HOME > ARTICLES

Simplification, as a scholarly vice, lies next to outright falsification, and when it comes to tackling the problem of underage drinking, nothing is more tempting than a simple fix. Banning, curbing, or otherwise regulating alcohol advertising has an international appeal, with each new study bringing, in the minds of those who would do the banning, the final piece of proof.

To be fair to the authors of “Early Adolescent Exposure to Alcohol Advertising and its Relationship to Underage Drinking” (Collins et al, Rand Corporation, Journal of Adolescent Health, April 2007), they took pains to note the problems of conducting research in this area, the mixed results of the most rigorous studies that have been conducted on the topic and the limitations of their own study.

Nevertheless, the New York Times reported that the study (along with another on tobacco) “suggested that fairly simple public policy changes might significantly reduce underage smoking and drinking.’

And this, at least in the case of drinking, was because

“After controlling for sex, race, parental education, school grades and more than a dozen other variables, the researchers found that exposure to advertising in sixth grade strongly predicted drinking in seventh grade. The most consistent predictor of drinking was ownership of a hat, poster, or T-shirt that advertised alcohol, they said, and the one-fifth of sixth graders who owned such items were almost twice as likely to take up drinking as those who did not.”

But looking at the actual data in the study should have given pause to such a simple conclusion.

The first set of data looked at partial associations between factors that could influence alcohol consumption. Where an odds ratio of 1.00 indicated that the factor had no influence on whether teens drank or intended to drink, or didn’t drink or intend to drink, the researchers found the following:

There was a positive association between exposure to TV beer ads in the sixth grade and drinking which ranged from 1.43 to 1.48 depending on the type of television ad. In other words, the sixth graders where 43 to 47% percent more likely to drink in grade seven based on seeing beer ads on tv.

But hold on – this was by no means the only factor to show such a positive association. The Times failed to mention that “Sports Activity” in the sixth grade delivered a higher odds ratio of drinking in the seventh – 1.60 or 60% more likely to drink. Low parental monitoring was greater than TV ads too (1.64), as was parental approval (1.69), and the approval of a friend almost doubled the chance of drinking (1.98).  Deviant behavior (2.00) and peer drinking (3.20) were even greater predictors of beer consumption.

The Times also neglected to mention that the Rand researchers tried to account for the effect of these factors on one another in a series of complex calculations. The result, at least according to the tabular data in the study, shows that the odds ratio significantly diminished for all advertising (1.08 to 1.13 for TV ads). The factors most strongly influencing seventh graders to drink were, once again, peer drinking (1.40) sports activity (1.52), friend approval (1.53) deviance (1.54), and above all, whether they had drunk in the sixth grade (2.32).

Nineteen percent of the students in the study did own beer promotional material, and the positive association between ownership here and drinking was strong, 3.54 in the partial analysis and 1.76 in the full multivariate analysis. So perhaps a ban on Budweiser caps might be in order. But even then, it may be overly simplistic to think that there’s a causal relationship between wearing a branded cap or t-shirt and the decision to drink. The latter decision may have come first, in which case the item reflects prior intent.

View the Technorati Link Cosmos for this entry