Beer Marketer's Insights - Cancer Warnings "May" Not Work as Intended, Causal Language Less Credible: Study

As public health advocates around the world rally behind efforts to mandate cancer warnings on alcohol packages, a new study emphasizes that the wording of such warnings would matter and that they may not be as effective at convincing folks not to drink as advocates hope. Though based on a relatively small sample of survey respondents, results surprised authors in a number of striking ways. Most crucially, "high-certainty causal language" &mdash; such as in the explicit warning that "drinking alcohol causes cancer" &mdash; "may neither increase cancer risk perceptions nor decrease alcohol consumption intentions, but may instead incite negative reactions to these messages," authors wrote. Ref 1 Small Sample; From "Causes" to "Contributes," Here a "May," There a "May" For this paper, researchers analyzed results of an online survey conducted back in February 2023, almost two years before January's release of the Surgeon General's Advisory on the topic. And it was only fielded to 799 American adults aged 18+. Each of those folks saw one randomly assigned version of a possible cancer warning before being asked a series of questions about: 1) "causal certainty" (how clear they think the link is); 2) their perception of the risk; 3) the credibility of the message; 4) "reactance" (whether they find the message manipulative, overblown or annoying), and 5) their "intention" to drink in the future.Respondents saw one of eight warnings, each one rated by anywhere from 78 to 111 individuals. On the most blunt, "high-certainty" side of the spectrum, some folks saw the message cited above: "drinking alcohol causes cancer." Other variants couched the warning to varying degrees, alerting respondents that alcohol "is linked," "contributes to the risk of" or "increases the risk of cancer." A different set of messages dampened the certainty of causality in the warning by inserting the word "may" into each of those statements (i.e. "drinking alcohol may contribute to the risk of cancer").Causal Language Garnered Negative Reactions, Not Confidence Contrary to what researchers expected, the most definitive statement did not do a better job of convincing respondents about either "causal certainty" or perceived risk than slightly softer messages using words like "linked" or "contributes." The addition of the word "may" resulted in lower certainty scores, but not in meaningful differences in how respondents perceived the risk of drinking. Also contrary to expectations, researchers found no meaningful impact from any of the variations in language "on future alcohol consumption intentions." In other words, respondents reported similar expectations for when they'd have their next drink regardless of which message they saw.The biggest differences showed up in measures of "reactance," or negative reactions to the messages, and whether folks thought the warnings were credible. And the most direct, causal statement was easily the most likely to elicit a negative response and least likely to be believed. The findings suggest that "causal messages about alcohol-related cancer risk may have paradoxical effects that could pose tradeoffs and undermine the goals of communicators," the authors summarized. Such messaging appears to have a "neutral effect" on raising awareness and "may incite negative reactions," suggesting "the need to strike a balance" and "utilize causal language that conveys at least some uncertainty."Why the Surprise? Finding the "Ethical" Option "There are a lot of possible explanations for this finding," one of the lead authors explained in the press release announcing the study's publication, suggesting "some underlying psychological or cognitive factors that we might not have initially accounted for." But the study itself cuts to the chase and presents the most obvious possible reason: such a simple, causal statement "may be perceived as inconsistent with the observable reality that the vast majority of people who consume alcohol are not diagnosed with an alcohol-related cancer." Maybe!Elsewhere, the authors repeatedly lean on WHO-backed research and rhetoric, the Surgeon General's Advisory and other studies and statements to unequivocally assert the "causal" relationship between alcohol and cancer. Yet, words like "may" or "could," they admit early on, "may be appropriate given variation in the health outcomes of individual persons." Recall, almost all of our understanding of alcohol's long-term health effects is based on observational research that cannot prove causality, though enough work has been done about cancer to make researchers and advocates comfortable with asserting causality. Still, the vast majority of alcohol-related cancer is linked to heavier consumption well beyond recommended limits, which this study's authors hint at only briefly by noting "the relative paucity of evidence on the cancer risk of low levels of alcohol consumption."Furthermore, constructing appropriate warning messages is not just about efficacy. There are ethics involved here, too. And only near the very end of the discussion of their results do authors acknowledge that "using terms such as 'increases the risk of' as opposed to 'causes,' or adding the modal verb 'may' might make audiences more receptive to the messages, and may also be ethically justifiable given the inherent limitations of all scientific evidence and the unpredictability of the future outcomes of specific individuals." What does that imply about the ethics or justifiability of telling consumers that the product they're holding "causes cancer" without caveat?

Cancer Warnings "May" Not Work as Intended, Causal Language Less Credible: Study

Publishing Info