Intended for healthcare professionals


Effect of seeing tobacco use in films on trying smoking among adolescents: cross sectional study

BMJ 2001; 323 doi: (Published 15 December 2001) Cite this as: BMJ 2001;323:1394

Preliminary study does not justify sweeping policy

There are so many unaccounted variables in the Sargent
et al study that it is unwise to base sweeping legislative
proposals on its findings. Many studies have found
adolescent and adult smoking are part of constellations
of high-risk and thrill-seeking behaviors, and this
constellation may well include heavy popular media
consumption--especially the sorts of movies where
characters themselves engage in risky behaviors such as
smoking. Just as violent individuals tend to prefer more
violent media, so smokers may prefer movies where
their habit is depicted frequently and positively, while
those disinclined to smoke may reject movies where a
habit they reject is depicted. Evidence for my
counter-hypothesis is found in the Sargent study’s
finding that the odds ratio for smoking in the
highest-movie-exposed set of youths drops dramatically
from 8.8 (based on 4.9% smoking incidence in the
least-exposed set to 31.3% in the most exposed) to 2.7
when the selected sociopersonal variables in the study
are controlled. This suggests the sociopersonal variables
powerfully mediate the tendency of adolescents to
attend movies where heavy smoking is depicted.

The second objection is that the measure (ever tried
smoking) is overly broad. Most adolescents try smoking;
only a minority of those who try take up the habit.
Absent is any showing that those adolescents for whom
movie depictions of smoking (as opposed to other
variables) form a significant motivation to try a cigarette
are likely to continue the habit, or whether these
comprise the one-time experimenters.

The chief objection, however, is the proposal by Action
on Smoking and Health and Dr. Stanton Glantz, among
others, for legislated restrictions on youths attending
movies in which smoking is depicted. First, the present
study finding is correlational only. The fact that the
variables the authors did control for sharply reduced
their initial correlation is a warning signal that control
for additional variables might reduce it to insignificance.
For example, it is crucial to know whether the
correlation is in reverse (i.e., do teenage smokers simply
prefer movies depicting heavy smoking more than do
teenage nonsmokers?) or mediated by more important
variables (i.e., do poorer youth both smoke more and
attend the types of movies where smoking is prevalent?).
The study authors acknowledge that their preliminary
study should not been seen as proof that movies cause
teen smoking, especially since directional causation
cannot be determined.

Second, I would invite those proposing legislation to
consider what rating movies depicting smoking as
Restricted (under 17 requires guardian) means in effect.
Such restrictions have been found ineffective in
curtailing adolescent media patronage in any case; the
only effective measure would be to rate such movies as
NC-17 (no youths allowed at all) or to ban such movies
altogether. Even assuming a way could be found to
prohibit youths from seeing smokers in movies they
want to see, consider the thousands of movies, including
virtually all classic war, newspaper, crime, and comedy
films to which restrictions would apply. Are Dr. Glantz
and ASH suggesting that a 14- or 16 year-old must have
a guardian to see “Casablanca,” “Pink Panther,” or “The
Graduate”? A quick glance at the American Film
Institute’s 10 “Best American Movies of All Time”
indicates at least eight would be restricted for youth
under the proposed policy. Setting age limits for films
depicting smoking is simply too great a curtailment of
young people’s basic right to participate in their culture
even given the policy’s laudable goal of reducing
smoking. It is all the more cavalier in light of the
dubious evidence for it and the more moderate and
focused measures (such as requiring filmmakers to
disclose promotional payments and taxing them at 100%
of value). It should be emphasized that more than
two-thirds of youths exposed to even the heaviest
depictions of smoking do not try cigarettes even once,
yet they would be punished by one-size-fits-all

American anti-smoking policies of the last decade have
been woefully ineffective precisely because
anti-smoking lobbies too quickly veer into popular
panaceas affirming media-message determinism and
involving ever-accumulating age restrictions.
Unfortunately, ineffectual popular salvos typically grab
legislative attention, reducing support for politically
difficult, effective measures. The fact is that the large
majority of teenagers who smoke come from families
and communities where adults smoke, and the most
effective policies to curtail smoking involve raising
tobacco taxes and restricting adult smoking. Dr. Glantz
has been a rare pioneer and leader in rational
anti-smoking strategies, and his concern over the corrupt
relationship between movie makers and the tobacco
promotion industry is not misplaced. But I ask that he,
ASH, and other anti-smoking lobbies to reconsider their
proposal to punish adolescents en masse by restricting
them from movie attendance and instead to propose
policies that directly and exclusively penalize the film
and tobacco industries at fault.

Mike Males, Ph.D.

Sociology Department

University of California, Santa Cruz
Santa Cruz, CA 95061-7842


Competing interests: No competing interests

19 December 2001
Mike Males
Sociology Department
University of California, Santa Cruz