Article Text

Download PDFPDF

Ethics of tobacco harm reduction from a liberal perspective
  1. Yvette van der Eijk
  1. Correspondence to Yvette van der Eijk, Centre for Biomedical Ethics, National University of Singapore, 10 Medical Drive 02-01, Singapore 117597, Singapore; yvette.eijk{at}


Mixed evidence on the possible harms, benefits and usage patterns of electronic nicotine delivery systems (ENDS, or ‘e-cigarettes’), has led to vigorous and ongoing debates on the issue. The ethical trade-off often represented is that, though smokers should be permitted access to ENDS as a less harmful alternative to smoking, this comes at the expense of non-smokers and children who may experiment with ENDS, become addicted to them, or experience health issues from long-term exposure to passive ENDS vapour. Lacking from many debates is a balanced analysis based on sound ethical reasoning, so this paper aims to examine the issue from a liberal perspective. More specifically, focus is on how ENDS policy can help to promote freedom in a broader sense, with ‘freedom’ considered as originating from having options and the necessary information and ability to autonomously choose between these options.

  • Autonomy
  • Ethics
  • Health Promotion
  • Public Health Ethics
  • Drugs and Drug Industry

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction: the tobacco harm reduction debate

Tobacco harm reduction is based on similar principles as harm reductive strategies for other drugs: to minimise the adverse health, social and economic consequences of drug use without expecting complete abstinence. A common example is the substitution of heroin with oral methadone for people with heroin addiction. From a public health perspective, harm reduction is generally considered ethical as it tends to reduce the negative impacts of drug use and can bring addicted drug users closer to treatment services.1 In the context of tobacco, a form of harm reduction that has become popularised recently is the use of e-cigarettes or ‘vaping’ devices—collectively termed electronic nicotine delivery systems (ENDS)i—as a substitute for cigarette smoking.

Since the invention of ENDS in 2003, their sales are booming: in 2014, over 450 brands were sold in a market worth roughly $3 billion.2 Restrictions on the marketing, design, and use of ENDS have been inconsistent,ii in part due to mixed evidence which has led to ongoing debates on whether a market for ENDS should be permitted. In terms of their harm potential, ENDS contain lower levels of harmful chemicals than cigarettes, and improvements in impaired lung function have been reported in smokers who have completely switched to ENDS.4 However, toxicants such as formaldehyde and acrolein are present in some varieties of ENDS,2 and studies indicate that ENDS can cause inflammation in the lungs which could result in chronic, fatal conditions such as emphysema.5 ,6 Further, due to the recentness of ENDS, their long-term health impacts—on self and others breathing in exhaled vapour—are unknown.7 It is known, however, that nicotine is highly addictive and has peripheral actions as a blood vessel constrictor and a tumour promoter.2 Nicotine-containing ENDS can therefore be highly addictive and may result in long-term complications related to cancer or heart disease, potentially for others exposed to ENDS vapour as well as ENDS users themselves.8 Although some argue that a market for ENDS should be permitted on the basis of their lower harm profile relative to cigarettes,9 others maintain that ENDS should be restricted as their widespread use would still have important public health impacts.10

Research on the use of ENDS is similarly inconclusive. For smokers, a switch to ENDS should be complete since cutting down smoking reaps little long-term benefit.11 Evidence indicates, however, that many smokers use ENDS to cut down smoking or to circumvent smoking bans,2 and that smokers who use ENDS are less likely to quit than smokers who do not.12 Yet, evidence also shows that many ENDS users have completely switched from cigarettes. In the UK, for example, an estimated 30% of ENDS users have completely quit smoking.11 Meanwhile, youth experimentation with ENDS has proliferated considerably. There are fears that this will encourage smoking initiation among youth, since ‘vaping’ may glamourise the act of smoking, and—as nicotine addictions develop from the use of ENDS—these youth will look to more potent nicotine sources such as cigarettes. While evidence on this ‘gateway effect’ remains unclear, the harms that may result from youth experimentations with ENDS, including the development of nicotine addiction, are a concern that most public health advocates share.13

Another widely expressed concern is the involvement of the tobacco industry (TI), particularly in the marketing of ENDS. The TI actively promotes ENDS as ‘healthy’ and ‘safe’, even though these claims remain uncertain. ENDS are promoted as effective cessation aids even though many—if not most—smokers engage in dual use; and as appropriate for use in areas where smoking is banned, even though ENDS vapour is not necessarily harmless to others.7 Moreover, the TI is strongly promoting ENDS to youth by selling them in a wide variety of flavours that mimic sweets—such as ‘coconut bomb’ and ‘vanilla cupcake’—and in an attractive array of shapes and colours.14 The marketing messages emulate ideas related to independence, sex appeal, success and lifestyle choice.2 Youth are highly influenced by such advertising, and more likely to experiment with ENDS as a result.15 In adults it may also trigger dual use or relapse into smoking, since adverts for ENDS have been found to trigger cigarette cravings in addicted smokers and ex-smokers.7

The public health community, which is usually united in its goal to minimise or eliminate tobacco-related harm, is now experiencing significant divisions on the ENDS issue. Most arguments ‘for’ or ‘against’ an ENDS market are consequentialist: that ENDS will improve or worsen public health. However, the mixed nature of current evidence, as well as the unavailability of long-term evidence, means that neither argument can be ascertained (for now). As it stands, there is little agreement on whether public health advocates should cooperate with the TI in reducing the public health impact of smoking, how children should be protected from experimenting with ENDS, or how smokers should be supported in their cessation or harm reduction efforts. On both sides of the debate, supporting evidence tends to be selectively presented with balanced ethical reasoning often missing.16

Differences aside, some important points are highlighted by these debates. First, better solutions for smoking cessation are needed. Just 2–7% unaided cessation attempts are successful; with approved cessation aids,iii success rate increases to a modest 25%.17 Second, the popularity of ENDS shows that many smokers welcome the availability of options, especially if these options are thought to provide long-term health benefits. In a broader context, arguments that support consumer choice and freedom have been influential in tobacco control policy. For instance, the TI successfully avoided tobacco regulations for decades by arguing that smoking is a ‘free choice’, and smoking bans were more readily accepted when framed as a means to promote the freedom of non-smokers (from the harms of passive smoke).18 It is therefore important to consider liberal reasoning in this context: how an ENDS policy can maximise freedom for each of the populations affected; notably smokers, non-smoking adults and children.

The aim of this paper is, from a liberal perspective, to determine what an ethical policy for ENDS would look like. It starts by describing a liberal theory that can be applied in the context of ENDS policy. This theory is then applied to populations affected differently by the issue (smokers, non-smoking adults and children), with some future indications for ENDS policy.

A liberal theory for tobacco harm reduction

Liberal theories consider freedomiv to be the pre-eminent principle of morality that policies should protect. Mill's liberal theory has been particularly influential,19 also in the public health context. In the public health ethics framework of the Nuffield Council, for example, Mill's liberal theory is developed into an ‘extended liberalism’ for public health, which goes further in recognising the responsibility of the state in helping people to live more healthily. This responsibility is then termed as the concept of ‘stewardship’.20 In Mill's liberalism, freedom is considered in a negative sense: freedom that arises from minimal state intervention. In this view, an individual's actions should only be restricted if they are harmful to others (the ‘harm principle’), and decisions should only be made in the best interests of those who are incapable of making rational and informed decisions; for example people with severe mental impairments, or children, who are not yet in the maturity of their faculties.19 Interventions outside these boundaries are considered paternalistic and therefore unethical.

An issue with this approach, particularly in the context of public health, is that justifications for a particular intervention are limited to the harm principle or the assumption that an individual is incapable of exercising a rational choice. Yet such a theory does not unfold so simply in many public health contexts. What if, for some adults, the ability to make a rational and informed decision is only partially impaired, and could be redeveloped with some help from the state? In the context of health promotion, this idea—as mentioned above—has been framed as the concept of stewardship. In the context of freedom, it may be more clearly understood when ‘freedom’ is considered in a positive sense; as in, as essentially connected to the concept of autonomy.

Positive freedom, unlike negative freedom, arises from having options (minimal state intervention), and from having the ability to choose between options (autonomy). Positive freedom, then, is created by interventions that promote or protect autonomy,21 with autonomy here defined as the ability to make free, informed decisions that are in line with one's own motives, reasons and values.22 Thus it differs from negative freedom, as it assumes that the necessary conditions for autonomy do not always pre-exist and may need to be created by the state. Interventions on individual actions are then ethically justified, provided that their aim is to promote or protect autonomy. In the context of tobacco control, it is important to also consider freedom in its positive sense, for several reasons.

First, nicotine is addictive, and nicotine addiction narrows one's ability to resist smokingv and therefore the ability to choose between two options: the option to use nicotine, and the option to not use Further, addiction affects most smokers: 90% of smokers regret having ever started smoking,vii and in any given year 40–50% smokers attempt to quit smoking,24 though very few succeed in the long term. Second, children, due to their developing neurobiology, are particularly vulnerable to developing addiction and may not be capable of fully appreciating the risks of initiating the use of nicotine products;25 yet youth experimentation with ENDS is increasingly common. Third, to make an autonomous choice requires that it is an informed choice. Yet the TI's marketing strategies leave little scope for making an informed choice regarding the use of ENDS, due to the misleading nature of its advertising and the inconclusiveness of evidence used to support many ENDS-related arguments.

Thus in the context of nicotine addiction, the distinction between ‘autonomous’ and ‘not autonomous’ is blurred as many smokers, though capable of exercising rational decisions in other life aspects, often experience great difficulties in quitting smoking. Would a state intervention that helps them overcome these difficulties be paternalistic? This paper argues not, because in this case the state is not dictating how the individual should or should not behave; rather, it is providing a type of stewardship that, rather than focusing on health promotion (as in the case of the Nuffield Council's stewardship), focuses on promoting autonomy. The liberal theory used as the basis of the arguments in this paper, then, aims to protect freedom in its positive and negative sense. It is a theory in which: (1) autonomy-undermining options are restricted; (2) autonomy-promoting options are permitted; (3) the best interests of children are protected; (4) actions harmful to others are restricted;viii and (5) informed choice is considered an important aspect of the autonomous decision-making process. It includes Mill's harm principle and the idea that the best interests of children should be protected. Yet is goes beyond Mill's liberalism in that it considers autonomy as an essential prerequisite for free action, and assumes that the necessary conditions for autonomous action are not always pre-existent.

In the context of ENDS policy, freedom then arises from having a set of options that do not undermine autonomy and do not harm others, having adequate information about these options, and having systems in place to protect the best interests of children. How would such a liberal theory apply to different populations affected by ENDS: smokers, non-smoking adults and children?

Alternative options for smokers

For the purpose of this analysis, the term ‘smoker’ includes current smokers and ex-smokers who are likely to relapse. For them, the ability to engage in the option to not smoke may be impaired to some degree. Some smokers may be unwilling to quit using nicotine, yet interested in minimising the health impacts of their smoking by using nicotine in a less harmful form such as ENDS.

It may be argued that, since ENDS are addictive and therefore autonomy-undermining, smokers should be denied the option to use them. However, this would be inconsistent with the fact that the use of cigarettes, which are also addictive, is still permitted.9 Further, for nicotine-addicted smokers, perceived options regarding nicotine use have become narrowed. A complete switch to ENDS is then an alternative option for those who consider quitting all nicotine use to be unfeasible, and is not autonomy-undermining since they are already addicted to nicotine. For those interested in eventually quitting all nicotine use, ENDS could be designed and provided in a way that helps people to wean off nicotine completely. This could be done, for example, by designing ENDS such that the nicotine delivery can be gradually reduced or by involving a health professional who can advise on the proper use of ENDS. Thus ENDS are not autonomy-undermining if they are provided as an alternative option to smokers who are already addicted to nicotine, and potentially autonomy-promoting as they could help to wean some smokers off nicotine completely.

However, as mentioned, ENDS are heavily marketed by the TI which is problematic for smokers as ENDS adverts can trigger cigarette cravings.7 This may lead some smokers to relapse, so ENDS adverts are potentially autonomy-undermining and should be restricted. Furthermore, marketing and information sources for ENDS are often misleading as they promote ENDS as a safe cessation option even though there is no robust evidence to back this claim. Impartial safety information on ENDS is rarely communicated to smokers, who have to rely on sources such as the media, the internet or owners of vaping stores. However, these owners often get their ‘facts’ from industry sources, and end up promoting ENDS as safe and healthy.26 These messages undermine smokers’ abilities to make an informed choice regarding the use of ENDS. An autonomy-promoting approach, then, would be to restrict ENDS marketing and to implement a system in which smokers can be properly informed on the potential benefits and risks of using ENDS, the uncertainty of current evidence and indications for proper use. This could be done, for instance, by restricting the supply of ENDS to smokers through a doctor or pharmacist, who is then trained to disclose up-to-date and unbiased information. With this approach, access to the most harmful varieties of ENDS—such as those containing carcinogenic additives—may be restricted.ix Though some tobacco harm reduction advocates may consider this approach to be restrictive, it is still concordant with harm reduction principles. Oral methadone, for example, is not sold on a free market; it is usually prescribed to heroin-addicted individuals by a doctor and taken under the supervision of a pharmacist.

Non-smoking adults and children

In this analysis, ‘non-smoker’ refers to adult never-smokers and ex-smokers who are unlikely to smoke again. Non-smokers are unlikely to benefit from using ENDS and may suffer long-term harm. As mentioned, early evidence indicates that ENDS may cause emphysema, nicotine-related complications and other conditions through exposure to toxicants such as acrolein. Although some varieties may not result in any—or very few—short-term or long-term complications, in all cases the nicotine-containing varieties are addictive. Thus non-smokers who have started using nicotine-containing ENDS and later wish to quit may find themselves unable to. Since these non-smokers would have developed addiction primarily as a result of using nicotine-containing ENDS, they are potentially autonomy-undermining to non-smokers. The use of nicotine-containing ENDS among non-smokers should therefore be restricted. This could be done by regulating the supply of ENDS such that they are only provided by a healthcare professional who ensures that nicotine-containing ENDS are only given to smokers.

However, a different analysis applies to nicotine-free ENDS since they are not addictive and therefore less likely to be autonomy-undermining. In this instance, a necessary precondition for autonomy—non-addiction—exists and a negative conception of liberty may apply. Within liberal reasoning, even if nicotine-free ENDS present health risks, it would not be unethical to sell them to non-smokers provided that they are able to make an informed decision on their use. Possible risks or uncertainties regarding their safety should then be clearly communicated, and all misleading marketing practices should be restricted. A potential issue remains, however, for non-smokers exposed to passive ENDS vapour. Although long-term evidence on the effects of passive ENDS vapour is not yet available, it is possible that permitting vaping in enclosed public spaces will result in an increased risk of chronic disease for others. The liberal theory described here would consider this unethical on the basis of the harm principle, and so vaping in public indoor areas should be banned; though these bans should be lifted if long-term, consistent evidence proves that passive ENDS vapour is safe.

In the case of children, there are important parallels between ENDS and cigarettes in terms of how both products have been marketed. The TI has a long and well characterised history of attracting children to cigarettes with flavours, attractive packaging, product placements and adverts designed specifically to target young children. These strategies were highly successful at creating new waves of addicted ‘replacement’ smokers.27 Similarly, the attractive designs, flavours and adverts for ENDS have triggered youth experimentations with ENDS which has raised concerns that a new generation of addicted, lifelong ENDS users is currently being groomed by the TI.13 Even if ENDS cause a fraction of the harm relative to cigarettes, nicotine-containing ENDS varieties are addictive and children—due to their young age—may not be able to fully appreciate the long-term implications of addiction. Decisions, then, should be made in their best interests. Further, children may risk developing chronic health complications from using ENDS, and are—due to their developing brains—more likely to develop nicotine addiction than adults. Since no long-term evidence on the safety of ENDS is available, a ban on the supply of all ENDS to children would be justified, particularly bans on the most addictive (nicotine-containing) varieties. Since children are highly impressionable, there should also be tight restrictions on all ENDS-related marketing practices. This should include bans on particular varieties of ENDS designed to target children (eg, bright colours and flavours that mimic sweets); certain product placements (eg, near sweets); and any other marketing designed to trigger a curiosity among children to try ENDS.


It may be argued that a liberal approach to ENDS policy entails simply permitting a free market for ENDS, with the fewest possible restrictions on their supply, marketing and use. This follows the negative conception of freedom, in which freedom is created through minimal state interference. This paper argues that this interpretation of liberalism is not sufficiently nuanced and inappropriate for the context of nicotine-containing ENDS, due to their addictiveness and the misinformation on ENDS which stems primarily from TI marketing activity. With minimal regulation, people otherwise not addicted to nicotine might experiment with these products and develop nicotine addiction. People might initiate the use of ENDS on the assumption that their safety has been tested and proven. Smokers would engage in dual use on the mistaken idea that this is beneficial for long-term health. Others would be involuntarily exposed to passive ENDS vapour, even though it is not yet known whether this is safe.

A ‘liberal’ policy can only truly be considered as freedom-promoting if it provides options and supports individuals’ abilities to make informed, autonomous decisions regarding their lives. Hence ‘minimal state’ approaches, which fail to adequately recognise the autonomy-undermining effects of nicotine addiction and the importance of having access to accurate, impartial information, do not necessarily promote freedom in a broader sense. A liberal approach to ENDS policy, then, should consider freedom in its positive sense and focus on creating options, and on creating meaningful options and ensuring that people are provided with the necessary information. In this view, a liberal ENDS policy actually requires a set of restrictions: to protect people from the misleading marketing practices of the TI, from developing addiction and from the potential health threat of passive ENDS vapour exposure.

Accordingly, this paper has made several recommendations for future ENDS policy. It was suggested that tight restrictions should be implemented on all ENDS-related marketing practices, particularly those targeted at children, those that may trigger cigarette cravings in nicotine-addicted smokers or ex-smokers, and those that are potentially misleading (eg, claims that ENDS are ‘healthy’). Vaping in indoor areas should be restricted to protect the health of others, at least until consistent, long-term evidence can prove their safety. Finally, it was suggested that smokers could be provided nicotine-containing ENDS through a prescription system. This would protect non-smokers—including children—from developing an addiction to them, while enabling smokers to make an informed choice to use a less harmful alternative.


I would like to thank Adrian Carter, Ben Capps, and two anonymous reviewers for their helpful feedback.


View Abstract


  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

  • i ENDS are electronic devices that heat a nicotine fluid and deliver nicotine to the lungs in a similar way as cigarettes. Although ENDS vary considerably in their ingredients and design, in most the fluid contains nicotine, propylene glycol, and other chemicals such as additives, flavourings, or contaminant by-products. They do not contain tar and are not burned like cigarettes, so vapour is only released through exhalation (by the ENDS user).

  • ii Countries such as Singapore and Argentina have banned ENDS completely. In Canada and Australia, ENDS require formal approval as a medicinal product. In the USA, ENDS remain largely unregulated although the FDA is assessing new regulations for ENDS. The UK has adopted a ‘middle ground’ approach by regulating some forms of advertising for ENDS under the new EU Tobacco Products Directive (which also covers ENDS). See ref. 3.

  • iii These may come in the form of nicotine substitution (e.g. nicotine patches, inhalers, lozenges); medicines (e.g. varenicline, bupropion); or non-medical therapy (e.g. cognitive behavioural therapy).

  • iv ‘Freedom’ and ‘liberty’ are used interchangeably in this paper.

  • v Or the ability to resist the use of some other nicotine product, including ENDS. It should be noted that nicotine addiction does not restrict one's ability to switch from tobacco to ENDS or to different brands of tobacco/ENDS; it simply makes it more difficult for one to completely quit all nicotine use.

  • vi Although the effects of addiction on autonomy are widely disputed, drug addictions are usually diagnosed through behavioural criteria that reflect some loss of control regarding drug use, for example, failed attempts to quit or continued use despite problems in other aspects of life, for example, work, family. In this paper, it is not assumed that addiction negates autonomy, but it is assumed that, to some extent, it does narrow one's ability to resist drug use.

  • vii Based on a survey of over 8,000 smokers in Canada, the USA, the UK and Australia. See ref. 23.

  • viii For the purpose of this analysis, ‘harm’ refers to potentially serious, physical harm. For example, exposing others to passive cigarette smoke would not be considered unethical on the basis that it is merely ‘unpleasant’, but it would on the basis that repeated and prolonged exposure can result in chronic disease.

  • ix It is noted that, with this approach, medical licenses may only be awarded to a small range of ENDS products which limits the range of options for smokers. However, it would give smokers access to a range of nicotine alternatives that have been tested and for which the contents and potential health risks are better characterised. This arguably promotes the notion of informed choice to a greater extent than, say, having access to over 400 varieties for which the contents and potential short-term and long-term risks are unknown or not properly communicated.

Linked Articles

Other content recommended for you