Back to the index


Table of Contents

Killjoys A Critique of Paternalism


The author compares soft (nudge) and hard (coercive) paternalism. He describes the characteristics of modern 'public health' paternalism and why he thinks it is bad, then finishes by giving some suggestions on better ways of regulating 'sin' products.

The author is generally fine with nudges (as long as they conform to the limitations laid out by Thaler/Sunstein) but dislikes any attempt to make a behaviour less attractive beyond the point necessary to account for externalities.

The author worries that modern public health paternalism considers its duty to restrict any behaviour that harms a significant number of people (where harm is measured by lifespan, ignoring any pleasure received from the behaviour). It has lots sight of its real responsibility, which is to protect from health risks that cannot be avoided by the individual. For example pollution or Ebola.

The focus is on the UK (especially the use of tobacco, alcohol and 'unhealthy food' in the UK) but the principles apply to most countries.

The author's viewpoint is essentially that of Mill in 'On Liberty', that is: "you should be free to do what you want, as long as it doesn't harm others – oh, and making people feel uncomfortable doesn't count as harm.". Disclaimer: My own views are similar to the author's.

I noticed a reference to a study from this infamous researcher:

Even putting a light above the fruit bowl can significantly increase the number of people who opt for fruit in school canteens (Wansink 2015).

which is an excellent reminder to actually check studies referenced in a book, check for other studies that might contradict the first study, and generally look at the research in case the author is ignoring everything that disagrees with their point of view. (I am not claiming this is what happened here)

Interesting points that would be worth thinking about in more detail:

  • At what point does something become harm? Clearly physical harm is not the only kind of harm, and mental harm exists, but I think Mill is correct that allowing anyone to claim harm when someone else does something that makes them feel uncomfortable will not work well. There seems to be a gradient between 'direct physical harm' on one end and 'I don't like knowing that gay people exist' at the other, and knowing where to draw the line is difficult.
  • How do we handle addiction? If you argue that addiction overrides someone's real desires, and prevents them from maximising their utility, then restricting addictive products more might make sense. Presumably people know that there products are addictive when they begin using them, and are making a choice which is rational at that time knowing that their preferences will be changed. Preventing children from using addictive products may still make sense, since we generally consider them to be not quite rational yet.

One argument that I have thought about for a while is the existence of the NHS. Since my taxes pay for the NHS, and any increase in NHS costs will result in an increase in my taxes (or a reduction in benefits, or a reduction in my children's benefits), then this essentially turns behaviour that would otherwise only pose a risk to the individual into a negative externality that affects everyone. It gives all tax payers a reason to want to control others' behaviour.

My personal bias means I conclude from this that we should no longer have the NHS, the author offers another way out: he suggests that the cost to the NHS of heavy drinkers and smokers is actually lower than average due to their lower life expectancy. (I have not checked these studies yet, but it seems feasible).

Either way, it does not seem defensible on economic grounds to set the tax on tobacco so high as it currently is. Tobacco duty brought in 8.8 billion last year. The duty (ie. sin tax) alone, excluding VAT, is currently 16.5% of the retail price plus £4.57 on a packet of 20, more than half the cost of an average pack of cigarettes is tax. The duty alone on wine is £3 a bottle (again, more than half the price for cheap supermarket wine). Even if smokers cost more to the NHS than they save by dying early, the tax is surely too high if the only goal is to pay for this externality they are imposing on other NHS users.

Here we are ignoring the pleasure smokers get from smoking, and the displeasure others get from seeing and smelling smokers, because it is hard to calculate. If we did try and do the calculation, I imagine smokers get more pleasure than others get displeasure.

The true reason for the tax is not really hidden, it is an attempt to protect smokers from themselves mixed with a little bit of moral puritanism. "Smoking kills, also it is a dirty habit." The moral side is obvious from a lack of lobbying groups trying to stop people from:

  • skiing
  • mountain climbing
  • running (lots of injuries here!)
  • not running (exercise is the best way to extend your life-span)
  • being fat (although this is starting to change!)
  • eating meat
  • not eating vegetables
  • not taking vitamin D supplements
  • etc.

'Sugar' is beginning to join the usual sins of gambling, drugs, cigarettes and alcohol, but is not quite at the point where it is acceptable to shun people who eat donuts in public.

Overall I enjoyed the book, since it made me clarify my own thinking on some of these issues. If you haven't read Mill's On Liberty, do that first. If you don't understand how someone could possibly be against banning smoking, then this book might be an interesting insight into a different point of view.


Paternalism and liberalism

loc: 95 Every day, people do things of which others disapprove. They do things that might seem unwise or immoral. They do things that are unhealthy or dangerous. They do things they might regret. This is a book about what happens when the government tries to stop them.

loc: 97 government paternalism has switched its focus from public morality to public health. Religion has lost its hold over politics. Free speech is far from absolute but blasphemy laws are no more and it is half a century since theatrical productions had to be approved by the Lord Chamberlain. Today, paternalist or ‘nanny state' regulation attempts to reduce the consumption of legal products that can have a negative effect on the health of the user if consumed in excess or over a period of many years. The usual targets are alcohol, tobacco, ‘junk food' and sugary drinks, with e-cigarettes and gambling products sometimes thrown into the mix.

loc: 102 typical policies including sin taxes, marketing bans and sale restrictions, all aimed at curtailing what paternalists call ‘the Three As': Affordability, Availability and Advertising. Mandatory product reformulation, graphic warnings, bans on branding and minimum pricing are also part of the armoury.

loc: 128 Mill himself described his golden rule as ‘one very simple principle' but generations of scholars have found it to be anything but. There is limitless disagreement about the meaning and application of the harm principle. Yet its fundamental idea – that government is justified in protecting people from others but never from themselves – resonated in Victorian Britain and still resonates today. To a large extent, it is this belief that distinguishes liberal democracies from states which require the individual to be subsumed by the religious, collectivist or nationalist beliefs of their rulers.

loc: 144 it is clear that he did not intend it to be defined so broadly as to include the psychological impact of taking offence, feeling sad or being bereaved. If we were to include such emotions as harm, it would allow far more government intervention than Mill would have countenanced. The mere knowledge that an irreligious or risky activity is taking place somewhere in the world could be enough to distress a moral puritan. When Mill wrote about harm, he meant only direct harm to an individual's person or property.

Can be tricky to draw this line

loc: 167 Mill feared that they would stop thinking at all, until ‘by dint of not following their own nature they have no nature to follow' (ibid.: 126). It might also be argued that a society that bans so much on grounds of safety lulls individuals into believing that everything that is legal is safe; that legality itself amounts to tacit encouragement (Miller 2010: 152). In this way, paternalism hinders our ability to make good decisions, first by giving us too little practice and then by giving us unrealistic expectations.

loc: 188 The dominant and recurring theme in On Liberty is Mill's belief that Britain was sinking under a tide of conformity and ‘collective mediocrity' (Mill 1987: 131). An exceptional and unorthodox individual himself, Mill argued that the average man has average tastes and little sympathy for free thinkers and non-conformists. If left unrestrained, he saw democracy becoming nothing more than a vehicle for the prejudices of the masses. ‘The likings and dislikings of society, or of some powerful portion of it,' he wrote, ‘are thus the main thing which has practically determined the rules laid down for general observance, under the penalties of law or opinion' (ibid.: 66). He berated his fellow intellectuals for spending time discussing ‘what things society ought to like or dislike' instead of asking the more fundamental question of whether society's ‘likings and dislikings should be a law to individuals'

The classical economist's view

loc: 199 Mill is quick to note that there is no objective measure of what is best for an individual, but so long as the person is equipped with ‘a tolerable amount of common sense and experience' we must assume that the life he has chosen for himself, within the constraints of his own circumstances and abilities, is more to his liking than the life that would be chosen for him by a committee, a king, or his peers (ibid.: 132–33). Only through liberty, therefore, can the individual maximise his utility.

loc: 235 But there is a major caveat to classical economists' laissez-faire approach to lifestyle regulation. If there are market failures, government action can be justified – so long as it will lead to better outcomes. The aim of such regulation is not to change people's behaviour, let alone their preferences. It is not designed to make people healthier or to make them better citizens. It is designed only to ensure that resources are allocated as efficiently as possible given consumers' preferences. Relevant market failures for our purposes include those which create information asymmetries, such as inaccurate labelling and false advertising, and negative externalities, such as financial costs forced onto unwilling third parties.

loc: 254 some paternalists argue that findings from behavioural economics prove that people are intractably irrational and, therefore, require more government paternalism than has traditionally been assumed. The following chapters will discuss the philosophical and economic arguments in favour of ‘nudge' (soft) paternalism, coercive (hard) paternalism, and ‘public health' paternalism.

Soft paternalism and nudge theory

loc: 262 There are not enough hours in the day for us to be perfectly informed about every decision we make and so we use shortcuts (heuristics) to help us reach an outcome that might not be perfect, but is good enough (satisficing). We use rules of thumb and best estimates. We rely on recommendations from friends, and trust brands that have served us well in the past. ‘It makes far more sense to say that people display bounded rationality than to accuse them of “irrationality”' says Sunstein (2014a: 11).

loc: 305 Even putting a light above the fruit bowl can significantly increase the number of people who opt for fruit in school canteens (Wansink 2015).

Oops. Wansink. Ig nobel winner. Many retractions. Most famous for being bad a science.

loc: 311 nudge must be evidence-based (‘alters people's behaviour in a predictable way'), cannot involve bans (‘without forbidding any options') and cannot make the activity less enjoyable or more expensive (‘[without] significantly changing their economic incentives').


loc: 345 the Behavioural Insights Team was set up under David Cameron. Popularly known as the Nudge Unit, it began life with the commendable pledge to close itself down if it did not produce a tenfold return on its £500,000 start-up costs. In his book Inside the Nudge Unit, the team's director David Halpern describes a string of nudging successes which only serve to demonstrate the limits of ‘libertarian paternalism'. The most significant of them include adding a note to income tax reminders telling the recipient that ‘most people pay their tax on time'; adding a photo of the driver's car to unpaid car tax bills; sending debtors a text message to tell them that the bailiffs are due to appear on their doorstep; and offering people a loft clearance service to increase uptake of subsidised home insulation (Halpern 2015: 3–4).9 All these nudges had the desired effect and, Halpern says, brought in tens of millions of pounds.

loc: 365 One can only speculate as to what legislative programme would emerge if a society was started from scratch based on nudge theory, but it would surely be more libertarian than any country currently in existence.

Book idea!

loc: 392 Mill does make one exception to his principle, albeit for a self-regarding action that is extremely rare if not non-existent. He says that nobody should be allowed to sell themselves into permanent slavery. Feinberg (1971) argues that he was wrong to do so.

Coercive paternalism

loc: 394 Few people are prepared to define themselves as coercive paternalists. An exception is Sarah Conly, whose 2013 book Against Autonomy: Justifying Coercive Paternalism is unusual in making the case for the nanny state openly and honestly.

loc: 406 At the core of Mill's view of individual liberty is the belief that people are best placed to make their own decisions because they have both the incentive and knowledge to make the optimal choices given their own preferences. In Principles of Political Economy, he wrote that ‘people understand their own business and their interests better, and care for them more, than the government does, or can be expected to do' (Mill 2004: 282).

loc: 418 A means paternalist is interested in helping people achieve their own goals whereas as an ends paternalist dictates what their goals should be and uses government intervention to direct people towards them. Sunstein and Thaler present themselves as means paternalists. They accept that people have different goals and preferences, but argue that individuals would make different – and probably better – decisions if the choice architecture was not working against them.

loc: 432 Her examples of objective well-being are uncontroversial on the face of it. It is trivially true to say that people would sooner be healthy than sick and would prefer to be affluent in old age than poor. These things are clearly important. The problem is that other things are also important and trade-offs have to be made. It would be a strange person who chose to live in poverty when they were young in order to be wealthy in retirement, and few people are prepared to live a life of austere self-denial to minimise every health risk.

Sounds like FIRE people

loc: 487 By contrast, he says that many ‘ultra-hazardous activities', such as bull-fighting, sports-car racing and mountain-climbing, should not be prohibited because there are ‘some risks – even very great ones – which a person is entitled to take with his life' (ibid.). But which ones? Climbing Everest has a death rate of one in fifteen and is significantly more dangerous than driving without a seat belt. We cannot know what goes through the minds of mountain climbers as they freeze to death on K2 (as one in five of those who attempt to reach the summit do). Perhaps some of them genuinely reflect on the fact that they are dying doing what they love best, but many of them may wish they had been restrained from attempting the climb.

loc: 495 That these exemptions tend to reflect the public mood of the day only confirms Mill's fears about the tyranny of the majority. Smoking and eating dominate both Against Autonomy and Sunstein's Why Nudge? as if they were in a separate class of risky pursuits. When it comes to activities that pose an acute risk of death at a young age, such as motorcycling and mountaineering, paternalists have little to say other than that participants should, perhaps, be forced to wear a helmet. There must be a suspicion that dangerous sports get a free pass because they are seen as daring, unusual and physically demanding whereas drinking, smoking and drug taking are undemanding, common and intoxicating.

Of course, they did not work from first principles, they started with vague feelings of right and wrong and attempted to justify them.

loc: 517 Slippery slope arguments are, strictly speaking, logical fallacies. In principle, the same arguments for banning opium can be used to ban alcohol, but there is no reason to assume that one will inevitably follow the other. Each policy can be debated on its own terms. However, the existence of one law makes people more likely to accept a similar law based on the same logic. Conly freely accepts this when she observes that laws which ‘may at first seem extreme, relative to the norm, can come to be seen as the status quo, which enables a step to what was considered extreme now appear moderate, and thus acceptable, regardless of merit' (Conly 2013: 115).

loc: 527 Seat belt laws are mentioned a great deal by advocates of paternalism. Along with motorcycle helmet laws, they represent a widely accepted precedent (in Britain, at least) for legislation against victimless crimes. This was always the fear of liberals. When the House of Commons debated seat belt legislation in 1979, none who opposed it denied that wearing a seat belt improved safety. Their concern was that such paternalistic legislation would become a runaway train. As one MP, Ivan Lawrence, said (Hansard 1979): Why should anyone be forced by criminal sanction not to hurt himself? That was never, at least until the crash helmet legislation, a principle of our criminal law. Where will it end? Why make driving without a seat belt a crime because it could save a thousand lives, when we could stop cigarette smoking by the criminal law and save 20,000 lives a year? Why not stop by making it criminal the drinking of alcohol, which would save hundreds of thousands of lives? In response, John Horan MP argued that ‘to regulate in these areas of smoking, sports, and so on, is to regulate people's pleasures and enjoyment. The Bill is really not the herald of some new era of prohibition, or something of that kind. To claim that it is is really too much' (ibid.). And yet the seat belt legislation has been cited ever since by parliamentarians seeking to justify everything from mandatory cycle helmets to water fluoridation to banning smoking in public places and private vehicles (Hansard 2004, 2006, 2014).

loc: 545 They did not oppose the legislation because they doubted it would ‘cut the number of deaths'. They opposed it because it was a minor infringement of liberty that would probably lead to major infringements. The very fact that politicians continually cite the seat belt law as an accepted precedent for further hard paternalism shows that their concerns were well founded.

loc: 590 The reality is that hard paternalists punish people for their lifestyle choices and require a minority or even a majority of their fellow citizens to sacrifice their welfare and liberty for the good of others – even when others could achieve the same outcomes (such as not smoking, staying slim, avoiding drugs) without everybody else being forced to submit to the same laws. In the final analysis, coercive paternalists not only breach the harm principle, they actively cause harm.

Neo-paternalism: an assessment

loc: 667 In the experiments above, the participants were given a straight choice. They did not have to pay for their food and films. There was nothing to sway them in the choice architecture, no nudging, no default option. Given that they opted for chocolate and Sleepless in Seattle, it would take a leap of faith to conclude that what they really wanted was celery and The Piano. Yes, they chose healthy food and highbrow films for their future selves, but putting something off until tomorrow is only one step removed from not doing it at all. At best, these experiments show that people know what an idealised version of themselves ought to do. Awkwardly, though, they also show what people really want to do.

loc: 697 Conly says that ‘the goal of paternalistic legislation is to allow us to be more like ourselves' (Conly 2013: 88). But which self? She does not mean us, the flesh and blood human beings reading her book. She means the people she imagines we would want to be if we were more ‘rational'. But to the paternalist, ‘rational' is a euphemism for their own tastes, prejudices and desires, giving the illusion of objectivity to fundamentally subjective preferences.

Be more like an 'ideal' Human, not like the actual self

loc: 715 Elsewhere in her book, Conly denies that her agenda is elitist but, as this quote suggests, coercive paternalism is not a collective effort in which every member of society recognises their shared fallibility and asks for a benevolent government to restrain them for their own good.

Otherwise it would be opt-in, like beeminder. Sting me if I buy cigarettes

‘Public health' paternalism

loc: 770 In their day-to-day political activities, ‘public health' paternalists rarely attempt to justify their position on ethical grounds, preferring instead to talk about ‘evidence-based policy' (see Whyte 2013). This keeps the conversation on the consequentialist turf of ‘does it work?' rather than opening up the question ‘is it right?', but their consequentialism is of a narrow sort. If a policy is believed to prolong life or curtail risky behaviour, then it ‘works' and becomes ‘evidence-based' per se. Other consequences are largely ignored, including the implications for people's welfare, unless they directly affect health.

loc: 793 The case for government action on public health issues is strongest when there is a threat to health that can only be countered by collective action. Collective action does not necessarily mean government action, but if the term ‘public health' meant anything in Mill's day, it meant tackling health risks in the shared environment which cannot be controlled by the individual, such as air pollution, or those involving people (or animals) who carry infectious diseases. Factories pumping coal smoke into a congested city and travellers coming home with Ebola pose a clear risk of unavoidable harm to others and are therefore a potential justification for coercion under the harm principle. It is not the scale of the risk nor the number of people affected that turns a health problem into a public health problem. It is the lack of consent from those who are put at risk and their inability to reduce the risk without collective action.

loc: 801 the modern ‘public health' movement ‘treats any health issue as one of public health so long as it affects large numbers of individuals'. This has led to a focus on personal habits which are risk factors for non-communicable diseases, such as heart disease, diabetes and cancer. ‘Our public health problems are not, strictly speaking, public health questions at all', write Gostin and Gostin (2009: 220). ‘They are questions of individual lifestyle'.

loc: 808 Obesity is now routinely described as a ‘public health epidemic' despite being neither contagious nor a disease. Smoking, drinking and even gambling have been brought under the umbrella of ‘public health' and are also referred to as ‘epidemics' despite having such long histories that they are the very opposite; they are endemic. An epidemic of non-communicable disease is an oxymoron, but the use of such terms encourages the public to tolerate the same level of state intervention in matters of lifestyle as they would if they were in the midst of a viral outbreak.

loc: 895 While the individual engages in a risky activity knowing that there is only a chance of being harmed by it and that death can only be postponed, never prevented (and, if he is statistically literate, that reducing risk from one disease necessarily increases the risk of other diseases),

Huh, true I guess

loc: 907 Breast cancer is much more common, affecting one in eight British women and yet the increased risk from light drinking is just 5 per cent, thereby turning an absolute lifetime risk of around 12.5 per cent into a lifetime risk of around 13 per cent. To the individual, these are little more than rounding errors. Few, if any, of us would sacrifice the pleasure of drinking alcohol in response to such a negligible health risk. It is for that reason that paternalists prefer to give us big numbers (‘10,000 deaths a year') or relative risks (‘17 per cent more likely') than to tell us the absolute risk.17 Binge drinking may increase our risk of liver cirrhosis sixfold but, as Allmark (2006: 4) notes, this statistic is ‘virtually useless for rational decision-making' unless we know what our odds of getting liver cirrhosis are to begin with (about 1 in 5,000 in any given year, in case you're wondering).

The politics of ‘public health' paternalism

loc: 945 By denying or ignoring the private benefits individuals receive from the consumption of tobacco, alcohol and ‘junk food', they bypass the need for cost–benefit analyses. If there are no costs incurred by suppressing these habits, suppression can only be positive. Even the most obvious economic consequences of their policies, such as the regressive impact of sin taxes or the damage done by smoking bans to the hospitality industry, are countered with denial and sophistry. For political reasons, ‘public health' policies must be presented as win–wins.

loc: 1,031 The perception that drinkers, smokers and the obese impose significant costs on others appears to validate ‘sin taxes' and, in Britain, the ‘cost to the NHS' argument is perhaps the most intuitively persuasive justification for lifestyle regulation. In a healthcare system funded by involuntary contributions, one person has to pay for another person's healthcare. This is an ‘induced externality' because state intervention has laid the foundations for externalities to thrive, thereby creating demand for further state intervention (Wiley et al. 2013: 89). If the government compels citizens to pay for one another's education, healthcare, pensions and welfare benefits, there will always be someone who can complain that somebody else has made them worse off.


loc: 1,042 As Le Grand and New (2015: 61) acknowledge, ‘the taxes on tobacco and cigarettes are more than sufficient to cover all the additional medical care costs incurred by smokers.' In fact, it is far from certain that smoking and obesity impose any financial costs on others once savings to the taxpayer are taken into account. Few wish to admit it openly, but premature mortality after the age of 65, when individuals take more out of the system than they put in, saves the government large sums of money in pensions, welfare and social care. This is significant because it is the net external cost, not the gross external cost that should be used to calculate a Pigouvian tax.

If so, this takes the teeth out of the NHS argument

loc: 1,058 The same principle applies to the intangible cost of a year of life. Monetary valuations of a life-year vary enormously and can only be arbitrary. Whatever estimate is used, it should be obvious that the benefits of being alive and, therefore, the costs of being dead, fall squarely on the individual.

When calculating externalities from smoking and drinking

loc: 1,061 it is wrong to include income tax that is ‘lost' when a person dies as a cost to society. People who take early retirement or choose to work part-time are not portrayed as a burden on others, and for good reason: failing to confer benefits on others is not the same thing as harming them.

loc: 1,066 The mere knowledge that an activity is taking place can be enough to cause psychic damage to some people. As Mill noted in On Liberty, the very existence of alcohol offended the prohibitionists of the UK Alliance and violated their vaguely defined ‘social rights'. Taken to its logical conclusion such extreme sensitivity could only result in making ‘all mankind a vested interest in each other's moral, intellectual, and even physical perfection, to be defined by each claimant according to his own standard'

Emotional externalities? Where to draw this line is an interesting question

loc: 1,070 emotional benefits to third parties must be counted alongside the costs. There are some positive externalities from drinking, smoking and overeating, such as those of the non-drinker who appreciates his local pub, but these are no easier to quantify than the costs. Most externality analysis therefore tends to be partial by construction and extraordinarily difficult to calculate.

loc: 1,074 For now, it is enough to observe that the sin taxes of ‘public health' paternalists differ fundamentally from the Pigouvian taxes of economists. Paternalists are not interested in finding the optimal level of consumption because they believe the optimal level to be zero.

Pigouvian would aim only to internalise the externalities (ie. reach the true equilibrium), not try to change demand

loc: 1,111 Parmet (2014) argues that tobacco marketing is not a self-regarding behaviour and Van der Eijk (2015: 3) accuses e-cigarette advertisements of being ‘potentially autonomy-undermining' on the questionable assumption that they trigger the desire to smoke. But if advertising undermines autonomy, it does so no more than any other form of free speech. It is no more ‘autonomy-undermining' than an invocation to stop drinking or lose weight, and since paternalists do not want to ban those messages we must conclude that it is the product being promoted, rather than the threat to personal autonomy, that they really object to.

loc: 1,133 Companies spend money on advertising because it encourages brand loyalty, increases the value of brands and encourages consumers to try new brands. In short, they use it to fight for market share, not to enlarge the market.

This responds to the argument "If advertising tobacco didn't make people smoke, those big companies would not do it" – they do it because they have competitors in the same market, not necessarily to expand the size of their market

loc: 1,156 It is said, for example, that teenagers who see a lot of alcohol advertising drink more alcohol (Snyder et al. 2006). It is also claimed that children who see e-cigarette advertising are more likely to become e-cigarette users (Singh et al. 2016) and may even be more likely to become tobacco smokers (Petrescu et al. 2016). All these studies have one obvious and fatal flaw. They do not measure how much advertising teenagers actually see, only how much they recall – sometimes many years later. It should not be surprising that a drinker pays more attention to alcohol advertisements than a teetotaller and that a heavy drinker recalls more alcohol brands than a light drinker.

Presumably there is some effect though

loc: 1,208 It is therefore true that the majority of smokers express a desire to quit, albeit with less fervour and in smaller numbers than is often believed. But stated preferences are unreliable, particularly when there is strong social pressure to conform. A smoker who says he wants to quit may be expressing a second-order preference (that is, he wishes he was the sort of person who didn't want to smoke) or he might be saying what he knows he should say in a society that disapproves of his habit.

loc: 1,230 We might take a more sympathetic view and acknowledge that many smokers genuinely struggle to give up and that their autonomy is undermined to some extent by nicotine's addictive properties. But taking a more sympathetic view does not mean giving more weight to vague aspirations than to revealed preferences. If people really wanted to give up smoking, we would expect them to express this desire strongly and make regular attempts to do so. But an Office for National Statistics survey found that only 22 per cent of smokers said they wanted ‘very much' to quit and only 26 per cent had made an attempt to quit in the past year (ONS 2009). Making at least one attempt to give up smoking in a year would seem the minimum requirement for someone who really wishes to quit. And even if we give credence to stated preferences, we would expect smokers who ‘very much' want to quit to say so. On either of these measures, the proportion of smokers who have a strong desire to quit appears to be closer to one in four than nine out of ten.

Real addiction is another slightly sticky point, another interesting question to think about

loc: 1,290 Paternalistic measures are employed in public health because public health wants a compliant population, not an informed one. When accurate labelling does not ‘work,' advocates move on to more coercive measures. Cigarettes have been appropriately labelled with increasingly dire warnings. The warnings have evolved from informing smokers that cigarettes ‘may cause cancer' to current warnings that say ‘cigarettes kill' and cause a variety of serious diseases and conditions. If we wished to have an informed smoking population, we would measure what smokers know about the risks of smoking. But success in labelling is not measured by what smokers know about the risks. Success is determined by the number of people who stop, or do not start, smoking. The goal is to control behaviour and have people do what we think is best for them.

loc: 1,315 Similarly, cigarette packs could be used to tell smokers that they will live nearly as long as people who have never smoked if they quit before the age of 40, but as this would not deter – and could even encourage – young people to take up the habit, this factually correct information is shelved in favour of the simple message that ‘Smoking Kills' (Jha and Peto 2013). Smokers and non-smokers alike tend to overestimate the risks of smoking. If they had a better grasp of the statistics, there would probably be more smokers (Viscusi and Hakes 2008) but, for obvious reasons, health campaigners have not attempted to correct that particular information deficit.

The consequences of hard paternalism

loc: 1,416 Affordability and availability do not drive consumption. They allow it.

loc: 1,444 There is, therefore, a ravenous appetite across government for taxes that are politically acceptable and do not cause too much economic disruption. Sin taxes fit the bill and politicians can rely on a small army of campaigners in the ‘public health' movement to lobby for them. Thanks to the pervasive, though mistaken, belief that smokers, drinkers and the obese are a drain on public resources, sin taxes are perceived to be equitable and so provide a rare opportunity to tax poorer groups in society, including pensioners, the disabled and the unemployed, without causing outrage.

Because they need to find money to pay pension entitlements

loc: 1,451 Commonly cited own-price elasticities are –0.79 for soft drinks (Andreyeva et al. 2010), –0.46 for beer (Wagenaar et al. 2009) and –0.48 for cigarettes (Gallet and List 2003). Caution is recommended when dealing with such estimates as they vary enormously between studies but, on the face of it, these figures suggest that a 10 per cent increase in price reduces consumption of these products by 7.9 per cent, 4.6 per cent and 4.8 per cent respectively. Sin taxes therefore reduce demand somewhat – and so fulfil their ostensible justification – but not enough to lower total revenue.

Tobacco "hit the peak off the laffer curve in 2012" so increases now drop revene. (it is very, very high currently, large proportion of the cost of cigarettres, like 70% or there abouts)

loc: 1,457 Alcohol duty (including the VAT on the duty) brought in £12.8 billion in 2015/16 (ONS 2016c) – the equivalent of nearly £500 per household. Tobacco duty (including the VAT on the duty) amounted to £11.4 billion, meaning that the UK's 9.6 million smokers were spending an average of £1,188 a year on this one tax (ONS 2016b).


loc: 1,472 The conspicuous failure of food and drink taxes to reduce obesity rates casts doubt on the first assumption and the high rates of smoking among the poor after decades of rising tobacco taxation cast doubt on the second (Hiscock et al. 2012).

Presumably low ability to trade short term pleasure for long-term benefit corallates with beng poor?

loc: 1,490 Putting a monetary figure on pleasure can never be an exact science, but the basic principle of counting lost consumer surplus should not be controversial. If consumers are rational, any increase in price is bound to reduce their consumer surplus. If they are biased or irrational in some way, the calculation should be adjusted to account for this, but the existence of bias does not justify ignoring consumer surplus altogether.

ie. pleasure of smoking

loc: 1,497 In fact, the FDA had more reason to include lost pleasure when looking at tobacco regulation than when it assessed calorie labelling. If calorie labelling makes a person change what they eat in a restaurant, it is because their previous choice was more fattening than they realised. Their revised choice better reflects their true preferences. Calorie labelling might create financial costs for the restaurant and it might create psychological costs for those who feel bad about ordering the least healthy option, but it does not create a cost for those who change their order as a result of being better informed. By contrast, if an individual is coerced into abandoning a pleasurable activity through bans, taxation or misleading information, he might enjoy a health benefit but will also incur a loss of utility (i.e. pleasure). This is a real cost and should be counted.

loc: 1,503 If the coercion fails and the individual continues to engage in the risky pursuit, he receives no health benefit and no welfare benefit. For example, a person who does not quit smoking after a smoking ban is introduced has to start smoking outdoors, often in unpleasant weather, enjoying himself less but incurring the same costs to his health. This is a net cost to the individual and such deadweight losses are the norm when ‘public health' interventions are made.

loc: 1,580 Others have argued against the use of stigma more forcefully. Scott Burris, a professor of law, describes stigma as ‘a barbaric form of social control that relies upon primitive and destructive emotions … a liberal society simply ought not to be in the business of shaming its citizens' (Burris 2008).

What about shaming bad parents? Surely works, and protects children, who cannot yet protect themselves

loc: 1,594 Anti-smoking messages such as the NHS television advertisement which ended with the words ‘If you smoke, you stink' are clearly directed at the person rather than the product.


loc: 1,601 There is, however, a tension between the desire to stigmatise individuals and the strategy of blaming industry for people's lifestyle choices. If industry is to blame, then the consumer is a victim – and it is hardly fair to demonise victims. This is the line taken by most paternalists in the field of obesity who blame the food industry, government and the ‘obesogenic environment' rather than the person who eats too much or exercises too little. Complaints about ‘fat-shaming' seem to be at odds with the active ‘denormalisation' of smokers, but this may reflect the fact that the anti-smoking campaign is at a more advanced stage than anti-obesity efforts.

loc: 1,632 As knowledge progresses, the scientific consensus can change. To take one of many examples, parents have long been told to avoid giving young children peanuts to prevent them from developing peanut allergies. Studies have since shown that children are considerably more likely to develop allergies to peanuts if they do not eat them at a young age. As a result, parents are now told to feed their children foods that contain peanuts before they are six months old (NIAID 2017). It would be unfortunate if children have developed peanut allergies as a result of the old advice, but the damage would have been multiplied if health professionals had found a way of coercing parents into following it.

loc: 1,639 Given the risks of government failure and the fact that an individual who chooses to ignore health advice will generally only harm himself, it would be better to permit what Mill called ‘different experiments of living' than to enforce expert opinion upon everyone which, if wrong, would magnify the consequences of error many times over.

loc: 1,663 It would take a moral cretin to miss the distinction between a health hazard which kills people at a young age through no fault of their own and one which shortens life by a few years as a result of lifestyle choices that are freely made. Until the day that infectious diseases are eradicated, diverting resources from the former to the latter is ethically dubious. Even in rich countries, where contagious diseases are relatively rare, it is doubtful whether the money spent on largely unsuccessful attempts to make people change their lifestyles would not be better spent on health and social care. For those who want medical help but are neglected because resources have been shifted towards lifestyle regulation, ‘public health' is a cost.

Towards better regulation

loc: 1,744 Is there a better way? I think there is. With the principles of liberalism laid out in the preceding chapters in mind, I now offer suggestions for how controversial lifestyle products should be regulated in the UK.

loc: 1,773 In the field of lifestyle regulation, only the smoking–cancer link has required a similar educational effort since smoking was so widespread in the mid-twentieth century and so many people were sceptical about the ‘tobacco scare' (as it was sometimes known). In that instance, special warnings on the product could be justified since they provided an official stamp of authority to a health claim that was still contested, not least by the tobacco industry. The warnings were all the more powerful for being unusual. The very fact that no other consumer product was emblazoned with such a warning made the public take it seriously. An argument could be made for further labelling on other products if there is good evidence that (a) large parts of the public are misinformed about certain risks, (b) those risks are meaningful and well-proven, (c) the warnings would make a difference to patterns of behaviour, and (d) it is not possible to effectively transmit the information by other means. In most cases, however, the case for warning labels fails to meet at least one of these criteria.

loc: 1,840 Smokeless tobacco products are little used in Britain but exiting the EU would allow us to start selling snus (see previous chapter). If so, we should not make the same mistake as the US in sending a negative signal by concealing important information. Smoking causes hundreds of thousands of deaths each year in the US whereas smokeless tobacco causes a few dozen cases of oral cancer.

loc: 1,870 Sin taxes can be justified if there are legitimate external costs borne by the state. A pure insurance-based health and welfare system would put an end to claims about unhealthy lifestyles being a burden on the taxpayer but so long as these services are paid for by the general taxpayer, sin taxes should cover the net costs associated with smoking and drinking: that is, the health and welfare costs minus the health and welfare savings. In practice, these taxes would be much lower than at present. As discussed above, the net costs of smoking are negligible, if not negative,


loc: 1,882 EU law currently prohibits taxing alcohol by the unit, but it could be implemented post-Brexit.


loc: 1,883 Obesity is more difficult since it has multiple causes. No single product can be held responsible and physical inactivity cannot be taxed. Fortunately, the question is largely academic since the net costs of obesity are much lower than is often claimed (Tovey 2017).

loc: 1,928 The tobacco industry has saved a fortune from the advertising ban, it is the consumer who has been disadvantaged.

Hah, zero sum game, they benefit from the ban because it allows all of them to coordinate a reduction in marketing costs.

Back to the index

Last modified 2019-08-16 Fri 16:27. Contact