policies will serve to exacerbate these discrepancies and further entrench pernicious attitudes about superiority and ‘station’. Whereas in less socially and economically stratified societies, a more benign picture of paternalism as seeking to correct for failures of rationality to which we are all susceptible may hold greater appeal.³⁷

It is crucial to recognize the contingency of the case against paternalism that emerges from the foregoing considerations. This is not just because doing so helps to correct our thinking about the wrongness of paternalism, though that is important. It is also important from a policy perspective to diagnose this objection to paternalism as depending heavily on social status, that is, on the way we are perceived by others. If paternalistic policies result in marginalization and hierarchy, that is as much a reason to change the attitudes that produce this result as it is to change the policies. If we acknowledge the contingency of this objection to paternalism, we open up the possibility of criticizing and reforming the attitudes that lead to people’s social status being damaged by their rational or volitional shortcomings. But if we package up these concerns as a denigration of people’s moral status that is constituted by paternalistic intervention, then at best, by laying the blame at the door of paternalism itself, we distract ourselves from the pressing social issues; and, at worst, we buy into these attitudes ourselves and elevate them from a regrettable aspect of human social organization into a philosophical truth.³⁸

STATE AND CITIZEN

In this section I use the term ‘citizen’ fairly loosely, to include all those who are systematically subject to the laws of a certain state. This will include long-term resident non-citizens.

The last section concerned the non-comparative argument from moral status. The claim, which I rejected, was that being motivated by a negative judgment about B’s abilities denigrates B’s moral status. In this section I move on to the comparative version of the argument. This holds that where the judgment is that B’s abilities are inferior to one’s own, what is called in question is whether B has equal moral status with oneself.⁴⁰ So far I have argued that Quong’s argument is unsuccessful in showing that either kind of judgment has any implications for B’s moral status. But now I want to argue that the comparisons that form the basis for the comparative version actually speak in favour of state, and against individual, paternalism.

What grounds the comparative wrongness in paternalism, Quong says, is that ‘the paternalizer frequently believes the paternalizee lacks the requisite judgment or willpower to make the right choice, but the paternalizer holds no such view about himself or herself’.⁴¹ I will argue, first, that what we care about here is not the state’s negative judgments about people’s rationality as such but the attitude of superiority that may accompany such judgments and, second, that this attitude is less likely to be implicated in state than in individual paternalism. To explain this, I need to introduce two distinctions. The first distinction is between two ways of seeing the state – or, more precisely, between two interpretations of the ‘state’ in ‘state paternalism’. On the first interpretation, the state is that abstract entity that continues through successive governments. This is the sense of ‘state’ in which a person is a citizen of the United Kingdom, for example. It is the sense in which a state is a member of the United Nations or is held responsible for its actions in international law. On the second interpretation, what is really meant by ‘state paternalism’ is ‘government paternalism’ I will use the term ‘government’ in a much broader sense than is usual, to mean not only the executive, and not only the traditional three ‘branches of government’ – the legislature, the executive, and the judiciary – but all the individuals who participate in the business of governing. This will include, for instance, civil servants, policy advisors, and anyone else who might be regarded as working for or on behalf of the state in the abstract sense or the government in the narrow sense. I do not think it will be necessary to specify the set of individuals any more precisely, since the nature of the set will emerge from its use in my argument, but those most relevant for my purposes here are lawmakers and others who create and implement paternalistic policies.

The second distinction is between two types of comparative judgment. Type (1) judgments are about who is better able to decide what is good for someone. They are comparisons of competence with respect to the pursuit of (some aspect of) a given citizen’s good. They can be imputed to the state in the first sense, as an abstract entity: the state can take itself to be in a position to act for my good better than I can act myself. Objections to paternalism on this basis are familiar. They question the state’s authority to make such decisions. It is not the state’s place, the objection runs, to make these decisions for me, because it lacks either the expertise or the standing (or both) to do so. Type (2) judgments are reflexive assessments of rational powers. They are comparisons of competence with respect to the pursuit of the good of the subject of the judgment: ‘I’m better at making decisions for myself than you are [for yourself]’. This is the kind of judgment that Quong’s comparative argument targets. The concern is that the paternalizer is setting herself up as somehow better than the paternalizee because she seeks to correct for flaws in the paternalizee that she does not acknowledge in herself.

What are we to make of this second kind of comparative judgment in the case of the state? For the ‘comparative wrongness’ to be present in state paternalism, it must be possible – and even common – for the state to be a party to this kind of comparison.⁴² My claims are that, where the state is understood as an abstract entity, it is not possible for it to be party to this kind of comparison, and that, where the state is understood as a collection of individuals, the real concern here is about the attitudes contingently held by certain individual agents, not one that is implicit in paternalistic intervention by the state simply as such.

Let us take the state as an abstract entity. It makes little sense to treat the state in this sense as an agent that can have views about its own judgment or willpower in the way that those faculties are enjoyed by the individuals who are the targets of paternalistic policies – in the way, that is, that would make the comparison a meaningful one. Type (2) judgments cannot be made by the state as an abstract entity because it is not a human person. When the state is the paternalizer, it is not an apt description of the state-citizen relationship to say that A ‘believes [that B] lacks the requisite judgment or willpower to make the right choice, but [A] holds no such view about himself or herself’.⁴³ The italicized portion of the belief-attribution is not applicable to the state in the first sense. The state in this sense simply doesn’t have judgment or willpower in a way that can be compared with the judgment or willpower of its citizens. It doesn’t have to resist the temptations of sugary and fatty foods, or hold clearly in mind its long-term interest when contemplating some highly risky behaviour for the sake of a minor short-term benefit. The state doesn’t drink Coke or wear seatbelts.

It might be argued that if the relevant comparative judgments are stated at a sufficiently high level of abstraction – ‘I’m better at making decisions about the execution of my projects than you are about yours’ – then the necessary comparisons are viable. But then they are just not the kind of thing that we should have any concern with. For Type (2) comparative judgments to be meaningful for us we need to be in a certain kind of social relationship with the comparator; it must make sense to speak of us as potential social peers. There is no reason for any human being to compare her own rational abilities to those of a complex group agent such as the state.

The people who govern us can make the relevant judgments, and are our (potential) social peers. Note first, however, that there is no reason to think that policymakers must regard themselves as specially immune from the flaws in judgment or willpower that make the paternalistic policy an effective measure. Type (2) comparative judgments are not a necessary feature of paternalism. Legislators and regulators may recognize their own cognitive shortcomings and weaknesses of will, and seek to correct for them, as much as they do for others’. The state actor who does take herself to be somehow constitutionally different from the ‘common person’ indicates by that attitude a failure of virtue or of self-awareness rather than a problem that is internal to paternalism.

But even if it is – as I have argued – a mistake for me to regard my moral status as denied or denigrated when the state imposes fines for failing to wear a seat belt or a tax on sugary drinks to discourage their consumption, my feeling this way may be a natural response to such impositions if, in fact, governments and politicians do sometimes hold in contempt many of the people they govern. There can be no doubt that many do hold such attitudes,⁴⁴ but they are not a necessary implication or an inevitable concomitant of paternalistic policies. Paternalism is motivated by a judgment of the paternalizee’s flawed rationality. That judgment may or may not be accompanied by contempt for the paternalizee. But since such contemptuous attitudes are so common among those who are in positions of power over others, it may not be unreasonable to presume that the members of a government whose policies are motivated by a negative judgment about its subjects’ rationality do not, in fact, recognize similar flaws in themselves and that they look down on the people they govern.⁴⁵

These considerations invite a new way of thinking about the supposed prima facie wrongness of paternalism. The state’s pursuit of a paternalistic policy in respect of a group of citizens does not constitute a repudiation of their equal status, but, given certain historical and social circumstances, paternalistic policies may give us reason to suspect that, in fact, the government looks down on those citizens. And it is incumbent on a state that implements paternalistic policies, not only not to look down on its citizens but to give them reasons to believe that they are not looked down on.⁴⁶ Paternalistic policies will be better justified, all things considered, if the state is able to reassure its citizens that the negative judgments that motivate such policies are not in fact accompanied by an attitude of superciliousness – that the government does in fact respect its citizens.

PATERNALISM AT A DISTANCE

Quong suggests an intriguing asymmetry as regards respect:

We are often happy to have our friends treat us in ways that would be considered wrong if done by a stranger. For example, amongst some friends it is perfectly permissible for one person to borrow something that belongs to another without asking, but this does not suffice to show that it would be permissible for a stranger or the state to behave in the same way. Friendships create particular expectations that change the nature of what treatment is considered respectful and permissible.⁴⁷

That is no doubt right, as far as it goes, but often things are the other way around. What might be morally problematic between individuals, and perhaps especially between friends, is not so between state and citizen. The relationship between state and citizen, I will argue, involves a kind of distance that changes the nature of what treatment is respectful and permissible.⁴⁸

Consider the example that Quong is addressing. Steven Wall contrasts two cases in which I am offered a financial incentive to visit a natural park in order to get me to appreciate the value of natural beauty. In the first, the incentive is $50 offered to me by a friend; in the second, the incentive is offered by the state, in the form of a subsidy. Wall invites us to conclude that no criticism on grounds of invasion of autonomy could be sustained against the friend, and there is therefore no reason to endorse a similar criticism against the state.⁴⁹ Quong’s response is to provide a reason to think that such incentive-giving objectionable in the case of the state where it would not be in the case of the friend:

The best liberal answer to Wall’s question is not that a state incentive to visit the park violates your autonomy, but rather that when the state acts in that way it shows citizens a lack of respect by treating them all, regardless of their individual differences, as if they lack the ability to learn about and appreciate valuable things without being offered a financial incentive to do so. Your friend’s action, though paternalistic and thus presumptively wrong, has a greater chance of being justifiable because we assume he or she will be acting on the basis of detailed information about your particular situation.⁵⁰

Here, then, is the argument from non-differentiation. Treating a competent adult paternalistically is pro tanto disrespectful because it denigrates her moral status. But where the paternalizer (A) is a close friend of the paternalizee (B), this disrespect can be diminished by the respect that A shows when she acts on the basis of detailed information about B’s individual characteristics and circumstances. The more this respect is able to counteract the disrespect manifested in A’s negative judgment about B’s capacities (and the greater the benefit to B) the more likely A’s intervention is to be justified, all things considered. On the other hand, of course, if A’s paternalistic intervention is not pursued on the basis of detailed information, then her disrespect of B is compounded by this failure. It adds a further source of disrespect.⁵¹

I have argued that the negative judgment that motivates paternalism in any case presents no threat to anyone’s moral status. It is not disrespectful to make or to act on such a judgment. So on my view there is no disrespect in need of counteraction. But my claim in this paper is that state paternalism has a pro tanto advantage over individual paternalism. There are two ways in which I try to support that claim here.

First, I argue that it is not so clear that the differentiation that Quong thinks will better justify paternalistic intervention is more respectful. In fact, I think the state’s undiscriminating approach shows more respect. Second, I draw a contrast between the state-citizen relationship and friendship in order to explain why, in one respect at least, paternalism sits more easily within the first than the second.

Respect and Non-differentiation

Quong regards the state’s inability to pursue its paternalistic ends using carefully differentiated treatment as a disadvantage:

One problem in justifying state paternalistic [sic] is thus that the state is a blunt instrument which is usually unable to make fine-grained distinctions in the way it treats different citizens. The nature of the law is such that it usually must be applied to everyone in the same way. The state, as an instrument, is unlikely to be able to assess individual cases to ensure that only those sane adults who, for whatever reason, temporarily lack the ability to rationally pursue their own good, are treated paternalistically. Individuals, on the other hand, are obviously capable of making informed and fine-grained distinctions with regard to their treatment of other individuals, and so it seems reasonable to assume that “person-to-person” paternalism is more likely to be justifiable all things considered than “state-to-citizen” paternalism. In other words, paternalism, although presumptively wrong, may be justified when a case is correctly assessed on its individual merits.⁵²

Our focus here is on legislation that applies to citizens in general. Paradigm cases of such legislation are the kind of subsidy envisaged by Wall and mandatory seatbelt laws. Consider first the situation of those who are helped by the policy, that is, those whose well-being is improved as a result of their being discouraged from engaging in, or encouraged to engage in, some behaviour that harms or benefits them. Here, I think non-differentiation goes some way to defusing the worry about social status discussed earlier. If people are identified as among those whose rational or volitional capacities are such that they would benefit from direction by the state, it is more likely that their status will be damaged in the eyes of their fellow citizens. This is a good reason to prefer less discriminating paternalistic measures.

But what about those whose choices the legislation affects but who are not helped by the policy because their choices are already adequately responsive to the good? In the natural park example, what about those who don’t need an incentive to visit the park but who are being treated as if they do?⁵³ We should remember that the question here is not whether these individuals are disrespected by the state’s being motivated by a negative judgment about them, because the point of the non-differentiation argument is supposed to be that making decisions on an individualized basis counteracts the denigration of moral status involved in making the negative judgment. The non-differentiation, in other words, is supposed to be a further source of disrespect. Notwithstanding the dialectical position, however, I think Quong’s example enables us to deny that, in the case of general legislation such as this, the state is motivated by a negative judgment about its paternalized citizens. It is motivated by a negative judgment about some of them, to be sure – about an indeterminate subsection of the population whose choices may be affected by the policy. But even their precise number has not been ascertained, let alone the individuals identified. This is an implication of the generality of legislation. Understanding this implication sheds new light on the supposed connection between the paternalizer’s negative judgment and the status of the paternalizee. My suggestion is this: although paternalism necessarily involves a negative judgment, the freer that judgment can be of any attachment to specific individuals, the fewer the implications for the paternalizee’s social status, and the less morally problematic the relationship between paternalizer and paternalizee.⁵⁴

Far from compounding the disrespect involved in being motivated by a negative judgment about B, the fact that general paternalistic legislation does not differentiate between individuals’ situations is a further reason to doubt whether, in these cases, paternalism involves any disrespect at all. And non-differentiation tends also to dispel reasonable concerns about the implications of paternalistic policies for the relationships between the paternalizee and her social peers.

To sum up the overall position here: (i) acting on a negative judgment about someone does not amount to disrespecting her (that was the argument of Section 3). But, in any case, (ii) when it passes general paternalistic legislation, the state – more specifically, the legislature – does not make a negative judgment about anyone in particular. This should lead us to doubt (further) that anyone in particular is disrespected. Even if these claims fail to convince, (iii) the state does not compound any disrespect by failing to differentiate between people’s situations; rather, it lessens the risk of social stigmatization by treating people alike.⁵⁵

It is easy to think of state paternalism simply as individual paternalism on a grand scale, as if the state maintained a personal relationship with each individual citizen. On this picture the state is the ‘person A’ acting paternalistically towards person B, person C, person D, and so on for all citizens. But this is misleading. Far from there being a multitude of paternalistic relations between individual actors, there are no such relationships. There is, rather, a relationship between the state and the public at large, in which the state helps to direct and coordinate the choices of the community.⁵⁶

But perhaps something has been missed here. Do we not in other cases find something objectionable in subjecting a large group of people to a policy without taking care to ensure that it applies to each of them? Quong suggests that a paternalistic policy is analogous to racial profiling:

If the state knows that an unknown member of racial group X has committed a crime, and so imprisons everyone from that racial group, we are not tempted to say that, although the policy is unjustified with regard to all the innocent members of X, it is justified with regard to the guilty member. Why should paternalistic policies be any different?

But there are important differences between this case and the normal case of a paternalistic policy. First, the treatment of the racial group violates their rights. It is a standard objection to consequentialist accounts of morality that the overall benefit to society cannot justify incarcerating the innocent.⁵⁷ Subsidizing natural parks and mandating seatbelt-wearing do not violate rights.⁵⁸ More significantly for our purposes, the imprisoned group is targeted on the basis of a criterion that is stigmatizing and perpetuates historical injustice. (Contrast, for example, the decision to detain everyone who was in the house when the murder was committed.) This feature is also absent in the case of general paternalistic legislation. People at large are not stigmatized by being offered an incentive to visit a natural park.⁵⁹

If the racial profiling policy is thought specifically disrespectful, as well it might be, I think it is for some or all of the reasons just mentioned: that it violates the rights of the innocent members of X, that it unfairly stigmatizes them, and that it perpetuates injustice against them. Since none of these features is present in the case of the paternalistic subsidy, I submit that it is not disrespectful.⁶⁰ Finally, I see no reason to ask whether a policy is justified with regard to some in particular who are affected by it. It is either justified or it is not, and its effects on all parties are relevant to deciding that question.⁶¹ If a paternalistic policy violates someone’s rights, or unfairly stigmatizes or causes injustice against some group, these are reasons against it. Even if it neither violates rights, stigmatizes nor causes injustice against anyone, if its benefits to the target group are outweighed by the curtailment of liberty or autonomy that it involves, it will not be justified.

It is useful to consider the relationship of the picture I present here to Ian Carter’s idea of ‘opacity respect’. Carter thinks that to respect persons in a certain way is to ‘treat them as opaque’.⁶² Opacity respect ‘is a substantive moral attitude that involves abstaining from looking behind the exteriors people present to us as moral agents’ and ‘refus[ing] to evaluate persons’ varying capacities’.⁶³ It is not universally appropriate, but it is most clearly apt to govern the relations in the public sphere, and especially the relationship ‘between political institutions and the citizen’.⁶⁴

Carter is trying to make sense of the value of human dignity and the obligations that it imposes on us. As I hope will be clear from the argument to this point, I doubt whether respect does require ‘evaluative abstinence’ in the way that Carter describes. Nor would I endorse the association of opacity respect with a political liberal conception of the role of public institutions. Political liberals believe, Carter notes,

that it is not business of the state, in its role as guarantor of basic rights, to evaluate the degrees to which individuals are able to make rational and responsible decisions, to form reasonable value commitments, to develop worthwhile life plans, and so on, for in doing so the state would show disrespect toward those individuals.⁶⁵

By contrast, I think that the state is justified, in light of its responsibility for its citizens’ well-being, in assessing the rationality of its citizens’ decisions and evaluating the degree to which they enable them to live a valuable life. But notwithstanding these disagreements, I think I can still recognize a crucial role for something like the notion of opacity respect.

Carter and I agree about the value of non-differentiation. It is better that the state enacts paternalistic laws that apply to everyone in the same way, rather than making the ‘informed and fine-grained distinctions’ that would enable it to take each case on its individual merits. But why? In Carter’s view ‘The appropriateness of opacity respect can be seen as supplying a reason for affirming [the Rawlsian] political conception of the person’.⁶⁶ In other words, opacity respect is the more ecumenical idea. The thought is that the appeal of opacity respect draws us towards political liberalism, rather than that a commitment to political liberalism leads us to recognize the value of opacity respect. But this leaves it somewhat mysterious in Carter’s account just why opacity is appropriate to the public sphere but not to private relationships. If it is not that it is implied by an account of the proper role of the state but rather that its independent appeal contributes to motivating the political liberal picture, then what is it that draws us to opacity in the public sphere in the first place?

One part of the answer, I think, is the problem of social attitudes that we have already noted. The state shows respect by not exposing me to the judgment of my peers.⁶⁷ But that is not the whole story. Insofar as we wish to remain opaque to the state in order to limit exposure to stigmatization, we value that opacity only instrumentally. We are opaque to the state in order to be opaque to other individuals. But the extent to which we are transparent to another person or institution is also constitutive of the kind of relationship we have with that person or institution.⁶⁸ And a relationship of full or even substantial disclosure is not the sort of relationship we want to have with the state.

Distance and Anonymity ⁶⁹

Hayek thought that a person was made less free only by (what he regarded as) arbitrary coercion, and not by being governed by general laws: ‘insofar as the rules providing for coercion are not aimed at me personally but are so framed as to apply equally to all people in similar circumstances, they are no different from any of the natural obstacles that affect my plans’.⁷⁰ This led him to some implausible views about the permissibility of the administrative state and the possibility of social justice. In my view, Hayek was right to think that there is an important qualitative difference between one’s behaviour being directed by the will of another and being guided by general legislation, but he misidentified the difference. It is not that being specifically targeted makes us less free. If restrictions are imposed on my actions, I am made less free. But the way in which restrictions are imposed makes a difference. General legislated rules are like ‘natural obstacles’ in the sense that they do not present themselves to me as aspects of my relationship with another person. They are obstacles arranged by the state, certainly, but they are arranged as an adjustment to the environment within which we all act rather than as an imposition on me personally.⁷¹

Opacity is the appropriate default stance for the state to take towards its citizens. The self-disclosure necessary for paternalism to work on an individual basis must be in the context of a personal relationship of trust, and a diffuse, monolithic institution such as the state is not the kind of entity that we can trust in the required sense. We cannot look the state in the eye. But state paternalism, at least through the kind of general legislation envisaged by Quong, does leave us opaque in the required way. State paternalism is paternalism at a distance.

In order to illustrate this, I want to contrast some features of the state/citizen relationship on the one hand, and personal relationships on the other. Consider first the specific type of paternalistic policy that Quong cites – a subsidy to incentivize behaviour thought to be beneficial. Quong’s suggestion is that the state shows less respect to me when it subsidizes a natural park than my friend does when she offers me $50 to visit it. His explanation, as we have seen, is the fact that the friend is in a position to act on the basis of detailed information about the subject and so treats her as a distinctive individual. It seems to me not only that this explanation is wrong but that the initial judgment is mistaken too. It is deeply disrespectful of a friend to offer you a financial incentive to participate in some activity that she thinks you have erroneously concluded will not be valuable to you. The fact that she has chosen to do this on the basis of her detailed knowledge about your preferences, beliefs, and behaviour – acquired in the course of, and partly constitutive of, your friendship – only makes things worse.

Let’s start with the explanation – the paternalizer’s acting on detailed information. Put aside the ‘state vs. individual’ distinction for a moment by dealing only with individuals. Suppose some anonymous (and entirely non-state-affiliated) busybody⁷² posts a letter through your door with a ticket to visit a natural park. In the letter she explains that she thinks people in general fail to appreciate nature, and so she is posting these free tickets through letterboxes at random throughout her neighbourhood. How odd, you think, that someone could be so exercised about this issue that she wanted to spend her money in this eccentric way. But you’re still not interested in natural parks, so you give the ticket to your cousin, who is. Now suppose that the anonymous writer explains in the letter that you are one of a number of her acquaintances whom she has noticed fail to appreciate nature, and she would like to persuade you to reconsider. She includes sufficient information in the letter to convince you that she does indeed know you personally.⁷³

Does it help that the second paternalizer has acted on the basis of detailed information? I don’t think so. There is the possible invasion of privacy, of course, but let’s assume for the sake of argument that the letter deftly conveys the justified impression that the information on the basis of which the paternalizer acts has not been acquired in an intrusive or underhand way. Still, this seems to me hugely objectionable – quite obviously more so than a state subsidy. Why?⁷⁴ The answer, I think, is that I am not anonymous to the paternalizer. Why should that make a difference? Because if I am known to the person who purports to be able to do better for me, an uncomfortable dynamic is set up between us in which I feel vulnerable to the other’s scrutiny and judgment. It is not that they fail to respect me – at least not in the fundamental way that Quong alleges⁷⁵ – but simply that it is unpleasant for me to be confronted so starkly with the reality of someone else’s opinion of me, an opinion which either I do not share, or – if I do – which I may not wish to be forced to come to terms with. (I may be affected by others’ judgments no less for the fact that I know or suspect them to be justified.) And nor should I be forced in this way to come to terms with it.⁷⁶ Opacity respect is a stance towards those not known personally to us that is justified by its tendency to preserve our ability to choose the scope and pattern of our self-disclosure. We are happy – by and large - to lay ourselves open to the scrutiny of friends because friendships are valuable relationships which we make possible by allowing others to learn about us – about our psychological strengths and weaknesses, about our foibles, hang-ups and trigger-points. These compensating benefits are for the most part absent in the public sphere.

If this is right, how should we make sense of the feeling that there is something more appealing about paternalism between friends than between state and citizen? The advantage of friendship, such as it is, is not the advantage of detailed information; it is the advantage that comes with the nature of the friend (or partner, etc.) relationship. Friendship is a relationship of mutual understanding, and mutual recognition. We make an honest effort to get to know our friends because we are genuinely interested in them as people, and we acknowledge and try to make sense of their own self-understandings. If we act paternalistically towards them, we do so sensitively, with due consideration for their likely reactions, both in the short- and the long-term, and for the likely effect on our relationship. To be sure, we couldn’t show this kind of sensitivity if it weren’t for the detailed information we have about each other – but it is the relationship, and not the information, that makes the difference.

These features of friendship all contribute to justifying – and indeed towards rendering respectful – paternalistic interventions within friendships. Now let’s return to Quong’s own example. It seems to me that these same features point towards such financial incentivization being entirely inappropriate in the context of friendship. It is peculiarly insensitive for a friend to offer a financial incentive for me to visit a natural park and, more broadly, to reconsider my outlook on the value of such things. In the context of a friendship the paternalizer may be justifiably accused of treating the paternalizee in a behaviourist fashion that is not fitting for their relationship. But such an approach is not out of place in the ‘arm’s-length’, and less discriminating, relationship between the state and its citizens.

We might be tempted to set aside financial intervention as a special kind of case on the ground that ‘money poisons everything’. Perhaps on the whole paternalism between individuals is still on firmer ground than state paternalism, if not because of the fact of detailed information, then because of the nature of the relationship. But even this is true only in a restricted sense. There are countervailing difficulties in personal relationships that give state paternalism a pro tanto advantage. The fact that we expect our friends to recognize and to credit our self-understandings makes it all the more hurtful when it turns out that they don’t – when it turns out that they think we have some flaw or failing that needs correcting or managing. It may be that, if I accept that my friend is responding appropriately to the evidence, and is genuinely concerned about my well-being, then I should not care about what she thinks of me. But this is hard to do. I want my friend to share my understanding of myself. If she does not – even if I accept that she has good reason not to – there is something missing in our relationship. Maybe I can find my way to accepting that the fault here is in me – that it is I that should see myself differently, rather than she. But even if that is the case, her seeing me in the way she does is still one aspect of the flaw in our mutual recognition. As such it is still regrettable. Consider also that it is impossible for a friend who acts paternalistically towards me to avoid implying any judgment about my competence: it is precisely because she worries that I am likely to make the wrong decisions that she intervenes in the way that she does. And while friends must no doubt be able to draw attention sensitively to each other’s faults without damaging their relationship, such forays must be delicately handled, as most of us will recognize. Often I simply do not want to know that my friend – of all people! – thinks of me that way.⁷⁷

Contrast the position of the state in these respects. For most paternalistic policies, the state does not make a presumption about any individual. It takes the view that many people’s lives will be improved by the intervention on account of some failure of rationality or willpower of theirs. But it does not single anyone out as the target. It need not point out to anyone the deficiencies in their own competence; indeed it need not (and, in any case, cannot) make any specific judgments about which individuals under its jurisdiction are likely to exhibit the rational or volitional deficiencies that justify the paternalistic intervention. No similarly problematic element is introduced into the relationship between the citizen and the state as a result of a paternalistic policy in the way that the paternalistic incentive affects the dynamic of a friendship.

These features of state paternalism, I claim, give it a pro tanto advantage over individual paternalism. In the title of their recent book, Bill New and Julian Le Grand ask whether we should think of government paternalism as a ‘nanny state’ or a ‘helpful friend’.⁷⁸ But in the best-case scenario it is neither, at least where general legislation is concerned. Both nannies and friends are individuals who take responsibility for, or show concern for, people they know intimately, and their interventions are inevitably complicated by being situated in the context of such intimate relationships. The state might not be able to work on the basis of the detailed picture that friends and nannies have can, but we, its citizens, are better off for that. In its ability to advance people’s interests while remaining at a distance, the state is a more appealing paternalist than either.

THE CHALLENGE OF STATE PATERNALISM

I have argued that what is sometimes called the ‘faceless bureaucracy’ of the modern state gives it a pro tanto advantage over individuals when it comes to paternalism. But while the state operates for the most part at a distance from us, it does not work by magic. The activities of the state are ultimately the actions of individuals. Paternalistic policies are the work of legislators. State paternalism is also government paternalism. The state is only able to take advantage of the arm’s-length nature of the state-citizen relationship if citizens believe that the individuals who comprise the government and the wider machinery of state do not in fact look down on them. If we suspect that the government does hold us in contempt, then the advantage of its arm’s-length operation will be cancelled. Although its paternalistic policies may still improve people’s lives, they will also reinforce a morally objectionable relationship between the governing and the governed.

It will not turn government paternalism into a version of individual paternalism, of course. The paternalizee remains anonymous to the paternalizer even when we start to think of the relationship being between those who govern and those who are governed rather than between state and citizen. But when it is no longer the abstract state arranging the circumstances in which we act, we nevertheless begin to feel the judgment of the human individuals involved.

It is sometimes pointed out that the relationship between state and citizen is not a relationship of equals. After all, the state makes a claim to comprehensive authority and has unparalleled power with which to back up that claim. These are among the basic facts of politics. They constitute the ground of our empirical concerns for our own vulnerability and our normative concerns about the state’s legitimacy. But the flipside of this is that the state is not a peer. We don’t relate to it in the way that we relate to other individuals. Where paternalism is concerned, I have argued, this is an advantage. But things are otherwise when the relationship that is most salient to us in this regard is the (would-be) peer-to-peer relationship between us and those who govern us, rather than the more abstract relationship between state and citizen. Without the mitigating features or the compensating benefits of close personal relationships such as friendships to help us out, paternalism may strike us as no more than a manifestation of contempt.

Much of what I have said in favour of state paternalism depends on citizens having a certain attitude towards the state. Some may think that the attitude I have sketched is unjustifiable. They may say that we should always remain more circumspect about the activities of the state. In response let me first reiterate that the case I offer here for state paternalism is limited. It is compatible with deep scepticism about the state’s epistemic position, operational competence, and practical authority. If the state’s actions outstrip its limitations in these areas, they will not be justified, whatever the attitudes of its agents. As for the question of motivation, the danger of contempt – and the very question of paternalism – only arises if the government acts with a view to improving people’s lives.

More importantly, the justifiability of the attitude to paternalism that I have advocated depends on the quality of government; and not just on the government’s making the right policy decisions, but in its having – and communicating – the right attitude towards its citizens. The case for state paternalism also depends on the government having a certain self-image. Respect-based criticisms of state paternalism often invoke precisely the objectionable attitude towards citizens that I have described here – of presumed social superiors taking control of the wayward lives of inadequate inferiors. The kind of respect that legislators and ministers ought to have for their citizens does not require them to ascribe to citizens rational or volitional capabilities that they do not in fact possess. It does require them to regard citizens as their social equals. The second need not imply the first.

It may be objected that, for some, general paternalistic legislation does present itself as precisely the uncalled-for intervention of personally identifiable supercilious meddlers that I have claimed it is not. But if I am right that such paternalism is not intrinsically disrespectful, then if its benefits are not outweighed by the cost to people’s autonomy, and if other concerns about the state’s authority to enact such measures can be addressed, then, I claim, we should not see state paternalism as such. On the other hand, if we know, or reasonably suspect, that those who govern us are in fact supercilious meddlers, then it will not be surprising if an instinctive dislike of paternalism persists. Governments must strike a very difficult balance. On the one hand they must remain ‘faceless’ enough for paternalism to operate at a distance; on the other, they must do enough to reassure the governed that the judgment that they can do better for them does not conceal the attitude that they are better than them.