Graphic Warnings on Cigarettes: Nudge or Shove? A Neuro-Perspective

Although the topic of cigarette packaging regulation may not leap immediately to mind when one thinks “neuroethics,” this Bob Greene opinion piece over at CNN nonetheless touched off a stimulating discussion among some of us at the Core recently. The neuroethics connection, in fact, struck us as quite natural: our group has researched (and blogged about) the ethics of “nudging” frequently of late, and, as I worded it when I first emailed the article around, “certainly the images at issue here are a kind of behavioural nudge.” The question that we grappled with was whether the kind of nudge that the graphic warning labels provide is warranted in the case of cigarettes. And, indeed, that discussion called my original characterization into question. Do these labels truly constitute a nudge – a subtle biasing technique that makes a particular option more cognitively accessible than another while preserving the freedom to choose between them – or are they something more akin to a “shove?”

One of the least gruesome of the proposed images for cigarette packs.

As with any highly politicized issue, the question of whether cigarettes ought to be labeled with disturbing imagery is likely to be filleted into oblivion by pundits, bloggers, legal experts, economists, et cetera, et cetera. All I hope to do here, then, is sketch some ways in which the view from neuroethics – informed as it is by philosophy and the cognitive sciences – can shed some interesting and hopefully useful light on the question. Here, then, are three loosely-connected points along those lines.

The crux of the issue is decisional autonomy. Indeed, our conversation ultimately connected up with some broader philosophical work on autonomy and addiction, particularly as embodied in this 2006 paper by Neil Levy. The gist of Levy’s perspective is that we would be mistaken to conceive of an addicted person as fully autonomous, but equally mistaken to conceive of such a person as entirely lacking autonomy. Because of the way addiction works on a temporal scale (frequently resistible in particular instances, yet incredibly difficult to hold out against over long periods of time), we can best understand addiction as an impairment of autonomy. Nicotine addiction is no exception. This understanding alters the topography of the issue in a few ways. For one thing, it challenges the familiar libertarian perspective that people’s decisions about cigarette use (like anything else) will be Homo economicus-esque. In doing so, it also serves to make the grounding for a regulatory intervention of this sort narrower: one can resist slippery-slope arguments about government regulation of speech by pointing out that the marketing of such starkly autonomy-impairing products constitutes a special case.

However, understanding nicotine addiction in terms of autonomy impairment is not, by itself, a sufficiently nuanced approach. We need to dig a little deeper, because there are – of course – different kinds of addicts, on whom these labels will work in different ways. A particularly handy framework for thinking about this can be found in Harry Frankfurt’s classic 1971 paper, Freedom of the Will and the Concept of a Person. In the paper, Frankfurt identifies two kinds of desires people have: first-order desires, which are desires for certain things or outcomes (e.g. food or sex, but also desires to avoid bad things like trouble with the law or interpersonal strife), and second-order desires, which are a person’s desires about which one of their first-order desires will win out and manifest as behaviour.  Specifically for our purposes, Frankfurt parses addicted persons into three groups, based on their second-order desires (and second-order volitions, but that level of detail must be glossed over here): unwilling addicts keep using a drug despite wishing that their motivations to quit would win out against their addiction, willing addicts identify more with their desire for the drug than their desire to avoid the consequences of its use, and wanton addicts have conflicting first-order desires but are simply not invested in one or the other winning out. In considering how graphic warning labels might be received in the minds of cigarette consumers, we feel it would be helpful to begin by thinking in terms of Frankfurt’s categorizations. I will only raise, rather than attempt to answer here, the question of how each group might be differently affected.

Finally, an important consideration here lies in discerning what, exactly, makes a particular intervention a shove as opposed to a nudge. There are (at least) two strategies for drawing the conceptual line here. One is in terms of sheer efficacy: how swift a kick in the amygdala can a warning label deliver before it can reasonably be said to swamp all capacity for resistance? And how can we tell from population-level data about the efficacy of such labels what proportion of consumers are being nudged rather than shoved? (That much is probably impossible, at least not without some innovative research methods.) The other approach focuses on how nudges / shoves interact with a person’s first- and second-order desires. It seems at least plausible to argue that nudges remain nudges when they merely augment or undercut the strength of a given first-order desire. When second-order desires are impacted by the intervention, then it becomes something decidedly more coercive – or so the idea goes.

On balance, our group concluded that the graphic warnings constitute more of a nudge than a shove, albeit a stronger nudge than the bland textual warnings from the Surgeon General that accompany cigarettes in the US today. This is not a conclusion we will attempt to establish by argument here, as that task is beyond the scope of a blog post. One piece of preliminary evidence that the labels do not rise to the level of a shove is the example of our own beloved Canada, where graphic warnings are already in use: a sizable percentage of the Canadian population continue to smoke. At the same time, graphic warning labels seem to be effective; otherwise the tobacco companies would not be up in arms about their sale. In our discussions, we settled quite comfortably on the idea that this kind of nudge, engaging as it does with our emotions, is entirely appropriate as a means to help individuals overcome the autonomy-impairing effects of addiction and achieve greater clarity on the question of whether it is worthwhile to them to continue smoking in full appreciation of its often-grisly consequences.