By Nathaniel Counts
In On Liberty, John Stuart Mill asserted that “[t]he only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others.” This has since become known as the harm principle and is foundational for much of American political discourse, especially for libertarianism and civil rights. At the time of Mills’ writing, On Liberty having been published in 1859, this logic could protect a lot of conduct that involved consenting adults or did not appear to directly impact others. If the harm principle was controlling, we could never have a soda ban. Today however, with our advances in social science, clear lines of harm and no harm have become fuzzy.
Few people buy cigarettes in a vacuum. Someone offers you cigarettes or you see other people smoking and then you buy cigarettes. This phenomenon, in which we make decisions based on the decisions of others in relationship to our esteem for them (if we see someone we respect smoking, we will be more likely to smoke; if we see someone we do not respect smoking, we will be less likely to smoke) is called social signaling (The Origin, Development, and Regulation of Norms is a great article on a related topic, the generation of norms, which explains this phenomenon quite well).
Social signaling is not inherently bad – we can receive signals to donate to charity, to be a better parent, to come to the aid of strangers – social signaling is neutral and dependent on the substance of the signals themselves. Some of these signals will make us worse off than if we were allowed to make an independent decision. It is hard to imagine many people who, having never encountered a smoker, would find cigarettes on the shelf in a store, figure out what they are for, and then buy them. Such people very well may exist, but it would certainly be a lower proportion of the population than currently smokes today. Because of this, many people may experience worse health outcomes caused, in part, by the actions of others.
This is in part why the non-smokers rights movement was so effective in reducing new smoker initiation (this study is a good example). The non-smokers rights movement made it illegal to smoke in many public spaces, so the act of smoking was less often a signal to others. Note that this does not militate for allowing smoking in private spaces – family members will still observe the smoking, these laws only limit the scope of the signaling. For some health concerns like obesity, prohibiting consumption of unhealthy foods when anyone else is able to observe would be insufficient as the obesity itself may be a signal to others.
Thus, social signaling presents a problem for the harm principle – actions we thought only harmed ourselves actually do harm others and no decision is without externalities. If we accept this, then the harm principle cannot be a basis for denying state regulation of certain activities that pose risks to public health. The soda ban cannot be proscribed on these grounds. To state this proposition in its starkest terms, based on the harm principle alone, the government could make it illegal for you to drink a soda at the movies because, in doing so, you are hurting the other patrons.