When warning is not informing: Alcohol labels will not be as straightforward as cigarette labels

News outlets have been discussing a call to require health warnings on alcoholic drinks comparable to those placed on cigarette packets. Amongst other recommendations, the All Party Parliamentary Group (APPG) on Alcohol Misuse has called on political parties to include a health warning on all alcohol labels, and to deliver a government-funded national public awareness campaign on alcohol-related health issues.

If this proposal is to be implemented, it is important to note that there is an important disanalogy with placing health warnings on cigarette packaging. Whilst cigarettes always damage health to some degree, a large body of evidence suggests that moderate drinking is not only non-harmful to health but may in fact promote it. A recent post by an addiction and public health specialist surveys the evidence which, in its totality, seems to support his claim that ‘even drinking more than is “perfectly” recommended, without displaying clinical symptoms of problem drinking or alcohol dependence […] is generally better for you than drinking nothing.’ Amongst the evidence he discusses, Peele cites a 2008 review from the Royal Society on Alcoholism which concluded that ‘a considerable body of epidemiology associates moderate alcohol consumption with significantly reduced risks of coronary heart disease and, albeit currently a less robust relationship, cerebrovascular (ischemic) stroke.’ The same report reviewed a range of biological evidence that suggested that moderate alcohol levels exert direct neuroprotective actions. It concluded, ‘In over half of nearly 45 reports since the early 1990s, significantly reduced risks of cognitive loss or dementia in moderate, nonbinge consumers of alcohol (wine, beer, liquor) have been observed’.

This is certainly not to say that frequently drinking large quantities of alcohol is good for you, and Peele is careful to note this. He also notes that whilst evidence suggests that drinking moderately each day confers health benefits (1-2 drinks if you’re a woman and 3-4 if you’re a man), drinking the equivalent of seven days’ moderate drinking all in one day is not going to do the same. Further, the data he discusses do not apply to women whose genetic makeup makes them prone to breast cancer – for these women, the health risks outweigh the health benefits.

The challenge this evidence raises for the lobbyists is whether the health warning labels that they recommend should also include information about the potential health benefits of moderate alcohol intake. If moderate drinking is actually better for you than abstinence, should this be noted alongside warnings about the effects of binge drinking? Given the APPG’s explicit intentions, the answer would seem to be yes. The say: ‘Information, education and awareness campaigns enable individual responsibility and are vital components of a comprehensive approach to reducing the harm from alcohol – consumers should have the right to make informed decisions about the products they purchase.’

Seemingly, promoting the consumer’s right to make informed decisions about alcohol consumption would require providing them with the information that moderate drinking may actually do them some good. The APPG go on to say that ‘In order to inform consumers about balanced risk, every alcohol label should include an evidence-based health warning as well as describing the product’s nutritional, calorific and alcohol content.’ Accordingly, it seems that the intention underlying the initiative is to empower consumers to make their own autonomous choices through presenting evidence. If this is the case, then by the APPG’s own lights, it would be neglectful for labellers to present only the pessimistic data associated with excessive alcohol consumption, omitting the more optimistic data associated with moderate consumption. Autonomous decision-making is best served by making agents aware of material information; and such information includes information pertaining to potential health benefits of moderate consumption as well as information pertaining to health risks.

Crucially, including information about the potential health benefits of moderate consumption alongside warnings about the potential harm that alcohol can cause would not involve conveying contradictory messages. The evidence that moderate drinking is net beneficial does not weigh against the evidence that binge drinking is net harmful. Rather, it demonstrates that the relationship between alcohol consumption and health is not linear.

It could be argued that by only including warnings about binge drinking no harm is done: whilst people may be steered away from dangerous drinking, warning labels will not stop people drinking completely. Whether this is the effect that warning labels would have is an empirical question. However, even if the effect was to reduce binge drinking I do not think it would be the case that no harm would therefore have been done. By excluding information about benefits, consumers would in fact be less empowered to make informed decisions about their alcohol consumption. To make a fully informed choice, consumers would have to be given all the evidence about the effects on their health exerted by all levels of drinking. They would then be in a position to assess what health risks they might be prepared to take in exchange for pleasurable effects and, crucially, whether or not they must restrict themselves entirely if they decide they are willing to accept only minimal or no risks to their health.

The labelling of alcoholic drinks is more ethically complicated than the labelling of cigarettes. Since the impact of alcohol on health is non-linear, the introduction of some (negative) health information seems to require the introduction of all (negative and positive) health information if the goal really is to promote the consumer’s right to make informed decisions about the products s/he purchases.

  • Facebook
  • Twitter
  • Reddit

2 Responses to When warning is not informing: Alcohol labels will not be as straightforward as cigarette labels

  • Phil says:

    While it doesn’t undermine the main thrust of the argument (which hinges on alcohol effects being nonlinear), the fact is that the health benefits of moderate drinking are strongly debated within the research community. Non drinkers tend to be a peculiar, heterogenous population, and the wide range of benefits attributed to alcohol consumption do not appear to have any connection to each other or any metabolic rationality, which would indicate the finding is, in fact, an artefact.

  • Anders Sandberg says:

    When I was in the US, I was told the reason there was no nutrient information on beer was that it was actually banned – it would confuse people into thinking it was healthy. The actual reason seems to be that the FDA and the BATF have a complex regulatory border on what is regulated as a food or alcohol, and bureaucratic messiness has slowed down the adoption of nutrient labels – with the industry torpedoing some of the proposals by arguing that if labels were added it would indeed fool some people into thinking it was good for them. After all, if something has vitamins, it must be healthy, right?

    Most likely buyers of something will tend to have their views reinforced by any labeling. The person thinking alcohol isn’t that dangerous will see the positive message about moderate drinking while the person worried about risks will notice the anti-binge warning and the stuff about alcohol and pregnancy. The problem might simply be that biased reading reduces the usefulness of honest information.

    There is also deeper nonlinearities: the cardiovascular protective effect of alcohol is nice, but it hurts nearly all other organ systems. But since cardiovascular risk is so dominant in our current health situation, the mortality rise in pancreatic damage is overshadowed. How do we explain that easily?


Subscribe Via Email