This article was downloaded by: [Seton Hall University] On: 27 March 2015, At: 17:22 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

The American Journal of Bioethics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uajb20

Mandatory Cancer Risk Warnings on Alcoholic Beverages: What Are the Ethical Issues? a

a

b

a

Jennie Louise , Jaklin Eliott , Ian Olver & Annette Braunack-Mayer a

University of Adelaide

b

Cancer Council Australia Published online: 18 Mar 2015.

Click for updates To cite this article: Jennie Louise, Jaklin Eliott, Ian Olver & Annette Braunack-Mayer (2015) Mandatory Cancer Risk Warnings on Alcoholic Beverages: What Are the Ethical Issues?, The American Journal of Bioethics, 15:3, 3-11, DOI: 10.1080/15265161.2014.998373 To link to this article: http://dx.doi.org/10.1080/15265161.2014.998373

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

The American Journal of Bioethics, 15(3): 3–11, 2015 Copyright © Taylor & Francis Group, LLC ISSN: 1526-5161 print / 1536-0075 online DOI: 10.1080/15265161.2014.998373

Target Article

Downloaded by [Seton Hall University] at 17:22 27 March 2015

Mandatory Cancer Risk Warnings on Alcoholic Beverages: What Are the Ethical Issues? Jennie Louise, University of Adelaide Jaklin Eliott, University of Adelaide Ian Olver, Cancer Council Australia Annette Braunack-Mayer, University of Adelaide The link between alcohol consumption and cancer is well established, but public awareness of the risk remains low. Mandated warning labels have been suggested as a way of ensuring “informed choice” about alcohol consumption. In this article we explore various ethical issues that may arise in connection with cancer warning labels on alcoholic beverages; in particular we highlight the potentially questionable autonomy of alcohol consumption decisions (either with or without labels) and consider the implications if the autonomy of drinking behavior is substantially compromised. Our discussion demonstrates the need for the various ethical issues to be considered and addressed in any decision to mandate cancer warning labels. Keywords: alcohol, cancer, ethics, public health, warning labels

Alcohol is a class-1 carcinogen (World Health Organization 1998), and alcohol consumption significantly increases the risk of certain cancers (including mouth, oropharyngeal, esophageal, liver, bowel, breast, and prostate cancer), even at low to moderate levels of consumption (Bagnardi et al. 2013; Room, Babor, and Rehm 2005; Stockwell 2006; World Cancer Research Foundation 2007). Public awareness of the link between alcohol and cancer, however, is low (Thomson, Vandenberg, and Fitzgerald 2012). This has led to recommendations that alcoholic beverages should carry mandated labels warning of the risk of cancer, often with explicit reference being made to the widespread use of warning labels on cigarettes (Room and Rehm 2011; AER Foundation; Cancer Council Australia 2013; UK Faculty of Public Health 2012; Campbell 2012; Institute of Alcohol Studies 2013; Al-hamdani 2014). There are indications that cancer warning labels would have a high degree of public support (Wilkinson and Room 2009), and may be an inexpensive and politically acceptable way to educate the public about the cancer risks associated with drinking (Stockwell 2006). This enthusiasm for warning labels, however, should be tempered by some consideration of the ethical issues, currently underexplored (Eliott and Miller 2014). Although the justification of cancer warning labels on alcohol might be perceived as a straightforward extension of that for

tobacco warning label policies, they raise some distinct ethical issues. There are undoubtedly similarities between alcohol and tobacco (especially in their public health costs), but there are also differences that may be relevant to the case for or against cancer warning labels on alcohol. For example, although smoking has previously been more ubiquitous in all areas of life than it is now, currently, drinking is deeply embedded in social practices in a way that tobacco is not; indeed, through public health strategies of denormalization in the public sphere, smoking has become a stigmatized activity (Bell et al. 2010). In addition, although there is a risk of cancer and other health problems attendant upon drinking, at low to moderate levels of consumption these risks are relatively small, and may be considered outweighed by cardioprotective benefits (acknowledging continuing debate: see Di Castelnuovo et al. 2006; Holmes et al. 2014); there are no similar benefits from smoking. Certainly, the ethical issues attendant upon cancer warning labels for alcohol overlap with those arising for tobacco warning labels and other educational health interventions. We submit that the ubiquity of drinking in most Western cultures (such that “drinking” is synonymous with alcohol consumption) and the general social acceptance of moderate levels of alcohol consumption suggest that consideration of whether we ought to implement

Address correspondence to Dr. Jennie Louise, School of Population Health, The University of Adelaide, Adelaide, SA 5005, Australia. E-mail: [email protected]

ajob 3

Downloaded by [Seton Hall University] at 17:22 27 March 2015

American Journal of Bioethics

cancer warning labels for alcoholic beverages in particular can serve as a useful lens through which the broader ethical issues (that arise for warning labels generally, and for public health interventions emphasizing education and information) may be fruitfully examined. Nonetheless, as we argue in the following, the suggestion of mandated cancer warning labels for alcohol must avoid making certain problematic assumptions about the way in which consumption decisions are made and the degree to which these are substantially autonomous. Again, alcohol is a useful example for discussion here: As we show later, it provides a good case study for ways in which autonomy may be compromised in ways that fall short of physiological dependence (addiction). In addition, the facts that the absolute risks of low-to-moderate alcohol consumption are not very high, and that there are purported offsetting health benefits, raise questions about the justifiability of emotive appeals. There are, moreover, various ambiguities regarding the suggestion that cancer warning labels be mandated for alcoholic beverages. Labels may take various forms, and may carry different informational content. This, in turn, is likely to impact upon their effectiveness; additionally, even assuming effectiveness, different kinds of warning labels may achieve this effect in different ways. For the purposes of this article, we make the simplifying assumption that what would be mandated is not just warning labels, but a particular format, content, and so on, and that the mandating of labels would be accompanied by other actions (e.g., educational campaigns, posters, websites) that aim to reinforce and enhance the message carried by the labels. However, as we discuss later, label design and content may be ethically relevant.

RATIONALES FOR MANDATED WARNING LABELS The first question we must address is that of effectiveness. This is a necessary (if not sufficient) criterion for any public health intervention to be ethically justifiable: If it is not effective, spending time and money on this intervention implies foregoing benefits that could have been achieved by spending those resources more wisely. Before assessing the effectiveness of alcohol warning labels, we must first be clear about what labels are intended to achieve. This may seem obvious: The aim is to decrease alcohol consumption, thereby ultimately reducing cancer incidence. In particular, the aim is to persuade consumers to voluntarily lower their alcohol consumption, by direct means and by changing attitudes to alcohol (so that drinking is more widely regarded as an inherently risky activity at any level). If this is the case, then the argument in favor of warning labels may be considered quite shaky, as the evidence that labels (for alcohol or tobacco) change behavior, as opposed to increasing awareness and understanding of associated risks, is weak. In the context of alcohol, at least 16 countries have instantiated compulsory warning labels on alcohol (International Center for

4 ajob

Alcohol Policies 2010), and while there is some evidence that these increase public awareness, there is little evidence that they have been effective in eliciting actual behavior change (Stockwell 2006; Wilkinson and Room 2009). We might also surmise that an alcohol warning label strategy may actually miss the mark, due to the fact that the message will be absent from many of the contexts in which most problematic drinking behavior occurs (e.g., in bars or parties where alcohol is served in glasses, which will not have the warning).1 However, it is quite common to see advocates of warning labels claim (or at least imply) that labels should be mandated in order that people can make an informed choice about their alcohol consumption. For example, the Australian Chronic Disease Prevention Alliance (2011) stated that mandated warning labels are “an opportunity to provide consumers with information on the safe use of alcohol and its potential harms at the point of sale or consumption, so they can make an ‘informed decision’ or a ‘healthy choice’ about their alcohol consumption levels.” A similar statement is made by Alcohol Concern UK (2009, 1), which referred to “the government’s own agenda of promoting choice and providing information so that individuals can make healthy decisions about their lifestyles.” It is important to understand that this is not necessarily a rationale different from that described earlier (persuading people to lower their alcohol consumption). In fact, if labels directly lower consumption, ensuring informed choice would presumably be the means by which they do so; people would voluntarily lower their consumption in response to an understanding of the risks involved in drinking. However, it is also possible that some stakeholders may have a different rationale in mind when advocating warning labels. Awareness and understanding of the risks might be seen as goals in their own right; that is, the aim may be to make sure people are aware of the risks, independently of whether this is expected to lead to lower consumption. This rationale recognizes that the information will have differing relevance and importance for different people, depending on their personal history, values and priorities, and health status. Alternatively, the thought might be that warning labels are not a way to reduce consumption directly, but rather a way to facilitate future interventions that would lower consumption. If warning labels changed public attitudes toward alcohol, so that it were seen as inherently dangerous and in need of further regulation (Wilkinson and Room 2009), this might create a political environment in which there is support for policies that are currently unacceptable to the public. We assume, in what follows, that lowered alcohol consumption (and consequent lower incidence of cancer) is the ultimate aim of warning labels, whether this is intended to be achieved directly through the labels themselves or indirectly by helping to create a cultural readiness for future interventions. The is because if awareness

1. Thanks to an anonymous reviewer for raising this point.

March, Volume 15, Number 3, 2015

Ethics of Cancer Risk Warnings on Beverages

and understanding of risks were the ultimate aim (rather than a means by which reduced consumption is to be achieved), then warning labels would be justified not in terms of eventual health benefits, but rather in terms of consumer rights—that is, that people have a right to information that is relevant to their consumption choices, and that the state must ensure that such information is provided to them. Of course, this is quite correct. However, it does not provide a justification for warning labels as a public health intervention, but rather would make it a concern for consumer protection agencies.2

Downloaded by [Seton Hall University] at 17:22 27 March 2015

CONSEQUENCES: INTENDED AND UNINTENDED If effective, an intervention aimed at reducing populationwide alcohol consumption would be expected to have numerous benefits. For example, given that over 5000 cases of cancer and approximately 2000 deaths per year in Australia are attributable to alcohol consumption (Winstanley et al. 2011), a reduction in alcohol consumption could reasonably be expected to see a reduction in cancer incidence and deaths—with attendant health care savings. Though not germane to this argument, we would also expect to see a reduction in various other (costly) harms associated with alcohol consumption, including antisocial behavior, liver disease, stroke, high blood pressure, and road injuries and fatalities (AIHW et al. 2007). Of course, the extent of the benefits will depend on the extent of the reduction as well as the groups in whom a reduction occurs. Greater overall benefits would result if we could persuade weekend binge drinkers to reduce consumption (Room, Babor, and Rehm 2005) than if the only effect were that occasional drinkers were persuaded to become teetotallers. There would also be costs, and potentially even harms, associated with lowered alcohol consumption. Alcohol is widely consumed because people derive important benefits from it, including relaxation, disinhibition that facilitates and eases social interaction, celebration, and enjoyment of the taste. Many people would find that the absence of alcohol diminishes their quality of life, and there may be disruption of established social and cultural traditions in which alcohol is a deeply embedded part. Additionally, for low to moderate drinkers, the cancer risk that is incurred by drinking might be offset by a reduction in risk of cardiovascular disease, meaning that even in 2. Clearly, “consumer protection” and “public health” are not entirely distinct, and there will be substantial overlap in terms of what is intended to protect consumers and what is intended to protect or promote public health. However, it seems to us that the rationale for public health interventions, even where consumer protection is involved, should be rather narrower than the protection of consumer rights per se. In other words, where public health concerns itself with consumer protection, this should be because it is expected to have some benefits (or prevent harms) in terms of people’s health. For example, it is not the job of public health agencies to protect people from Ponzi schemes, while it is within their remit to protect people from unsafe food handling in restaurants.

March, Volume 15, Number 3, 2015

terms of health, drinking (for some) may be on balance beneficial.3 An additional concern is that while the costs of lowered consumption will fall on individuals, many of the benefits—especially in relation to cancer incidence—will mainly be apparent at the population level. Individuals lose something by drinking less, but it cannot necessarily be said that, as individuals, they gain something that is commensurate to this loss. This is not to say that there are no individual benefits; for example, people will save money if they drink less, and especially for heavier drinkers, there will be further benefits such as fewer hangovers, acute injuries, embarrassing misjudgements, and so on. Moreover, there may also be population-level costs— for example, the need to find new sources for sports sponsorship if it no longer benefits the alcohol industry to maintain such funding in the face of reduced consumption. However, as far as health risks (especially cancer) are concerned, alcohol consumption is another instantiation of the “prevention paradox,” whereby “a preventive measure which brings much benefit to the population offers little to each participating individual” (Rose 1985, 38). Most low to moderate drinkers would not have developed a cancer attributable to alcohol use, so may gain no tangible health benefit; at the same time, if we restricted our efforts to those (heavier, irregular) drinkers who would derive sufficient individual benefit from reduced consumption, the population benefits would be much smaller. Thus, as Room, Babor, and Rehm (2005, 521) noted, “The optimum average level of drinking for the population as a whole is likely to be lower than that for an individual.” Clearly, then, the ethical justifiability of an intervention intended to lower alcohol consumption will depend on a careful balancing of the expected benefits against the expected costs. It must be shown that the extent and kind of benefits at the population level are proportional to the costs imposed on individuals, and consideration must be given to the potential harms. However, this is precisely where the greatest strength of warning labels may be considered to lie. As noted, the means by which warning labels would lower consumption are either through direct persuasion or through more gradual attitude changes. They are not an attempt to force people to adopt an “optimum level of drinking,” or even to make it more difficult for them to consume more than this “optimum.” No claims are made about what is an optimum level of drinking, for either the population or the individual. Rather, warning labels leave the choice up to each individual, who can decide—in light of that person’s own values, priorities, and concerns (including other risk factors for cancer)— whether the benefits of drinking are worth the risks.

3. It is worth observing nonetheless that the compounds in alcohol thought to improve cardiovascular health are also available in fresh fruits, which confer other health benefits and do not incur the health harms associated with alcohol consumption.

ajob 5

American Journal of Bioethics

Downloaded by [Seton Hall University] at 17:22 27 March 2015

Of course, the price of allowing individuals to make this judgment for themselves is likely to be that the average level of alcohol consumption across the population is higher than optimal. And, as previously noted, research on warning labels suggests that while they may be effective in raising awareness of risks, they are less effective in changing drinking behavior (Stockwell 2006). In other words, while they may be one component of a strategy that succeeds in persuading individuals to reduce their consumption, they will not be sufficient by themselves.4 However, this may be acceptable, since the alternative to allowing people to decide for themselves would be to disregard people’s own judgments about the level of drinking that is optimal for them, and instead impose a judgment based on what would produce benefits at the population level. This would clearly be ethically problematic (as well as impossible in practice)—not necessarily unjustifiable, but raising additional questions of justification.5

THE RIGHT TO AN INFORMED CHOICE The main argument in favor of mandated warning labels is therefore that it would allow consumers to make an informed choice about their alcohol consumption in light of their own values, concerns, and attitudes to risk. Rather than interfering with autonomous choice, warning labels facilitate it by ensuring that consumers have the information that is relevant to their decision-making around alcohol consumption. It appears that awareness of the link between alcohol and cancer is currently very low. Thus, people’s choices regarding alcohol consumption are not fully autonomous, since they do not understand the nature of the choice they are making or the potential consequences of that choice. Giving people this information allows them either to decide voluntarily to assume this risk, or to lower their risk by reducing or eliminating their alcohol consumption. Indeed, as noted earlier, it is plausible to suppose that people have a right to this information, insofar as cancer is something most people would care about and consider relevant to their decisions. They should be 4. An anonymous reviewer suggests that other strategies such as increased taxation would also “persuade” consumers to lower their consumption. While this may be true, in that higher prices would lead consumers to choose to consume less, it would not persuade them in the sense meant here, that is, by convincing them that it is better to consume less alcohol, all things being equal, because of the health risks. 5. It might be argued that warning labels constitute a coercive interference with the alcohol industry, namely, in disregarding their preferences regarding labeling and presentation of their products. However, this objection is easily met by noting that the Harm Principle allows for interference with choices where these may cause harm to others. Alcoholic products carry a risk of health harms, and if people are not aware of these risks, they are taking on a risk of harm without having consented to it. It is therefore the duty of the industry to ensure that people are aware of these risks, and it is the duty of the state to ensure that this obligation is met.

6 ajob

given the opportunity to decide for themselves whether the risk of cancer is something they are prepared to accept in order to gain the benefits of drinking. This is reasonable as far as it goes—that is, we agree that there is a right to the information about cancer risks, and that people should be able to make up their own minds on this question. However, it is important to note that this argument relies upon certain assumptions, not only about the effectiveness of warning labels in producing awareness and understanding of the risks, but about the nature of the decisions made once this information has been received and understood. That is, there is an assumption that mandated warning labels would make consumers informed about the risks, and that choices made by informed consumers are fully autonomous choices that reflect the person’s considered weighing of the potential benefits and costs in light of their own values, concerns and goals (Wilkinson and Room 2009). Both of these assumptions, but particularly the second one, may be in error, and this has ethical implications. Put simply, if we are arguing for mandated warning labels on the grounds that the desired health outcomes will result from facilitation of autonomous choice, then it matters that warning labels may not facilitate autonomous choice, or that people’s consumption decisions may still be substantially nonautonomous even when “informed.” This does not mean that we ought not to bother providing information, of course— as noted earlier, even if information is not sufficient to ensure autonomous choice, it will be necessary. However, this raises the question of whether there is an obligation to go beyond warning labels in order to create the conditions for genuinely autonomous choice.

THE AUTONOMY AND RATIONALITY OF CONSUMPTION DECISIONS We reiterate that warning labels would generally be implemented as part of a suite of measures, rather than as a stand-alone strategy.6 However, contemplating warning labels as a public health strategy—that is, as a way of reducing consumption rather than as a means of providing information for its own sake—involves an assumption that individuals can and will make appropriate use of this information in deciding whether and how much alcohol to consume. In this section, we consider ways in which this assumption may turn out to be incorrect. 6. An anonymous reviewer raises the objection that other facets of a public health campaign, for example, advertisements, would have a different purpose to that of warning labels. For example, advertisements (or even warnings on alcohol advertisements) would have the purpose of counteracting the force of alcohol industry persuasion. However, while the proximate purpose may appear different, the ultimate purpose—reducing alcohol consumption, and/or creating awareness of alcohol consumption as a risky activity—is the same.

March, Volume 15, Number 3, 2015

Downloaded by [Seton Hall University] at 17:22 27 March 2015

Ethics of Cancer Risk Warnings on Beverages

One way in which warning labels might fail to facilitate autonomous choice is by failing even to ensure that people are fully informed about the risks. There will be a trade-off between comprehensibility/accessibility and information: In order to be easily understood, especially by those with low literacy or English comprehension, the labels will have to be simple and direct. Thus, one of the questions we must answer is how specific a person’s understanding of the risks must be for them to be considered “informed” about these risks. If a person knows that any drinking carries a risk, and that the risk increases with higher consumption, are the person’s choices informed even if the person has no idea of the absolute degree of risk? As noted, warning labels would not stand alone, and this concern might therefore be addressed by having more detailed information available, such as on websites. However, it is likely that few people will seek out this information, and we cannot assume that those who do not do so simply do not care to know. The problem of how to achieve sufficient awareness in those with low literacy, technological skills, time constraints, and so on is a problem for all forms of health education. The particular issue in relation to warning labels is the need to convey an appropriate kind and amount of information in the (very small) space available, and the trade-off that must be made between making a label easily understood (and accessible to a more diverse set of literacy levels) and conveying important nuance. The relevant information includes not just the fact of the alcohol–cancer link, but the doseresponse nature of the link, and the balance between risks and benefits at low levels of consumption. People would probably think that the most relevant information was the level at which the risks become significant enough to outweigh at least the other health benefits. A more important concern, however, is that even when people are appropriately informed, and even if they have used this information to form beliefs and preferences about their desired level of drinking, their actual choices may fail to connect appropriately with those beliefs and preferences. This is obviously true of those with a physiological dependence on alcohol, where drinking is to some extent “out of their control.” However, it may also be true of those who are not alcoholics (or physiologically dependent). Decisions about drinking, once we understand the risks of cancer (and other long-term health effects), involve balancing present benefits against potential future harms. And much research on human decision-making shows that we have a systematic tendency to undermine our longer term, more considered preferences in favor of shortterm rewards. While most of us adopt a range of cognitive and practical strategies to thwart this tendency (e.g., “selfbinding” so as to make it impossible to choose the shortterm reward), these are not always available or successful. The general phenomenon of time-inconsistent preference reversal can easily be seen to apply to drinking behavior. For example, my considered preference, after weighing up all of the evidence in light of my own values and circumstances, may be to drink only on special occasions,

March, Volume 15, Number 3, 2015

and then to have no more than one drink. However, in certain situations (after-work socializing, dinner with friends) I regularly—but temporarily—reverse this preference and decide I’ll have a drink (or two). There are various theories about the mechanism that drives this phenomenon. Ainslie claimed that when people discount the value of future goods, they devalue too steeply, according to a hyperbolic rather than exponential function (Ainslie 2001). This allows smaller goods close to the present to temporarily seem more rewarding than larger benefits in the future, causing temporary preference reversal: I prefer getting $100 in a year over getting $20 in 6 months, but at the 6month mark I prefer the $20 now to waiting for the $100 (Ainslie 2001). In relation to drinking behavior, this implies that when an occasion for drinking is close at hand, the perceived benefits will temporarily loom larger than the too steeply discounted benefits of avoiding cancer. An alternative explanation involves so-called “dual process” theories of decision making, which hold that we have two different “systems”: a “rational” system, which is slower and makes use of evidence, logic, and analysis; and an “experiential” system, which is faster and makes use of heuristics and intuitions (Slovic et al. 2005). People vary in the degree to which they rely on each of these systems. However, we generally rely more on the experiential system when under stress, rushed, distracted, or subject to increased cognitive load. Emotions, or emotional associations, play an important part in experiential decision making, and—importantly for our purposes—have an effect on perceptions of risk. As Slovic and colleagues (2005) noted, activities that carry positive emotional associations are perceived to be less risky (and more beneficial), while activities that carry negative emotional associations are perceived as more risky. In the case of drinking, we can see that warning labels—which encourage weighing of risks against benefits—may engage the rational system, but actual drinking choices are more likely to be made using the experiential system, where powerful positive emotional associations may cause the risks to be downplayed. Insofar as time-inconsistent preference reversal is a phenomenon, it is plausible to think that it could manifest itself in relation to alcohol consumption. In fact, there are other aspects of drinking that make it an even more likely candidate for such behavior. First, like much consumption behavior, the future bad consequences are not the result of a single act of consumption, but rather of a pattern of consumption over time. This creates a situation (analogous to a collective action problem) where I can, on each occasion, truly say to myself that having a(nother) drink now is not going to appreciably alter my cancer risk, even though a long enough sequence of such occasions certainly will. Second, the physiological effect of alcohol itself is likely to exacerbate the tendency toward overvaluing short-term rewards. Alcohol consumption lowers anxiety, leading to increased risk-taking (Room, Babor, and Rehm 2005) and impulsive behavior. It also increases emotional responses and impairs cognitive functioning (Room et al. 2005),

ajob 7

Downloaded by [Seton Hall University] at 17:22 27 March 2015

American Journal of Bioethics

which could increase reliance on experiential decision making. And finally, external forces—social pressures, advertising, cultural norms—create both positive emotional associations with drinking, and negative emotional associations with not drinking (Caswell 2012). Finally, the notion of relational autonomy is also very relevant here, as it reminds us that individuals do not form preferences—or act on them—in a social vacuum, but rather that “social relations influence and perhaps constitute agents‘ senses of themselves and their capacities” (Mackenzie and Stoljar 2000, 8). In other words, people’s decisions about whether to drink may depend strongly upon how those decisions impact upon their relationships with others. Not drinking may mark a person as nonconformist and may be construed by others in their social circle as carrying implicit disapproval, thus leading to perceptions that they are a “killjoy” or a “wet blanket.” Thus a person’s desire to maintain an identity as part of a social group (e.g., as “one of the lads”) and to maintain good relationships with others may come into conflict with their desire to refrain from drinking in any given situation. In fact, to the extent that such social forces may limit what combinations of preferences and choices are possible, they may substantially impair autonomy as regards alcohol consumption decisions. The problem, then, is that people’s drinking choices may not be sufficiently autonomous even if they are aware of the risks and have used this information to form considered preferences about alcohol consumption. If people do not change their drinking behavior in response to awareness of the cancer risks, this may not be because they have decided that the benefits outweigh the potential harms. Rather, it may be due to the failure to maintain considered preferences in the face of incentives to consume, and the unavailability or impracticality of various strategies to manage time-inconsistent preference reversal in relation to alcohol and drinking. We began this section by supposing that alcohol warning labels might avoid the potential ethical problem of balancing the benefits and costs of reducing alcohol consumption, because they were supposed to achieve this end by the facilitation of autonomous choice. However, if warning labels do not in fact facilitate autonomous choice, then we must consider the possibility that we have a responsibility to do more. If people would prefer to avoid or reduce their risk of cancer, but are hampered in their ability to act on these preferences, then there may be an obligation to take further measures to support fully autonomous choice—for example, price controls or advertising restrictions.

NONRATIONAL DECISIONS NOT TO DRINK We have discussed the possibility that warning labels might, through rational persuasion, lead people to form preferences to drink less, but that these preferences might fail to connect appropriately to their drinking choices

8 ajob

because of the way in which rational decision making can be trumped by short-term or experiential reasoning. Here, we consider the converse scenario: namely, that cancer warning labels might cause people to change their drinking behavior, but via nonrational means. In other words, people might be induced by the warning labels to drink less, but not because they have made a considered decision to do so, and perhaps when their interests, all things considered, would favor drinking. This scenario may seem much less plausible than the previous one. It may be more likely, however, if the labels are specifically designed to elicit an emotional response (as arguably the case for the graphic warnings on cigarette packs in, for example, Australia and Canada, depicting tobacco-related diseases), or if the risks are (intentionally or otherwise) overstated, as may happen because of lack of nuance necessary to convey a message clearly. Certainly, there is convincing evidence from tobacco labeling research that health warnings that include graphics are more effective in impacting smokers’ behavior than those that do not (Borland 2009). In a comprehensive review on the evidence of tobacco health warning labels in more than a dozen countries, Hammond concluded that graphic “health warnings that elicit strong emotional reactions are significantly more effective” in impacting smoking behavior than (smaller) text-only messages (Hammond 2011, 327). In a Canadian study, researchers sought to address criticisms that graphic tobacco warnings would cause unnecessary or excessive emotional distress, engender avoidance of the message, and even prompt increases in consumption. They found that graphic labels certainly elicited strong negative emotional reactions (such as fear and disgust), but that smokers who reported higher levels of these were more likely to have acted to reduce or quit smoking, regardless of whether they avoided the warning or not (Hammond et al. 2004). The authors argued that health warning labels can be effectively used to prompt behavioral change through nonrational means, namely, through eliciting an emotional reaction, and specifically through communicating “health risks that are manifestly frightening and harsh”; they further implied that to fail to effectively engender this emotional arousal is to “fail to communicate these risks in a truthful, forthright manner” (Discussion {2.) Nonetheless, with regard to alcohol labels warning of cancer risk, as Slovic and colleagues (2005, S38) have pointed out, cancer is a “highly-dreaded” disease; therefore, warnings mentioning cancer might carry stronger negative emotional associations, which (as previously mentioned) tend to make people perceive an activity as more risky. Agostinelli and Grube (2002, 19) reported that among U.S. college students, warnings with the words “poison” and “cancer” “elicited more avoidance” of alcohol than the standard (U.S. Surgeon General) warning label. Thus, it is at least possible that some people would react to warning labels by overreacting to the actual risk, and reducing their consumption below the level that would in fact be optimal for them.

March, Volume 15, Number 3, 2015

Downloaded by [Seton Hall University] at 17:22 27 March 2015

Ethics of Cancer Risk Warnings on Beverages

This concern may seem rather trivial; it is slightly odd to consider a complaint that a person would, under conditions of informed rational deliberation, have formed a preference to drink more, but in fact drank less because a warning label caused the person to fear getting cancer. We must also not fall into the trap of thinking that any elicitation of emotion is manipulative or somehow illegitimate: An emotional response will often be quite appropriate. It could be argued (and, as mentioned earlier, was implied by Hammond et al. 2004) that fear or dread is entirely appropriate in the face of a risk of cancer, and the elicitation of this response in fact aids rationality by providing the emotional emphasis necessary for us to pay proper attention to relevant aspects of our choice. This does not entirely resolve the concern, however. It would have to be shown that the kind and degree of emotional response were in fact appropriate to the situation— in particular, that the labels did not cause people to feel fear or dread disproportionate to their actual risk. Alternatively, it would have to be shown that there was a justification for eliciting disproportionate fear, or otherwise manipulating people to prompt behavior change. The debate over “counter-advertising” and similar tactics is ongoing, and the ethical issues are complex. One of the considerations likely to be important in relation to alcohol is the likelihood that, from a public health point of view, it would be desirable to aim to lower consumption to a level below what is in fact optimal from the individual’s point of view. If the means of achieving this does not involve rational persuasion, the ethical issue of justification becomes more difficult.

EQUITY A final ethical consideration is that of equity: Warning labels may exacerbate, or fail to properly address, existing health inequalities in society. This could occur if the main benefits of warning labels accrued among those who were already healthy, or they were systematically less effective in those who are already disadvantaged. We can see from the preceding discussion that there are various ways in which such inequities might arise. First, the effectiveness of warning labels (in raising awareness as well as changing behavior) depends on their being understood. Those with low literacy, or who are not fluent in English, may not comprehend the warning labels, and may live in communities where others likewise do not understand them. Insofar as these groups are already disadvantaged and less healthy, the expectation that they will not receive the message raises serious concerns. Any implementation of warning labels must consider alternative means of getting information to people who will not be able to read or understand the labels. As noted earlier, supplemental information sources such as websites raise similar problems, in that disadvantaged groups may lack the ability or opportunity to seek out these sources of information.

March, Volume 15, Number 3, 2015

A less obvious source of potential inequity, but one that is no less significant, arises from the idea that individuals will use information about cancer risk to form considered preferences about alcohol consumption. For those who are socially and economically disadvantaged, as well as less healthy, the weighing of risks and benefits may be systematically different from those who are more fortunate. In particular, it is not implausible to suppose that some of the benefits of alcohol consumption—relaxation, enjoyment— may be far more important for those whose lives are otherwise difficult and unrewarding; moreover, the risk of cancer at some distant point in the future may be less relevant for those who face many larger risks of harm in the short term. Moreover, people living in such uncertain circumstances are also precisely those for whom social support and belonging may be most important; the risk of negative social and relational consequences for choosing not to drink may be unacceptably high. If this is correct, then for more disadvantaged people, the benefits of alcohol consumption may in fact outweigh the future risks; it would be wrong, however, to simply accept this, rather than trying to address the factors that shape this outcome. A similar observation may be made in relation to the potential for considered preferences to fail to connect to drinking behavior. While time-inconsistent preference reversal is a phenomenon that is more or less universal, it may be more of a problem for certain disadvantaged groups. This doesn’t just mean that those who have grown up in impoverished circumstances may have failed to learn strategies for managing the phenomenon, though this may be true. It also means that the environment (the amount of advertising they encounter, the norms amongst their peers, etc.) may make it more difficult to resist preference reversal, or to resist acting on it. If such inequities are found to arise from warning labels—that is, if the labels’ effectiveness systematically excludes some of the more disadvantaged members of society—then we must again consider whether we have met our obligations if we do not take further action. Obviously, addressing social and economic disadvantage is difficult, expensive, and beyond the ability of public health interventions alone to remedy. However, this does not necessarily make it permissible to accept that our attempts to improve health will simply fail to reach those who need it most.

CONCLUSION In this article we have explored the ethical issues that arise in relation to mandated cancer warnings on alcoholic beverages. One line of argument would have it that warning labels are ethically preferable to more “intrusive” or coercive approaches, as they allow individuals to decide for themselves whether the benefits of reduced alcohol consumption are worth it to them. However, while there are no grounds for thinking that mandated warning labels are unethical per se (i.e., considered in isolation), there are various reasons why they may in fact represent a suboptimal

ajob 9

Downloaded by [Seton Hall University] at 17:22 27 March 2015

American Journal of Bioethics

intervention where the goal is to reduce alcohol consumption. In particular, even if we stipulate that this outcome should be achieved via facilitation of autonomous choice by individuals, there is reason to believe that mere provision of information falls short of ensuring that individuals’ drinking behavior will be autonomous. Having said this, we do not believe that mandated warning labels are unjustified—only that, if implemented, attention must be given to the intended aim of the policy, and the potential ethical problems should be addressed. As we have noted, even if some other intervention would be more effective, the feasibility of such interventions may depend upon first changing community attitudes so that other measures are accepted. Warning labels may be a necessary step in creating greater awareness of the risks and thereby building support and acceptance for further measures designed to reduce consumption. Note, however, that this requires us to be very clear about what exactly the purpose of the warning labels is. They may be ethically problematic as an intervention designed to reduce alcohol consumption, but not as an intervention designed to pave the way for future measures that are aimed at reducing alcohol consumption. In fact, our discussion shows that the justifiability of warning labels may depend very much upon the context— the other interventions and policies that are adopted concurrently or subsequently. If labels are intended to produce long-term changes in attitudes that will lead to public support for further interventions that are more effective in producing behavior change, then it is important that there be plans for further interventions. Conversely, if we design labels for emotive impact—to change behavior at the point of choice via affective salience rather than rational deliberation—then we must be aware of the possibility of manipulation, and consider whether this is ethically justifiable. We must also concern ourselves with the best way to reach those for whom warning labels are likely to be least effective, and to consider whether we are aiming at reducing consumption to a “socially optimal” or an “individually optimal” level. &

REFERENCES AER Foundation Policy Paper. 2011. Alcohol product labelling: Health warning labels and consumer information. Available at: http://www.fare.org.au/policy-advocacy/alcohol-product-label ling (accessed September 1, 2014). Agostinelli, G., and J. W. Grube. 2002. Alcohol counter-advertising and the media: A review of recent research. Alcohol Research & Health 26(1): 15–21. AIHW, S. Begg, T. Vos, et al. 2007. The burden of disease and injury in Australia 2003. Catalogue number PHE 82 2007. Canberra, Australia: AIHW. Available at: http://www.aihw.gov.au/publica tion-detail/?idD6442467990. Ainslie, G. 2001. Breakdown of will. New York, NY: Cambridge University Press.

10 ajob

Al-hamdani, M. 2014. The case for stringent alcohol warning labels: Lessons from the tobacco control experience. Journal of Public Health Policy 35 (1): 65–74. Available at: http://dx.doi. org/10.1057/jphp.2013.47 Alcohol Concern (UK). 2009. Message on a bottle—Does the public have enough information about what they are drinking? London, UK: Alcohol Concern. Australian Chronic Disease Prevention Alliance. 2011. Health information and warning labels on alcohol: Position statement. Available at: http://www.cancer.org.au/content/pdf/ACDPA/ 110930-Final-ACDPA-PS-Alcohol-labelling-PS.pdf Bagnardi, V., M. Rota, E. Botteri, et al. 2013. Light alcohol drinking and cancer: a meta-analysis. Annals of Oncology 24: 301–308. Available at: http://dx.doi.org/10.1093/annonc/mds337 Bell, K., A. Salmon, M. Bowers, et al. 2010. Smoking, stigma and tobacco ‘denormalization’: Further reflections on the use of stigma as a public health tool. A commentary on Social Science & Medicine’s Stigmas, Prejudice, Discrimation and Health Special Issue (67: 3). Social Science and Medicine 70: 795–799. Available at: http:// dx.doi.org/10.1016/j.socscimed.2009.09.060 Borland, R. 2009. Impact of graphic and text warnings on cigarette packs: findings from four countries over five years. Tobacco Control 18: 358–364. Available at: http://dx.doi.org/10.1136/tc.2008.028043 Campbell, D. 2012. Alcohol packages should carry graphic health warnings, urge doctors. The Guardian, July 13. Available at: http:// www.theguardian.com/society/2012/jul/13/alcohol-packaginggraphic-health-warning/print (accessed September 1, 2014). Cancer Council Australia. 2013. Position statement: Consumer information and labelling of alcohol. Available at: http://wiki. cancer.org.au/policy/Position_statement_-_Consumer_informa tion_and_labelling_of_alcohol (accessed September 1, 2014). Caswell, S. 2012. Why have guidelines at all? A critical perspective. Drug and Alcohol Review 31: 151–152. Available at: http://dx. doi.org/10.1111/j.1465-3362.2011.00376.x Di Castelnuovo, A., S. Costanzo, V. Bagnardi, et al. 2006. Alcohol dosing and total mortality in men and women: An updated metaanalysis of 34 prospective studies. Archives of Internal Medicine 166 (22): 2437. Available at: http://dx.doi.org/10.1001/archinte. 166.22.2437 Eliott, J. A., and E. R. Miller. 2014. Alcohol and cancer: The urgent need for a new message. Medical Journal of Australia 200(2): 71–72. Available at: http://dx.doi.org/10.5694/mja13.10426 Hammond, D. 2011. Health warning messages on tobacco products: A review. Tobacco Control 20(5): 327–337. Available at: http://dx.doi.org/10.1136/tc.2010.037630 Hammond, D., G. T. Fong, P. W. McDonald, et al. 2004. Graphic Canadian cigarette warning labels and adverse outcomes: Evidence from Canadian smokers. American Journal of Public Health 94(8): 1442–1445. Available at: http://dx.doi.org/10.2105/AJPH.94.8.1442 Holmes, M. V., C. E. Dale, L. Zuccolo, et al. 2014. Association between alcohol and cardiovascular disease: Mendelian randomisation analysis based on individual participant data. British Medical Journal 346: g4164. Available at: http://dx.doi.org/10.1136/ bmj.g4164

March, Volume 15, Number 3, 2015

Ethics of Cancer Risk Warnings on Beverages

Institute of Alcohol Studies. 2013. Alcohol causes cancer—Experts call for health warnings. The Globe 3: 19. International Center for Alcohol Policies. 2010. Health warning labels. Available at: http//www.icap.org/table/HealthWarningLabels Mackenzie, C., and N. Stoljar. 2000. Introduction: Autonomy refigured. In Relational autonomy: Feminist perspectives on autonomy, agency and the social self. New York, NY: Oxford University Press. Room, R., T. Babor, and J. Rehm. 2005. Alcohol and public health. Lancet 365: 519–530. Available at: http://dx.doi.org/10.1016/S01406736(05)17870-2 Room, R., and J. Rehm. 2011. Alcohol and non-communicable diseases—Cancer, heart disease and more. Addiction 106(1): 1–2. Available at: http://dx.doi.org/10.1111/j.1360-0443.2010.03223.x

Downloaded by [Seton Hall University] at 17:22 27 March 2015

Rose, G. 1985. Sick individuals and sick populations. International Journal of Epidemiology 14(1): 32–38. Available at: http://dx.doi. org/10.1093/ije/14.1.32 Slovic, P., et al. 2005. Affect, risk, and decision making. Health Psychology 24(4 Suppl.): S35–S40. Available at: http://dx.doi.org/ 10.1037/0278-6133.24.4.S35 Stockwell, T. 2006. A review of research into the impacts of alcohol warning labels on attitudes and behaviour. Victoria, BC, Canada: Centre for Addictions Research of BC.

March, Volume 15, Number 3, 2015

Thomson, L. M., B. Vandenberg, and J. L. Fitzgerald. 2012. An exploratory study of drinkers views of health information and warning labels on alcohol containers. Drug and Alcohol Review 31(2): 240–247. Available at: http://dx.doi.org/10.1111/ j.1465-3362.2011.00343.x UK Faculty of Public Health. 2012. Health warnings needed on alcohol. Available at: http://www.fph.org.uk/health_warnings_nee ded_on_alcohol (accessed September 1, 2014). Wilkinson, C., and R. Room. 2009. Warnings on alcohol containers and advertisements: International experience and evidence on effects. Drug and Alcohol Review 28: 426–435. Available at: http:// dx.doi.org/10.1111/j.1465-3362.2009.00055.x Winstanley, M. H., I. S. Pratt, K. Chapman, et al. 2011. Alcohol and cancer: A position statement from Cancer Council Australia. Medical Journal of Australia 194(9): 479–482. World Cancer Research Foundation. 2007. Food, nutrition, physical activity, and the prevention of cancer: A global perspective. Available at: http://www.dietandcancerreport.org World Health Organization. 2009. Alcohol drinking. Summary of data reported and evaluation. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans 44. Available at: http:// monographs.iarc.fr/ENG/Monographs/vol44/volume44.pdf

ajob 11

Mandatory cancer risk warnings on alcoholic beverages: what are the ethical issues?

The link between alcohol consumption and cancer is well established, but public awareness of the risk remains low. Mandated warning labels have been s...
140KB Sizes 0 Downloads 2 Views