Abstract Many public debates over policies aimed at curbing alcohol consumption start from an assumption that policies should not affect ‘responsible’ drinkers. In this article, I examine this normative claim, which I call prudentialism. In the first part of the article, I argue that prudentialism is both a demanding and distinctive doctrine, which philosophers should consider seriously. In the middle sections, I examine the relationship between prudentialism and two familiar topics in public health ethics: the prevention paradox and the relationship between responsibility and solidarity. I argue that standard positions in these debates do not necessarily undermine prudentialism. In the final part of the article, I outline an alternative, more successful, argument against prudentialism: that the categories of ‘responsible’ and ‘irresponsible’ drinking behaviour are not ‘apt’ for use in policy. I show how this objection relates to Elizabeth Anderson's arguments against the more familiar doctrine of ‘luck egalitarianism’. There are over 1 million alcohol-related hospital admissions in England per year, and alcohol consumption is the leading risk factor for premature mortality and morbidity among 19- to 45-year-olds (Public Health England, 2016: 6). Alcohol is associated with many other social problems, from vandalism to domestic violence. Banning the production, sale or consumption of alcohol entirely would reduce these harms, but seems a disproportionate restriction of liberty. However, more liberal measures—such as education campaigns—are notoriously ineffective. A third option, less draconian than prohibition but more interventionist than education, is ‘Minimum Unit Pricing’ (MUP), imposition of a legal requirement that the cost of alcoholic drinks is proportional to their alcoholic content. Already trialled in Scotland, a recent report claimed that ‘policies that reduce the affordability of alcohol are the most effective, and cost-effective, approaches to prevention and health improvement’ (Public Health England, 2016: 7). In this paper, I focus on one aspect of the controversy over such policies: their effects on ‘responsible’ drinkers. In an earlier round of public debate, the British Retail Consortium attacked MUP as a ‘tax on responsible drinkers’ (Winnett, 2012). David Cameron conceded such concerns, stressing that pricing policies are ‘not about stopping responsible drinking’ (Rigby and Lucas, 2012). Who is the responsible drinker? Why, if at all, does it matter whether policy affects him/her? To put matters more formally, some opponents and proponents of MUP seem to agree on ‘prudentialism’: All else being equal, public policy aimed at changing individual behaviour may disadvantage those who engage in that behaviour in an ‘irresponsible’ manner, but should not disadvantage those who engage in that behaviour in a ‘responsible’ manner. This article asks whether we should endorse prudentialism in the context of alcohol policy. Ultimately, I argue we should not. Prudentialism is politically popular: opponents of ‘sin taxes’ claim they would punish those who indulge sensibly; outside public health policy, low interest rates punish the ‘prudent saver’. However, philosophers have typically ignored such claims. Therefore, the ‘The Non-Trivial Nature of Prudentialism’ section develops what I take to be the most defensible version of prudentialism. This may seem rather odd: Why build up an unfamiliar theoretical position only to knock it back down? I take it that philosophers have rarely discussed prudentialism because it seems, at worst, a piece of rhetoric, and, at best, a mangled version of more familiar concerns. However, even if invocations of prudentialism are often dog whistles to thirsty or hungry voters, this does not mean that they should be rejected out of hand; by analogy, objections to the Nanny State are often little more than appeals to self-interest, but that does not imply liberalism is false. Secondly, as I argue below, properly understood, prudentialism is a distinctive mid-level doctrine, which differs from familiar claims that policy should be efficient or that we should adopt ‘choice-sensitive’ distributive principles. By studying this doctrine, we gain a better understanding of more familiar debates over such topics as the prevention paradox (‘Prudentialism and Population Health’ section), the relationship between responsibility and solidarity (‘Solidarity and Responsibility’ section) and the interplay between epidemiology and ethics (‘Prudentialism and Categorisation’ section). MUP is a paradigm ‘population-based’ public health policy; prudentialism is a powerful concern in public debate; thinking through the latter as it relates to the former provides a prism for thinking through alcohol policy specifically and public health ethics more generally. The Non-Trivial Nature of Prudentialism My aim in this section is simple: to construct an argument in favour of prudentialism in public health policy. I have defined ‘prudentialism’ as the claim that, all else being equal, policies may target the irresponsible, but not the responsible. There are various ways of spelling out ‘responsible’. First, a ‘strict anti-paternalist’ interpretation: behaviour is ‘irresponsible’ when it poses significant risks of material harm to non-consenting third parties; it is ‘responsible’ when it does not. Secondly, what I call a ‘health consequentialist’ interpretation: behaviour is ‘irresponsible’ when it poses either the first kind of risk or when it poses a risk to the individual’s own future health. The ‘strict anti-paternalist’ and the ‘health consequentialist’ might agree that someone who, each evening, drinks before driving home behaves ‘irresponsibly’, but disagree over whether someone who drinks heavily but always walks home drinks ‘irresponsibly’. The sense of ‘responsible’ used in prudentialism, I claim, lies between these two interpretations. For the prudentialist, a behaviour is irresponsible when it is ‘imprudent’: it poses either a risk of material harm to others or a significant, foreseeable risk to the individual’s own future health. While the ‘prudentialist’ might agree with the health consequentialist that the heavy drinking walker drinks ‘irresponsibly’, she might disagree that someone who regularly enjoys a tipple at the end of a stressful day—hence slightly increasing her chance of alcohol-related morbidity—drinks irresponsibly. This sense of ‘responsible’ tracks public debate. On the one hand, the category of the ‘irresponsible drinker’—whose behaviour it is permissible to engineer—seems to include drinkers whose behaviour poses risks to their own health, rather than to others. When David Cameron stressed that over 10 million people were drinking ‘dangerously’ (Rigby and Lucas 2012), he stretched the category of the ‘irresponsible’ beyond those who threaten disorder or domestic violence. On the other hand, despite the mounting evidence that even low levels of consumption increase alcohol-related morbidity risk (Public Health England, 2016: 28–30), no one holds that all alcohol consumption is ‘irresponsible’. I suggest that this would persist even if it were conclusively proven that there is no ‘safe’ level of consumption; the responsible drinker balances risks against benefits, rather than being risk-free. Prudentialism seems to face a challenge: either it is too paternalistic or it is insufficiently paternalistic. A standard objection to MUP is that it is paternalistic because it diminishes drinkers’ ‘opportunity to choose’ to drink, on the basis that this is in their self-interest.1 A prudentialist seems unable to object to MUP on such grounds; the principle apparently licenses action against some drinkers—the ‘irresponsible’—whose drinking does not threaten harm to non-consenting others. Even if the prudentialist avoids this charge, she faces a second challenge: she is insufficiently paternalist. It is clearly possible that an effective public health intervention, such as MUP, might affect both ‘responsible’ and ‘irresponsible’ drinkers. It seems, however, that the prudentialist must reject such effective paternalistic policies. In the rest of this section, I will argue that, rather than being overly paternalistic, prudentialism can best be understood as based on the same underlying moral concerns which typically motivate anti-paternalism (I return to efficiency concerns in the ‘Prudentialism and Population Health’ section). To frame my discussion, consider a second aspect of public debate over policies such as MUP: the use of anticipated ‘direct savings’ to health services as a justification for policy. We might assume that the underlying justification for adopting preventive health policies is their effect on overall population health. However, arguments for MUP often consider a second set of outcomes: its likely effect on NHS expenditure. For example, the recent PHE report stressed that introducing MUP at 60 p per unit would ‘save’ c.£250 million a year on ‘direct’ healthcare costs (PHE, 2016: 96). Such measures are not mere proxies for decreased mortality and morbidity. The aggregate benefits of MUP appealed to by PHE included both ‘QALY gains’ and ‘direct healthcare costs’; either this was ‘double counting’ or the second category has independent normative weight. I will now sketch one possible justification for treating ‘savings’ as a relevant outcome for justifying alcohol policy in general, before turning to how these arguments relate to prudentialism specifically. Imagine that you believe (correctly) that you have a strong obligation to help out your neighbour, Bill, if he is in need. Bill, aware of this fact, gambles heavily. As such, it is likely that you will soon have no morally permissible option but to pay his bills. Plausibly, you have a legitimate interest in changing Bill’s behaviour, even if other moral rights and duties limit what you can do. (By contrast, you may lack any legitimate interest in changing the behaviour of a second neighbour, Ben, who also gambles, but whom you are not obliged to help out.) A concern to minimize the future costs of meeting one’s obligations is not to deny the moral weight of those obligations. Rather, it is to take them seriously: it is precisely because you know that morality demands that you must do something personally costly in some situation that you seek to avoid that situation arising.2 Assume that institutions of socialized healthcare, such as the National Health Service, can be justified—at least partly—as systems which ensure that each complies with a basic moral obligation: to help those in serious medical need. [As Buchanan (1984) has it, nationalized healthcare enforces an obligation of beneficence.]3 Let us also assume, in turn, that each contributor to that system has a legitimate interest in reducing the costs of meeting their obligations. This concern provides a ground for preventative health policies, reflected in concerns about ‘savings’. In support of this normative understanding of preventative health policies, consider a puzzling feature of public debates: that they often focus on the provision of healthcare although other socially controllable factors make a much larger contribution to overall health outcomes (Segall, 2007). My model explains this asymmetry: the obligation to provide care to the needy, rather than a more generalized obligation to promote good health, lies at the centre of our normative thinking about health policy. In turn, this understanding of preventative health can motivate a non-paternalist account of prudentialism’s first clause: that it is permissible to change the behaviour of the ‘irresponsible’ ‘self-harmer’. Consumption patterns which do not pose direct risks of physical harm to (non-consenting) others may still pose a risk of financial loss to them. In turn, such risks can justify otherwise ‘paternalistic interventions’—it is not ‘paternalistic’ to stop you taking my money. The ‘loss’ we suffer through providing care to others may, however, seem very different from the ‘loss’ we suffer when others steal our money. The former ‘loss’ seems mediated by a choice of the ‘loser’—to provide care at all—hence rendering such losses ‘consensual’, in a way in which the latter are not. (Note that here I place to one side issues around coercion and taxation.) However, this difference in the causal mechanisms of loss may not be morally relevant. Rather, as long as the relevant ‘loss’ cannot be avoided in a morally acceptable manner—it is a loss we must suffer if we are to meet our obligations—then it might be ‘non-consensual’ (in the morally relevant sense). Prudentialism is not strictly paternalistic because it does not licence interference in ‘irresponsible’ individuals’ behaviour for the sake of preventing harm to themselves; rather, it justifies such interventions in terms of the interests of ‘non-consenting’ third parties, who risk paying a morally unavoidable cost for others’ behaviour. On this approach, other alcohol policies, such as Sabbatarian anti-drinking laws (Mill, 1989: 90), remain unjustifiable because the relevant ‘costs’ are not imposed on third parties in virtue of their basic moral obligations. Students often claim that apparently ‘paternalist’ interventions—such as a ban on smoking—are not in fact paternalist because smokers impose a ‘cost’ on others, via taxation; my arguments spell out this common undergraduate argument for behaviour change policies! (Of course, it may well be false that smokers, who typically die younger than non-smokers, do impose net costs on others: that is a different issue!) The objection that prudentialism is overly paternalistic is misguided. However, the ‘prudent’ drinker’s behaviour might also pose some risk of alcohol-related morbidity, and, hence, impose morally inescapable costs on others. Why treat his/her behaviour as outside the scope of legitimate interference? To address this challenge, return to the example above. Imagine that you have a third neighbour, Bob. As with Bill, you believe (correctly) that you have a strong obligation to help Bob if he is in need. Bob is not a compulsive gambler, but enjoys the occasional flutter. As a matter of fact, unbeknown to Bob, there is a chance that one of these bets will go horribly wrong, leaving Bob destitute. As such, you also have some interest in changing Bob’s behaviour, much as you do in changing Bill’s behaviour. However, it may also seem that there are things you may legitimately do in response to the ‘threat’ posed by ‘irresponsible’ Bill, which would be an improper response to the ‘threat’ posed by ‘responsible’ Bob. Broadly, there are three arguments for a normatively significant difference between Bill and Bob. First, one could argue that simply because the risks the former impose are higher than those the latter impose, different responses are legitimate. In general, the greater the risk of physical harm I impose on you, the more you may legitimately do to stop me; a similar principle might apply when the risks are of financial loss, contingent on meeting my obligations. Secondly, one could argue that the size of the risk is not key, but, rather, the attitudes expressed by their behaviour: Bill seems to be taking advantage of the fact that I am under an obligation to help him in a way in which Bob does not. As such, Bill renders himself liable to response but Bob does not, much as those who intentionally place me at risk of harm render themselves liable to a wider range of permissible responses than those who unintentionally place me at risk. Thirdly, there may be long-term (broadly) consequentialist reasons to encourage Bob-like behaviour and discourage Bill-like behaviour. Consider, again, the relationship between nationalized healthcare and preventative medicine. For any of the three reasons above, there may be a moral difference between people who choose to act in ways which they can reasonably assume will significantly increase the chances that they will require care, and others who do not. The prudentialist claims that public health policies, aimed at reducing the expected costs of meeting our obligations of care may aim to change ‘irresponsible’, but not ‘responsible’, behaviour. For reasons I explain in the ‘Prudentialism and Categorization’ section, I do not think prudentialism is, all things considered, defensible. My aim in this section, however, has been constructive: to show how we might generate a ‘prudentialist’ principle via an obligation of beneficence without thereby endorsing paternalism. Of course, these arguments are disputable, and they may only hold in systems where medical care is socialized. However, they imply that prudentialism is worthy of attention. In the next two sections, I develop this position by showing how it relates to two more familiar debates: over the prevention paradox, and over the relationship between responsibility and solidarity. Prudentialism and Population Health In this section, I consider the objection that prudentialism is insufficiently paternalistic. Policy-makers seem to disagree over whether or not MUP would burden ‘responsible’ drinkers. The recent Public Health England report claims it would have little effect on ‘moderate’ drinkers (PHE, 2016: 7), but the Scottish government has framed MUP as tackling ‘population-level consumption’ (Katikireddi et al. 2014). Note, however, that both ‘sides’ to this debate agree that even ‘moderate’ alcohol consumption poses some risk of morbidity or mortality. To assess this disagreement, consider Geoffrey Rose’s ‘fundamental axiom’ of preventive medicine: ‘a large number of people exposed to a small risk may generate many more cases [of disease] than a small number exposed to a large risk’ (Rose, 2008: 59). For example, even if women with both BRCA1 and BRCA2 mutations are at higher risk of breast cancer than women between the ages of 40 and 70, many more cases of breast cancer occur among the second group than the first, because the second group is much larger. Typically, ‘population strategies’, which target the ‘moderate risk’ mass of the population, will have a greater effect on overall population health than strategies targeting ‘high risk’ populations. Given Rose’s arguments it seems likely the Scottish government is closer to the truth: if MUP is an effective means of reducing population-level harms, then it will affect both moderate (‘responsible’) and high-risk (‘irresponsible’) drinkers. Assuming this analysis is correct, then prudentialism implies we should reject MUP, despite the fact that such a policy would predictably have positive population-level health outcomes. (Of course, strictly, prudentialism has a ceteris paribus clause, but I assume that if the principle applies anywhere, it applies to such quotidian cases as pricing policy.) From the viewpoint of maximizing population-level outcomes, prudentialism seems to place extremely strong limits on public health policy, by requiring that we choose policies which target ‘high risk’ populations, even when more good would be done by population strategies, which affect both ‘high’ and ‘moderate’ risk populations. (A proponent of MUP might note, further, that increasing the cost of alcohol is not the kind of intervention which typically exercises anti-paternalists; it does not remove or infringe anyone’s rights, to involve coercion or manipulation.) It might, then, seem that prudentialism is morally problematic. This section responds to this concern. Rose’s ‘fundamental axiom’ implies that targeting individuals at a small risk of disease may often be the best way to reduce the population-level incidence of disease. However, as Rose himself noted, the implications of his fundamental axiom can seem paradoxical. One problem with ‘population policies’, which Stephen John (2014) calls the ‘absolute prevention paradox’, is that they can seem to burden each individual agent more than they benefit him/her, even if their expected impact on overall population health is positive. For example, imagine that taking a cholesterol-inhibitor reduces cardiovascular risk for men over 60 by 1 per cent. Although a small absolute risk reduction, we can reasonably expect that if every man over 60 took this drug, there would be a major reduction in overall cardiovascular morbidity. Plausibly, we might judge that this population-level benefit morally outweighs the aggregate cost of slight annoyance felt by every man who takes the pill unnecessarily. However, it also seems possible—indeed, plausible—that each individual man might judge the ‘benefit’ of a 1 per cent risk reduction outweighed by the burden of annoyance. If each man did take the pill, each would seem to lose, while the population benefits!4 Is this a genuine paradox? Maybe not (Thompson, 2017). What it does reveal is a tension between two perspectives on normative assessments of the distributional effects of policies: one focused on expected ex post population-level benefits, and one focused on each individual’s ex ante prospects (Fleurbaey and Voorhoeve, 2013). In this case, we seem torn between two concerns: we think that ex post a world where each took a pill would contain more good than harm, but taking the pill does not seem in the ex ante interests of any individual. It might seem that this tension is simply an artefact of the difficulties in accounting for such things as aggregate annoyance across the population; one might think that if we properly accounted for the ‘externalities’ of the policy, we would judge the expected outcome is, in fact, worse than the status quo. However, there are two problems with this response. First, many philosophers reject strict aggregative reasoning (i.e. that lots of small sums of annoyance can add up to outweigh lives saved), but denying aggregation seems compatible with thinking that the policy is problematic because, ex ante, it burdens each more than it benefits her (John, 2014). Secondly, similar problems arise regardless of whether we focus on aggregation or other distributive principles. For a non-prudentialist example in alcohol policy, consider one aspect of Tom Walker’s rich discussion of MUP (Walker, 2010). Walker assumes that MUP is supposed to be justified on paternalistic grounds. He claims, however, that even if we might legitimately act paternalistically towards ‘heavy drinkers’, it is more problematic to act paternalistically towards ‘moderate drinkers’. Furthermore, he argues that such problematic treatment would be unequally distributed; MUP would have a significant effect on the consumption of poorer, ‘moderate’ drinkers, but not richer drinkers. Walker’s arguments appeal, then, to a concern that MUP would lead to a kind of inequality. However, proponents of MUP also appeal to egalitarian concerns: they argue that, because alcohol-related mortality and morbidity rates are higher in lower socio-economic groups, it would narrow health outcome inequalities (PHE, 2016: 95). In the disagreement between Walker and proponents of MUP, both sides appeal to egalitarian, rather than aggregative, concerns. Where the disagreement lies is in whether we should think of equality in ex ante or in ex post terms. Which perspective is more fundamental? I will not resolve this debate here.5 What is clear, however, is that sometimes ex ante concerns can seem morally relevant. Analogously, the prudentialist can concede that, judged from the expected ex post population perspective, population strategies have better overall outcomes than targeted strategies, even taking account of the fact that such policies require a much greater loss in the pleasures of intoxication. What she stresses, however, is that this perspective is not the only morally relevant perspective on alcohol policy; rather, we should also consider the fact that the effects of such policies on the ex ante prospects of many drinkers—those who are responsible—are negative. In turn, then, she can appeal to the arguments outlined in the ‘The Non-Trivial Nature of Prudentialism’ section to claim that there are good moral reasons not to impose such burdens. Intuitively, tackling heavy drinking seems the most effective way to reduce alcohol-related morbidity. However, this is not necessarily true. Rather, there is a potential gap between the categories ‘irresponsible drinker’ and ‘drinker whose behaviour we should change if we wish to promote best outcomes’. In adopting the former category as the starting point for normative analysis of policy, the prudentialist is not making an epidemiological error, but, rather, expressing a preference for the ex ante over the ex post perspective. Unless we wish to simply dismiss other arguments in this arena—such as Walker’s egalitarian concerns—we should take this preference seriously. Solidarity and Responsibility Should we let uninsured motorcyclists die by the roadside? Responsibility-sensitive accounts of justice—such as so-called ‘luck egalitarianism’—imply that doing so would not be unjust. Some object that this conflicts with ‘solidarity’, central to the functioning of nationalized healthcare systems (Thompson et al., 2003). Prudentialism, like luck egalitarianism, invokes a concept of solidarity. As such, it seems that it, too, might be subject to a ‘solidarity critique’. In this section, I argue against this possibility. In doing so, I clarify the concept of prudentialism, broader debates over the relationship between responsibility, healthcare and solidarity more generally, and set up the final section, where I present a related, but more general, concern about how prudentialism categorizes sub-populations. There are three reasons why even if luck egalitarianism, and similar responsibility-catering accounts of justice, are subject to a solidarity critique, prudentialism is not. First, the concept of responsibility is used in multiple ways in health policy; for example, Norman Daniels has argued that the ‘luck egalitarian’ cannot justify policies which promote individuals’ ‘responsibility’ for their own health.6 Similarly, the concept of ‘responsibility’ at play in ‘prudentialism’ differs from that used in debates over ‘responsibility-sensitive’ healthcare distribution. The former concept picks out a pattern of behaviour—one which does not pose significant risks of harm on others—whereas the latter refers to a relationship between an agent and an outcome. Secondly, as the ‘Prudentialism and Population Health’ section showed, there is a general difference between thinking about policy from the ex ante perspective and thinking about which distributions would be just ex post; arguably, many claims about the importance of ‘responsibility’ in healthcare focus on the latter question—which distribution of goods would be fitting—rather than, as with prudentialism, the former. Prudentialism is not a restatement of the claim that healthcare distribution should be responsibility-sensitive. Thirdly, and most importantly, my proposed justification in the ‘The Non-Trivial Nature of Prudentialism’ section implies that prudentialism is not incompatible with the value of solidarity, understood as ‘shared practices reflecting a collective commitment to carry “costs” (financial, social, emotional or otherwise) to assist others’ (Prainsack and Buyx, 2011: 59). Rather, it rests upon such a commitment. The ‘savings justification’ for preventative health policies starts from an obligation to treat the uninsured motorcyclist should she have an accident. What it adds is a justification for requiring insurance in the first place, grounded in citizens’ legitimate interest in avoiding the cost of meeting this obligation. Of course, one could hold the ‘savings justification’, but deny ‘prudentialism’. However, prudentialism is parasitic on a solidaristic obligation because such an obligation is necessary to characterize the distinction between ‘responsible’ and ‘irresponsible’ behaviour. These features of ‘prudentialism’ have important implications for how we think about both responsibility and solidarity in health contexts. First, a key selling point of many luck egalitarian theories of justice is that they capture the intuition that it is unfair to treat the prudent and the imprudent the same way, incorporating ‘within [egalitarianism] the most powerful idea in the arsenal of the anti-egalitarian right: the idea of choice and responsibility’ (Cohen, 1989: 933). However, everyday public discourse around the prudent drinker has less to do with ideals of what would be cosmically just, and a more specific concern about respecting those who play ‘within the rules’. Secondly, one might have thought that even if ‘responsibility’ has multiple uses, the different uses all raise concerns about solidarity. For example, Prainsack and Buyx (2011: Chap. 8) argue that the practices of ‘responsibilitization’ discussed by Daniels, where citizens are divided into ‘at risk groups’ and encouraged to change their behaviour, may threaten solidarity. However, prudentialism does not imply that people must be made into more ‘responsible’ citizens. The principle treats responsibility as limiting policy—do not burden the ‘responsible’—rather than as a positive ideal. What is true, however, is that prudentialism splits the population into different groups: the responsible and the irresponsible. Does any such distinction undermine solidarity? Prainsack and Buyx (2011: Chap. 8) seem to suggest so in their discussion of risk stratification. However, this seems implausible. For example, inviting some groups but not others to a breast cancer screening programme does not necessarily undermine social solidarity. Indeed, inviting all women, with a concomitant rise in false positives, would probably do far more to undermine public support. What is true is that much depends on how we decide who to invite: inviting only rich women, for no reason other than that they are rich, seems problematic in a way in which inviting only women between 40 and 70, on the grounds that the expected ‘benefits’ of screening this group outweigh the expected ‘costs’, does not. Are prudentialist categorizations of behaviour as responsible or irresponsible likely to undermine solidarity? To a large extent, this can only be answered empirically. Distinguishing between those who knowingly impose on solidaristic systems and those who do not may do more to support, than undermine, support for those systems, given their reliance on norms of reciprocity (Buyx, 2008). Even if the solidarity critique is not fatal to prudentialism, it points us towards an important question, which the next section addresses: whether the categories of ‘responsible’ and ‘irresponsible’ are ‘apt’ for use in policy (John, 2013).This general question can be separated from more specific concerns about solidarity; for example, egalitarians might equally well worry about screening only the wealthy. Prudentialism and Categorization Some writers apparently assume that health policy should, insofar as possible, be undergirded by ‘naturalistic’ categories; for example, that we should use a ‘naturalistic’ account of health and disease to demarcate between clinical services the State is obligated to provide—treatments—and those it is not—enhancements (Daniels, 2000). One possible argument against treating the categories of ‘responsible’ and ‘irresponsible’ drinking behaviour as inapt is that they do not necessarily track underlying epidemiological natural kinds. To explain: either, as many epidemiologists now argue, all alcohol consumption raises morbidity risk slightly or it does not. According to the position sketched in earlier in this paper these arguments are largely irrelevant to distinguishing between responsible and irresponsible behaviour. Even if all alcohol consumption increases morbidity risk, it does not follow that all alcohol consumption is imprudent in the sense of foreseeably imposing a significant risk to the health of self or others. By analogy, the fact that any bank might go bust does not show that there is no such thing as the ‘prudent’ saver. Indeed, even if there is some ‘safe’ level of alcohol consumption—i.e. a level of consumption below which there is no risk of harm—it does not follow that this level of consumption marks the upper bound of ‘responsible’ drinking. However, there are at least two good reasons to think that the mere fact that some categorization is shaped by moral and political concerns does not show that it is inapt for use in policy. First, there are multiple, equally ‘scientifically legitimate’ ways of carving up the social world, such that it is unclear that any one approach is privileged (Ludwig, 2016). Secondly, there may be ways of categorizing the world which are not employed within any ‘purely’ scientific categorization scheme, but which are still extremely important for moral or political purposes. Even if racial categorizations do not track biological reality, we might still be concerned with racial injustice. To decide when some category is ‘apt’ requires, then, that we ask whether it is normatively acceptable. How, though, should we make such judgments? Even though ‘prudentialism’ is not merely a confused form of ‘luck egalitarianism’, Elizabeth Anderson’s objections to the latter theory provide a useful starting point for understanding these issues (Anderson, 1999). We have good reason to believe that people who are not conventionally attractive do less well than the attractive along a series of important parameters. Plausibly, very few people are responsible for their degree of attractiveness. Therefore, luck egalitarians, committed to the claim that we should compensate people who suffer from ‘unchosen’ misfortune, seem committed to ‘compensating’ the ugly; for example, in Anderson’s example, through tax rebates! There are multiple ways of reading this example: most commonly, it is cited in discussions of worries that luck egalitarianism would seem to require ‘shameful self-revelations’ (Wolff, 2010). However, I will focus on a less familiar way of reading Anderson’s concern: the causal path from having a particular physiognomy to various kinds of social disadvantage is mediated by various social norms which are, themselves, problematic from the viewpoint of social justice. For example, ‘beauty norms’ may implicitly reflect sexist views about proper gender roles. Even if there is a legitimate general principle of justice according to which we should rectify unearned disadvantage, we should not respect these concerns by using categories such as ‘ugly’, because doing so ignores, endorses or perpetuates norms which are independently problematic from the viewpoint of social justice. Regardless of the underlying normative theory, the actual system of categorization required for its implementation is not ‘apt’. (Similar concerns might be relevant in cases of racial inequality; when certain inequalities arise because of how a dominant group ‘racializes’ another group, there may be some argument for rectification. However, using the dominant groups racial taxonomies to implement the independently plausible account of justice may be problematic if doing so simply perpetuates historical injustice.) Consider, then, the case of ‘prudentialism’. Two factual claims about ‘responsible’ and ‘irresponsible’ behaviour are uncontroversial. First, individuals’ choices to behave in particular ways are mediated and shaped by various social norms and social institutions; it may be much easier to behave ‘responsibly’ given certain socially defined roles, resources and so on, than others. Secondly, any attempt to capture the underlying facts about who is ‘responsible’ or ‘irresponsible’, in the normatively relevant sense, is likely to be shaped by background social norms and institutions. For example, in societies with a long Protestant tradition, ascriptions of behaviour as ‘responsible’ may track the ‘Christian virtues’ of ‘hardwork, thrift and will power’ (Grant 2012: 19), rather than imposition of risks on others. In-and-of-themselves, neither of these claims show that the concepts of ‘responsible’ and ‘irresponsible’ are inapt for use in policy; all behaviour is shaped by norms and institutions, and any attempt to capture normative phenomena will be filtered through local values. What they do suggest is that we cannot move smoothly from the thought that prudentialist concerns might in general express normatively relevant concerns to a willingness to endorse any particular prudentialist claim. Rather, a division of behaviours into ‘responsible’ and ‘irresponsible’ may rest on or assume politically problematic social practices or norms. In the specific context of alcohol policy, there are at least two reasons to be suspicious of how powerful social actors divide drinkers into the ‘responsible’ and ‘irresponsible’ categories. First, within many—although, of course, not all—sub-communities within the UK, alcohol consumption is tied to many other aspects of communal life (Ally et al., 2016). In turn, social norms in some communities are such that high levels of alcohol consumption may be important as a way of accessing other goods. For example, drinking to excess may be an important aspect of taking part in social gatherings, and, hence, social life or the job market. Of course, these social norms can be extremely harmful, in that they do, often, lead to harm to self and others. However, the relevant behaviour may be sensible, or all-things-considered rational, when these norms are in place. Treating excessive consumption as if it were an individual ‘failing’, rather than a rational response to social circumstances, may not only be ineffective (the underlying norms may remain in place); more fundamentally, it risks disrespecting citizens, by ignoring the reality of their situation. Secondly, designations of individuals as ‘responsible’ or ‘irresponsible’ drinkers are shaped by gender or class-based norms which may, themselves, be problematic from the viewpoint of social justice. For example, even when a male middle manager and a working-class ‘ladette’ consume in similar proportions, the former’s behaviour is typically treated as a form of ‘well-earned relaxation’, and, hence, ‘responsible’, whereas the latter’s consumption is treated as showing a ‘irresponsible’ lack of self-respect (Atkinson et al., 2011). In turn, the underlying social norms and expectations which drive these different judgments—a valourization of white-collar work, or a fear of female sexuality—should themselves be challenged, rather than (implicitly) endorsed.7 Note that both this concern and the previous concern are not, necessarily, concerns about the effects of the categorization on those who are categorized. As such, they differ from Voigt’s (2013) worries that the language of responsibility might lead to negative effects for individuals, for example, by making them feel stigmatized, or ashamed, when they cannot act responsibly. Rather, these are concerns that to use certain sorts of categories can be to ignore or endorse norms which State actors should not ignore or endorse. By analogy, it need not be the case that every woman is harmed by an image of women as homemakers for us to be concerned that State policies should not ignore or endorse this way of viewing women. I suggest, then, that the ways in which we categorize alcohol consumption as responsible or irresponsible are not ‘apt’ for use in policy, and, as such, we should reject prudentialism in alcohol policy. This may seem rather a weak conclusion. After all, even if many descriptions of behaviour as responsible or irresponsible in political debate or the media are problematic, it is not necessarily true that all ways of distinguishing behaviour must fall into the traps above. For example, respected bodies issue guidelines about ‘healthy’ alcohol consumption; might these not provide some objective standard by which to distinguish between the ‘responsible’ and the ‘irresponsible’ drinker? Of course, guidelines for alcohol consumption might provide a way of sorting drinkers into the responsible and the irresponsible. In turn, some of these results may be surprising; the middle-class manager might turn out to be more irresponsible than the ladette. Therefore, using guidelines might seem to avoid the problems of ignoring or endorsing problematic norms. However, any set of guidelines are, at best, a kind of advice, indexed to the average person and encoding expectations about which risks it is ‘reasonable’ to take with one’s health, the role which alcohol consumption ‘should’ play within a life, and so on. In describing individuals who fail to take such advice as ‘irresponsible’, we risk overlooking these epistemological and ethical complexities. Although guidelines allow us to apply the categories of ‘responsible’ and ‘irresponsible’ with a degree of mechanical objectivity, this does not imply that the results are necessarily apt; rather, we need to look at how guidelines are made. Of course, these comments do not show that actual guidelines must be flawed, still less that there can be no ‘normatively apt’ way of drawing the distinction between ‘responsible’ and ‘irresponsible’ drinking behaviour. My claim, rather, is that, given the complex roles which social norms and expectations play in how we think about alcohol, it is difficult to see how we might extract any potential nuggets of truth in prudentialism without ignoring or endorsing broader, potentially problematic social norms. Conclusions Should MUP be introduced? I do not know. What I do suggest is that we should not use the categories of ‘responsible’ or ‘irresponsible’ drinking to debate its introduction. This is not because prudentialism is clearly wrong-headed. Rather, as the sections ‘The Non-Trivial Nature of Prudentialism’, ‘Prudentialism and Population Health’ and ‘Solidarity and Responsibility’ argued, such concerns can be grounded in a complex account of citizens’ interests in reducing the expected costs of meeting their solidaristic obligations. Instead, I have argued that, as a matter of fact, designations of behaviour as responsible or irresponsible are inapt—their use ignores or endorses problematic norms—and there is no easy way around these problems. This conclusion leaves open further issues: whether the problems with prudentialism in alcohol policy can be solved; whether prudentialism in other domains of public health policy falls prey to similar problems; and whether the ‘The Non-Trivial Nature of Prudentialism’ section’s arguments for prudentialism live up to sustained scrutiny at all. What, then, have I shown? First, we should not simply dismiss prudentialist claims as bad epidemiology or some kind of proto-luck egalitarianism. Such claims are richer than they first appear, and thinking through them clarifies the limits on population thinking and the relationship between responsibility and solidarity. Secondly, public health ethics gains from working bottom-up, rather than top-down; starting from the rhetoric of debate, rather than familiar theories. I do not think we should be prudentialists, but we gain much from thinking about why not. Acknowledgements An earlier version of this article was presented at UCL in January 2016. I am particularly grateful to Emily McTernan, Martin Wilkinson, Gabriele Badano and James Wilson for their useful feedback. Footnotes 1. I borrow the claim that changes to ‘opportunity to choose’ constitutes paternalism from Archard (1990: 36); of course, this definition is controversial. Still, even if one denied that MUP is paternalist avant la lettre, it seems paternalist in the loose sense of treating adults’ choices as levers to be manipulated for their own well-being. 2. For a related discussion, see John (2015). 3. See Kelleher (2014) for an argument that the relevant obligation may be one of justice, not beneficence: nothing which follows turns much on this issue. 4. This example is inspired by debates over the use of statins, although that case is complicated both by ongoing arguments over the size of the risk reduction, and by the controversy over whether statin use can cause diabetes. 5. In favour of the ex post perspective, see Fleurbaey and Voorhoeve (2013); in favour of the ex ante perspective, see John (2014) and Frick (2016). 6. For further discussion, see Voigt, 2013; although note that Wikler (2006: 111) seems to treat the different senses as closely related. 7. For related concerns, albeit expressed in terms of ‘arbitrariness’, see Wikler (2006: 128–9). References Ally A. K., Lovatt M., Meier P. S., Brennan A., Holmes J. ( 2016). Developing a Social Practice-Based Typology of British Drinking Culture in 2009–2011: Implications for Alcohol Policy Analysis. Addiction , 111, 1568– 1579. Google Scholar CrossRef Search ADS PubMed Anderson E. ( 1999). What Is the Point of Equality?. Ethics , 109, 287– 337. Google Scholar CrossRef Search ADS Archard D. ( 1990). Paternalism Defined. Analysis , 50, 36– 42. Google Scholar CrossRef Search ADS Atkinson A., Elliot G., Bellis M., Sumnall H. ( 2011) Young People, Alcohol and the Media . London: Joseph Rowntree Foundation. ebook only, available from: http://www.ias.org.uk/uploads/pdf/Underage%20drinking%20docs/young-people-alcohol-mediaEBOOK.pdf [accessed 20 October 2017]. Buchanan A. ( 1984). The Right to a Decent Minimum of Health Care. Philosophy and Public Affairs , 13, 55– 78. Google Scholar PubMed Buyx A. ( 2008). Personal Responsibility for Health as a Rationing Criterion: Why We Don’t Like it and Why Maybe We Should. Journal of Medical Ethics , 34, 871– 874. Google Scholar CrossRef Search ADS PubMed Cohen G. A. ( 1989). On the Currency of Egalitarian Justice. Ethics , 99, 906– 944. Google Scholar CrossRef Search ADS Daniels N. ( 2000). Normal Functioning and the Treatment-Enhancement Distinction. Cambridge Quarterly of Healthcare Ethics , 99, 309– 322. Daniels N. ( 2011) Individual and Social Responsibility for Health. In Knight C., Stemplowska Z. (eds), Responsibility and Distributive Justice . Oxford: Oxford University Press. Fleurbaey M., Voorhoeve A. ( 2013). Decide As You Would with Full Information. In Eyal N. (ed.), Inequalities in Health: Concepts, Measures, and Ethics Oxford: Oxford University Press. Frick J. ( 2016). Identified versus Statistical Lives. Philosophy and Public Affairs . Grant R. ( 2012) Strings Attached: Untangling the Ethics of Incentives . Princeton, NJ: Princeton University Press. John S. D. ( 2013) Cancer Screening, Risk Stratification and the Ethics of Apt Categorisation. In Ethics in Public Health and Health Policy . Netherlands: Springer, pp. 141– 152. Google Scholar CrossRef Search ADS John S. D. ( 2015). Efficiency, Responsibility and Disability. Politics, Philosophy and Economics , 14, 3. Google Scholar CrossRef Search ADS John S. D. ( 2014). Risk, Contractualism, and Rose's Prevention Paradox. Social Theory and Practice , 40, 28– 50. Google Scholar CrossRef Search ADS Kelleher P. ( 2014). Beneficence, Justice and Health Care. Kennedy Institute of Ethics Journal , 24, 27– 49. Google Scholar CrossRef Search ADS PubMed Katikireddi S. V., Hilton S., Bonell C., Bond L., Jenkins N. ( 2014). Understanding the Development of Minimum Unit Pricing of Alcohol in Scotland: A Qualitative Study of the Policy Process. PLoS One , 9, e91185. Available from: http://dx.doi.org/10.1371/journal.pone.0091185 [accessed 20 October 2017]. Google Scholar CrossRef Search ADS PubMed Ludwig D. ( 2016). Ontological Choices and the Value-Free Ideal. Erkenntnis , 81, 1253– 1272. Google Scholar CrossRef Search ADS Mill J. S. ( 1989). On Liberty. In Collini S (ed.), On Liberty and Other Writings . Cambridge: Cambridge University Press. Prainsack B., Buyx A. ( 2011). Solidarity: Reflections on an Emerging Concept in Bioethics London: Nuffield Council on Bioethics, London. Public Health England ( 2016). The Public Health Burden of Alcohol and the Effectiveness and Cost-Effectiveness of Alcohol Control Policies . London: Public Health England. Rigby E., Lucas L. ( 2012) Cameron Targets Binge Drinking. The Financial Times, 22nd March 2012, available from: https://www.ft.com/content/64793be0-744f-11e1-9e4d-00144feab49a [accessed 20 December 2016]. Rose G. ( 2008). The Strategy of Preventive Medicine . Oxford: Oxford University Press. Google Scholar CrossRef Search ADS Segall S. ( 2007). Is Health Care (Still) Special?. Journal of Political Philosophy , 1515, 342– 361. Google Scholar CrossRef Search ADS Segall S. ( 2007). In Solidarity with the Imprudent: A Defense of Luck Egalitarianism. Social Theory and Practice , 33, 177– 198. Google Scholar CrossRef Search ADS Thompson A., Robertson A., Upshur R. ( 2003). Public Health Ethics: Towards a Research Agenda. Acta Bioethica , 22, 157– 163. Thompson C. ( 2017). Rose's Prevention Paradox. Journal of Applied Philosophy , doi:10.1111/japp.12177, in press. Voigt K. ( 2013). Appeals to Individual Responsibility for Health. Cambridge Quarterly of Healthcare Ethics , 22, 146– 158. Google Scholar CrossRef Search ADS PubMed Walker T. ( 2010). Why We Should Not Set a Minimum Price Per Unit of Alcohol. Public Health Ethics , 33, 107– 114. Google Scholar CrossRef Search ADS Wikler D. ( 2006). Justice, Socioeconomic Status and Responsibility for Health. In Sen A., Peters F. (eds), Public Health, Ethics, and Equity . Oxford: Oxford University Press. Winnett R. ( 2012). Minimum Price for Alcohol to be Set. The Daily Telegraph, 22nd March 2012, available from: http://www.telegraph.co.uk/news/politics/david-cameron/9161781/Minimum-price-for-alcohol-to-be-set.html [accessed 20 December 2016]. Wolff J. ( 2010). Fairness, Respect and the Egalitarian Ethos Revisited. The Journal of Ethics , 1414, 335– 350. Google Scholar CrossRef Search ADS © The Author 2017. Published by Oxford University Press. Available online at www.phe.oxfordjournals.org
Public Health Ethics – Oxford University Press
Published: Apr 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.
All for just $49/month
Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.
Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.
It’s easy to organize your research with our built-in tools.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera