• public health ethics

Preventive regulations are designed to address obesity by curbing access to unhealthy foods and include measures such as soda taxes and trans fat bans. One argument in favour of preventive regulations is that they correct for what is often seen as a lack of choice. The obese are disproportionately worse off and therefore ‘forced’ to choose less expensive, less healthy foods. A key element of Wilkinson’s case against preventive regulations is that this argument in fact makes two distinct claims. The first is that badly off people have poor options, and the second is that badly off people have problems in how they choose between their options. Wilkinson argues that on neither interpretation of the choice problem do preventive regulations improve equity. In the first instance, where obesity is explained in terms of poor options, preventive regulations are inequitable because they worsen the already poor choice sets of the worst off. The equity case for preventive regulations then hangs, Wilkinson argues, on the notion that badly off people have problems in how they choose between their options, specifically that they choose against goals that they themselves consider more important. In the absence of evidence that obesity is explicable in these terms, Wilkinson argues that this equity case for preventive regulations also fails.

I will argue, with respect to the first of these arguments, (A) that preventive regulations do not necessarily worsen the options of the worst off and (B) that even if preventive regulations do worsen the options of the worst off, they are not necessarily inequitable.

Wilkinson’s first argument for the inequity of preventive regulations can be stated as:

1. It is inequitable (unjust and unfair) to reduce the already poor options/choice sets of the worst off.

2. Preventive regulations reduce the already poor options/choice sets of the worst off.

3. Therefore, preventive regulations are inequitable.

The problem with this argument is with premise 2, the assumption that preventive regulations reduce or worsen the choice sets of the worst off. Wilkinson argues that choice is reduced ‘when people have a worse choice set than before’ and that one choice set is better than another when it has ‘the same options as before plus some good ones or it has different options that the chooser prefers’. These definitions are slippery in the context of preventive regulations. As an example, Wilkinson assumes that a change in options from a cheap soft drink to a more expensive soft drink (eg, via a soda tax) worsens a choice set. Why is the choice set with the cheaper soft drink a better choice set than the one with the more expensive soft drink? It cannot be because it has more options, because the number of options is the same (calling into question Wilkinson’s equivocation between ‘reducing’ and ‘worsening’ a chooser’s options). Nor can we assume that the choice set with the cheaper soft drink is better because it has options that the chooser prefers, because we cannot assume that the worst off prefer cheaper soft drinks, especially on the assumption that the worst off ‘have a problem with the options from which they can choose’. According to the definition Wilkinson gives, the change from a cheaper to a more expensive soft drink could count as ‘a different option that the chooser prefers’, making the choice set with the more expensive soft drink the better of the two. If the choice set with the cheaper soft drink is not better than the choice set with the more expensive soft drink, it is unclear how the change imposed by the preventive regulation makes the choice set worse.

Wilkinson takes the notion of preference even further in his argument, claiming that if equity is concerned with the welfare of the worst off, preventive regulations that change the consumption of the worst off mean ‘the worst off would consume what they prefer less, which would reduce their welfare’, which would be inequitable. The problem with this claim is that we have no reason to assume that the worst off are currently consuming what they prefer, particularly since the issue under discussion is the poverty of options. If the available options are bad ones, then we cannot draw any conclusions from the choices made about what people prefer. Moreover, this change reduces welfare in only one sense (consuming less of what is preferred); in another sense, welfare (in the form of health) is potentially enhanced by this change.

The second problem with Wilkinson’s argument is his contention that preventive regulations are inequitable even if it is also the case that they reduce obesity among the worst off and therefore improve health equity. Wilkinson notes that it is an open question whether it is just to pursue an improvement in health of the worst off that reduces a health inequity if that improvement comes at the cost of some non-health benefits of the worst off. In the case of preventive regulations, Wilkinson believes that such a trade-off is unjustified: ‘When the problem is a poverty of options, restricting them further would be perverse. It would still be perverse even if restricting them reduced inequalities of health’. As argued above, it is not necessarily the case that preventive regulations do restrict, reduce or worsen options, but if we concede that they do, it is still not clear that they are therefore inequitable, even if they do reduce health inequities. Wilkinson is concerned to avoid imposing public health measures that make the worst off worse off simply because they reduce a health inequity and rightly cautions against ‘fetishizing health as a good independent of its contribution to welfare and overvaluing it at the same time’. Those who argue that any improvement in health or gain in health equity should be pursued regardless of its impact on overall welfare would be guilty of both. On a more moderate view, however, there are some things that we might be willing to trade for an improvement in the health of the worst off, or a gain in health equity, and some that we are not. Consider a community in which the only viable employment option is coal mining. A job in this industry poses serious risks to health but is the only available option to earn a living. A preventive regulation that reduces food choice seems to be analogous, in Wilkinson’s argument, to a regulation that would close down the coal mine on health grounds. An argument for this regulation could claim that the health risks of coal mining fall disproportionately on the worst off and that the health inequity between those who work in the mine and those who do not can be reduced by removing the option of coal mining, thereby improving the health of those who would otherwise work in that industry. However, since closing down the mine would leave all of these people unemployed, reducing their already impoverished employment choices in this way would be inequitable, even if doing so reduced health inequities. This is an example in which an inequity is reduced by worsening the position of the worst off. However, while it would seem justified in the coal mining case to object to the regulation on the grounds that it worsens the position of the worst off, preventive regulations are not sufficiently similar to closing the coal mine to make the same objection. This is because not all choice sets are as significant for people’s lives as others. Choices about employment are arguably more significant for people than choices of beverage, so changes that significantly limit the former are likely to have a greater impact on overall welfare. In rejecting the argument that a ban on trans fats in food unjustifiably limits liberty, Lawrence Gostin notes that ‘it matters a great deal how serious the limit on freedom is’.1 Removing a harmful ingredient in food, or making soda more expensive, is not ethically on a par with truly liberty-limiting public health measures such as quarantine or compulsory treatment, or with policies that ‘fetishise’ health relative to non-health considerations. While it can be inequitable to prioritise gains in health equity over losses in non-health benefits (as in the coal mining case), determining whether a regulation or public health measure is inequitable for this reason is to some extent a question of proportionality. More expensive soda represents a loss in non-health benefits but, as a test case, is arguably not ethically on a par with the loss of an individual’s sole employment option. Nor is it only health and non-health benefits that can be weighed against each other. We might give priority to reducing a health inequity between population groups even if doing so means a smaller gain in population health overall. Questions about how to weigh overall health gains against reductions in health inequity are ones on which reasonable people will disagree.2 This is also the case for questions of how to weigh gains in health, or reductions in health inequity, against losses in non-health benefits. If a change in options around food consumption produces a sufficiently large gain in health equity, it may justify some restriction of options. Allowing the gain in health equity to ‘outweigh’ the loss of non-health benefits in some instances does not mean fetishising health, but rather appropriately tallying the scale of one versus the other. It is not obvious that the potentially significant health benefits of a soda tax, for example, are outweighed by the loss of the non-health benefits of cheap soda (whatever those might be).