Many public health ethicists will disagree with the reasoning outlined here. For example, some have argued that beneficence, non-malfeasance, and autonomy are less important in public health ethics than in other bioethical domains [22]. Others have argued that the ethics of public health is inseparable from the ethics surrounding contextual environmental factors linked to health, such as environmental protection [23], or to the social relationships that enable society to be sufficiently stable, resourced, and organized to pursue health.

These arguments have important additional limitations. First, they assume that any changes in normative standards induced by a policy would not sufficiently alter the benefit-to-harm calculus such that the policy would become favorable for a majority of persons. For example, if a salt restriction policy allowed a majority of salt-lovers to willingly forgo salt because they would be released from the normative pressure of eating salty foods, as hypothesized by Rose, then the individual ethical calculus would no longer be negative. Second, this discussion does not apply to behavior change that involves easily extinguishable preferences or habits that are easily adapted, such as seatbelt-wearing. On the other hand, it clearly applies to preferences that are rooted in evolutionary selection and reinforced by potent activation of the brain’s reward circuitry, such as an affinity for salty or fatty foods [20]. Third, implicit cost transfers need to be considered in the harm to benefit calculus (such as costs incurred to society by spending for salt-related health care costs). Fourth, this discussion assumes that health and wellbeing effects on different individuals do not have important communicable components, whether direct (a communicable pathogen) or indirect (an influencing behavior). Vaccines for COVID, influenza, and other highly communicable pathogens offer stark illustrations of how communicability may render population effects more ethically important than individual effects. Fifth, it is worth noting that an individual’s ethical calculus may be very sensitive to the stringency with which mandates pursue a health goal. For example, moderate reduction of salt in packaged foods, as has been accomplished in the U.K. and several other countries [24], may be ethically favorable because the gain in wellbeing from health benefits exceeds the mild loss in wellbeing. In contrast, stringent reduction of salt to less than 2 g per day may impair wellbeing sufficiently to outweigh health benefits for many people. Sixth, it is difficult to disambiguate empirically derived estimates of inequality-aversion (disliking the ethical consequences of unequal distributions) from risk-aversion (disliking the uncertainty regarding an individual’s position in a distribution as it becomes more unequal). Nonetheless, spillover from risk-aversion into inequality-aversion does not diminish the ethical relevance of that aversion. Finally, individual preferences may change, and therefore some scholars argue that any calculus that depends on the stability of preferences is inherently flawed. However, preferences rooted in evolution and reinforced by mesolimbic reward system exhibit great stability. While virtually all countries worldwide have attempted to reduce the salt intake of their populations, only 12 have documented subsequent reductions in daily salt intake. Among these 12 countries, the largest proportional reduction (China, 28%) was only sufficient to reduce salt intake to 12 g per day, and populations in all 12 countries continued to consume between 7 and 15 g per day, far above recommended levels [24]. In another example of the stability of reward-reinforced behaviors, the long-term success of non-procedural weight loss interventions is strikingly low (pooled estimate, 2.4 kg weight loss) [25].