For Your Own Good?

The freedom to pursue informed self-harm has a long and noble tradition….
by Eric Crampton

After the Christchurch earthquakes, the army and civil defence barred business owners from entering their downtown offices. Even if you were prepared to take the risk of entering your own building to retrieve critical files, you weren’t allowed to take on that risk. Officials noted that the risks did not just fall on those entering the building willingly: if further substantial aftershocks meant that you needed to be rescued, that would place a burden on the Search and Rescue teams. Because of the risk placed on emergency workers who would be called on to save anyone trapped in a building, business owners were forbidden entry lest they be trapped.

Pity then the business owner who took the government’s explanation perhaps too seriously. After hiring a private team of qualified Search and Rescue technicians and announcing that he was going into his building to collect his files, and that no burden was placed on public responders as he had hired his own, he was still barred entry. For his own good.

But how can anyone ever really know what is for someone else’s own good? Paternalistic interventions around all kinds of private risk-taking do work to protect people against those risks, but without privileged access into the minds of others, how can anyone tell whether those interventions really are to the benefit of those so-constrained?

Worse, public efforts to improve health outcomes fail to distinguish between negative outcomes of voluntarily chosen risks, and deaths and illnesses consequent to contagious disease. In the midst of an Ebola epidemic in which aid workers in West Africa struggled for resources, World Health Organisation Director-General Margaret Chan attended the WHO’s Framework Convention on Tobacco Control conference in Moscow, where she praised Putin for his robust anti-smoking measures. She later explained her attendance, noting “Ebola is important, but there are other important issues, like tobacco control.”

On a metric that weighs only mortality and morbidity, it would be hard to fault Chan’s decision. Smoking kills far more people than does Ebola. But while public health efforts in ensuring that smokers know the risks they’re taking are laudable, efforts beyond that miss one rather important detail: people can knowingly and willingly take on the risks of unhealthy lifestyles because they enjoy them. Nobody enjoys Ebola. For all the warnings about second-hand smoke and rhetoric about smoking’s being contagious, you do not need to be placed in quarantine for having come into contact with a smoker. Public health efforts to stem communicable disease, and especially those with potential to cause pandemic, cost only the resources needed in research, discovery, and preparation; campaigns against lifestyle diseases also cost us an important part of our freedom.

I argue that modern paternalistic regulation far too often supplants individuals’ visions of the good, as they see things, for the good as judged by the regulator or the public health expert. Many of the common justifications for paternalistic regulation, like costs borne through the tax system, simply fail to withstand scrutiny. We are then left with an official’s vision of the good life constraining or supplanting the equally legitimate visions of others, at a cost to our freedom, and at a cost to our preparedness for pandemics.

I. Paternalistic archetypes: the Carlyle-Mill debates

In the absence of strong evidence to the contrary, economists, myself included, generally expect that, on average and in most cases, people make the best decisions they can given the information available to them and the goals they have. For this we’re sometimes criticised as rationalistic and for expecting too much of humanity’s crooked timber. The criticism does not generally stand: economists’ conclusions do not rely so heavily on fragile rationality assumptions. All we really need is that people do the best they can with what they have, and that the costs of errors from presupposing others’ ends and dictating to them their means outweigh the costs of errors individuals might make on their own. The alternative view holds that individuals generally cannot be expected to find their own paths and walk them; instead, they need nudges and constraints against harmful choices and encouragement of better ones.

British essayist Thomas Carlyle made the case for such constraints in his 1850 pamphlet when he wrote, in opposition to the political economists of his day,

“The true liberty of man, you [economists] would say, consisted in his finding out, or being forced to find out the right path, and walk thereon. To learn, or to be taught, what work he actually was able for; and then, by permission, persuasion, and even compulsion, to set about doing of the same! … If you do know better than I what is good and right, I conjure thee in the name of God, force me to do it, were it by never such brass collars, whips and handcuffs, leave me not to walk over precipices!”

Carlyle’s discussion of collars, whips and handcuffs was not metaphorical: his pamphlet urged America not to abandon slavery, which he viewed as an enlightened paternalistic intervention that furthered the best interest of black American slaves who, in his view, otherwise would never learn the discipline of work and consequently could never become fully human. John Stuart Mill’s opposition to Carlyle’s views on slavery led Carlyle to name economics the Dismal Science.

Mill began from a polar opposite point. In “On Liberty”, he wrote that the state has no business interfering in the lives of individuals except where strong risk of harm to others obtains. He argued this not because we do not care about each other, but rather because that care has to be expressed in ways other than lashes, whether physical or metaphorical. We can persuade and cajole, but we cannot force. He noted that while the drunkard may neglect his family, the harm is in the neglect, not the drunkenness; someone else who is similarly callous towards their family while sober does as much harm. Mill writes,

“No person ought to be punished simply for being drunk; but a soldier or a policeman should be punished for being drunk on duty. Whenever, in short, there is a definite case, or a definite risk of damage, either to an individual or to the public, the case is taken out of the province of liberty, and placed in that of morality or law.”

Mill later presciently argues that if we start counting as harm the distaste one feels about the actions of another, then there is little bound.

“The evil here pointed out is not one which exists only in theory; and it may perhaps be expected that I should specify the instances in which the public of this age and country improperly invests its own preferences with the character of moral laws. … And it is not difficult to show, by abundant instances, that to extend the bounds of what may be called moral police, until it encroaches on the most unquestionably legitimate liberty of the individual, is one of the most universal of all human propensities.”

The basic arguments around the scope of paternalistic regulations have not changed much in 150 years, although, thankfully, slavery is well off the table.

Economists typically start from a Millean perspective. As we do not know the goals that others seek, nor what brings them joy, we must be reticent to interfere with their choices except where those choices impose strong harm upon others. For a paternalistic intervention to be justifiable, it should be welcomed by those subject to its constraints. Less justifiable interventions instead push individuals to fulfil the ends sought by the planner, rather than helping people to achieve their own goals. Carlyle imagined that, were he in the place of American slaves, he would welcome the lash; he does not weigh what the slaves themselves might prefer.

II. Paternalism and market failure

But what of the case in which consumers will reliably make poor choices because of market failures? Suppose that some product carries hidden health costs unknown to consumers; consumers consequently enjoy too much of the product relative to the amount they’d purchase were they fully aware of the health costs. For decades, tobacco was in this kind of position; the health risks of smoking are now very well known. We today instead hear that those buying food at chain fast-food restaurants do not have enough information about caloric content or that people underestimate the risk of eating sugar.

In these kinds of cases, it can be reasonable to suggest providing information about nutrition and health consequences. Consumers then armed with that information should be trusted to make their own choices.

New York City mandated in 2008 that chain restaurant menus carry calorie counts. The intervention was paternalistic: the information provision was at least somewhat costly, and consumers, if they had wanted to know, could have chosen to frequent restaurants providing calorie information voluntarily. But, it was rather less intrusive than taxes or bans. If lack of information had been the problem, consumers would have changed their consumption patterns.

But instead a series of studies showed no effects of calorie benchmarks or nutritional information on consumers’ choices. If anything, consumers had overestimated the caloric content of fast food. The response of obesity campaigners was telling. Rather than concluding that individuals might be making informed choices and weighing trade-offs around food, obesity, convenience and taste, they instead decided that tools other than fast-food menu nutritional labelling were necessary in the fight against obesity. The health lobby’s goal was not to empower people to make their own choices in pursuit of their own ends, it was rather to steer people towards healthier behaviours, whether or not they wanted them.

While it’s possible that other kinds of simplified nutritional labelling could be more informative for some consumers, oughtn’t we consider the possibility that many people simply like eating food that makes them fat and that they prefer being overweight to changing their diet? The right to swing one’s burger ought to extend to one’s own mouth. The health planners too often take their own goals to be the only permissible ones; people can legitimately have goals other than being as healthy as possible.

a. Market failures and fiscal externalities: getting the signs right

What if your burger’s path routes via my wallet? We all pay taxes to cover the costs of the public health system; those choosing less healthy lifestyles impose some costs on others through the tax system. However, this does not provide sufficient justification for barring others’ liberties, for four main reasons.

First, the public health system is compulsory: while you may choose to attend a private hospital or to buy private health insurance if you earn enough money to do so, there is no contract you can sign to opt-out of public healthcare entirely. Consequently, it seems a bit perverse that at least some individuals are offered the following bundle as compulsory option: “In exchange for a health insurance product you might not want, the State will ban, tax, nudge or constrain you against doing many things you enjoy doing that do no harm to anyone else, but that might impose costs on our insurance scheme.” This is not a very good deal.

Second, individuals still have reasonable incentive to look after their health even though the public pays the bill for doctors and hospitals: being ill is not pleasant, so most people try to avoid it.

Next, where costs borne through the public health system form the justification for restrictive policy, it is less than clear what bounds the potential scope of public regulation of private lives. Few things we do are riskless; we trade off potential costs against other benefits all the time. While none of us are perfect in anticipating the exact costs and benefits of our actions, we at least have some incentive to consider both; bureaucrats designing regulations to keep health care costs down might care a bit less about how much any of us enjoy our risky activities, whether risky mountain climbing or risky sodas. Where the metric is illness and injury avoided with no accounting for individuals’ own choices or preferences, a death in a skiing accident is of equal policy consequence to a death from Ebola.

Perhaps most importantly, the costs of smoking, drinking, and obesity to the health care system have been rather strongly overstated. Figures in the billions of dollars are often cited as the “social costs” of smoking, drinking and obesity, but those figures typically tally all of the costs borne by smokers, drinkers, and the obese themselves as being costs to the country in general; costs to the government are generally a small fraction of the headline costs. In fact, while smokers do impose some costs on the health care system, their excise tax payments are almost three times higher than their costs to the health system, or at least according to the O’Dea report commissioned by Action on Smoking and Health and the SmokeFree Coalition. And that’s before accounting for the savings to the superannuation system where smokers do not live as long. Similarly, van Baal and coauthors showed that while the obese rack up heavy costs in obesity-related illness, lifetime health expenditures were the highest for healthy-living people, lowest for smokers, and moderate for the obese: abolishing smoking and obesity would cut the costs of smoking-related and obesity-related illnesses, but they would increase aggregate health care expenditures. Costs, or benefits, to the public accounts are a horrible basis for regulating individuals’ choices. But if we are going to use that kind of metric at all, we should at least get the sign of the effect right.

III. Paternalism’s costs to freedom, and to our health

It might not be the greatest trick the Devil ever pulled, but it’s up there: convince the thinking and caring classes that, because Big Industry exploits the poor, we need the power of the State to protect lower-income shoppers from bad choices. Once that’s done, you can use the State to target all the lower-class behaviours you find distasteful while claiming you’re doing it for their own protection.

Consider a few examples.

Because poorer people are more likely to smoke than richer cohorts, and because poorer people have less disposable income, tobacco excise taxation disproportionately hits lower-income families. In 2007, a report commissioned for Action on Smoking and Health and for the SmokeFree Coalition found that a 20% tobacco price increase would cost Decile 2 smoking households $8 per week, at a time when average weekly disposable income was around $350. Tobacco prices doubled from 2006 to 2014 on several rounds of tax increases; I’m rather sure that Decile 2 incomes have not doubled over that same period. Meanwhile, electronic cigarettes, while much safer than smoked tobacco, are still shunned by the public health establishment. Their adoption in place of cigarettes would reduce mortality and morbidity while allowing the enjoyment of nicotine; I wonder whether it’s that enjoyment, rather than second-hand smoke, that gets up the public health advocates’ nostrils. Nicotine seems a lower-class pleasure that needs to be curtailed, even when it is consumed in fairly riskless ways.

The war on drugs nominally affects everybody, but enforcement hardly seems class-neutral.

New Zealand has not implemented any kind of soda tax, though they are proposed from time to time. When New York City sought to ban large-sized sugary drinks, those with enough milk content were exempt: richer Americans drink relatively expensive milky Frappuccinos; poorer people drink soda. And the best evidence suggests that taxes on soda may reduce soda consumption, but the effects are offset by increased consumption of other products.

Fat taxes also would also disproportionately hit the poor. For meat of equal fat content, an excise tax per unit of fat would do more to increase the price of mince than it would to hit a nicely marbled Scotch Fillet, in proportionate terms. TV news specials on obesity typically pan to below-the-neck shots of larger people in t-shirts, rather than suits, and highlight fast food rather than chocolate mousse. When public discussion turns to mandatory calorie counts for restaurant menus, the cream sauce gets a pass.

Alcohol excise does more, proportionately, to affect the price of Tui and cask wine than it does to increase the cost of a Penfolds’ Bin series or a $9 craft beer.

It is hard then to look at the existing and proposed constellation of paternalistic health interventions and fail to conclude that they largely reflect upper-middle-class anxiety about the characteristics and habits of the poor. While it is true that, were those consumption behaviours changed to reflect the often healthier behaviours of richer cohorts, life expectancy would rise and disability would fall, those behaviours are chosen. It can be useful for public health advocates to provide information about the consequences of those choices, and about alternatives. But individuals have to be free to make those choices. Conversely, would richer cohorts welcome it if Tui drinkers banned our drinking more expensive beers, for the sake of our wallets? The interference is not welcome; it is insulting.

As mortality due to contagious disease dropped in the developed world, public health’s focus shifted to other causes of mortality and morbidity. In the developed world, that often means individuals’ lifestyle choices. But the appropriate framework for dealing with lifestyle choices isn’t the epidemiologist’s rule of Disability-Adjusted Life Years (DALY) it’s the economist’s metric that takes very seriously individuals’ own choices. In one interesting lab experiment, mice kept on a severely reduced diet were found to live longer, but were also far less happy and were more likely to bite their handlers. Both choices can be entirely legitimate; imposing those choices is not.

In other words, not all Disability-Adjusted Life-Year reductions are created equal. If I choose a richer diet and a lifestyle with more exhilarating risks, and a somewhat reduced life expectancy, the DALY reduction is simply of no policy consequence as compared to the DALY reduction if a flu pandemic reaches New Zealand, or the DALY reduction from Ebola in Africa. In the first case, I’m buying fun at the cost of health. In the second, there are only costs. By shifting focus away from diseases that have costs but no benefits towards lifestyle issues where people are making trade-offs, public health organisations like the WHO risk doing real harm. Lifestyle issues like alcohol, tobacco, salt and obesity received 32 mentions in the latest World Health Organisation proposed programme budget. Pandemics received only 22 mentions, and Ebola got none. A dollar spent on obesity control might possibly do more to increase DALYs than would a dollar spent in Ebola preparation, but the former does not count the costs experienced by those who do not welcome the interference in their dinners.

Perhaps it is time that public health advocates return to classic public health problems: encouraging vaccination, preparing for pandemics, watching for outbreaks of contagious disease, and providing information about health risks to help consumers make informed choices.