Sargent et al commence their criticism of our paper (1) by noting with a hint of tedium that I have now written critical articles about this issue three times in as many years(1-3). By comparison, Web of Science returns 68 papers on “Sargent J* AND movies”, Stoolmiller 13 and Wills 9. All three of our critics have thus staked out major career paths researching exposure to (mostly) tobacco in movies, but also to alcohol and violence. They have repeatedly endorsed adult-rating as the major policy response that should be adopted. To describe criticism as “demeaning” suggests a little hubris may have entered the debate. Skepticism is at the heart of scholarly engagement.

Sargent et al note that in my earlier papers, I agreed that it was undeniable that smoking scenes in movies could function in the same causal way as tobacco advertising in influencing smoking and that it was unarguable that smoking in movies contributed to the cultural iconography and appeal of smoking. I did indeed write this, but as they know, this was in the preamble to argument that controlling commercial speech through advertising(4) ought not be regarded as equivalent to controlling “cultural speech” in cinema, theatre and literature. They call my views here “philosophical” and resent the comparison with despotic nations like North Korea where heavy regulation of cultural speech is a routine part of the ideological state apparatus.

We readily agree that expanding the criteria for film classification into arguments that various exposures might promote undesirable behaviours requires “philosophical” values to be made explicit. This is because there is a long and often disturbing queue of single-issue advocates who would wish to see greater state intervention in cultural expression. Such a door should be opened with great caution.

Since writing those papers(2, 3), I have reviewed work on this subject for journals and began to be troubled by the blind spot that the Farrelly et al paper(5) exposed: the near complete entanglement of smoking with other “forbidden fruit” dimensions of movies. The Keynesian principle that “when the facts change, I change my mind” did not inhibit me from then going on to express concerns about the research base that underpins the policy proposal. To be clear, we do believe that smoking scenes in movies can influence smoking, but we believe that the precision and magnitude of estimates now being routinely recycled in policy advocacy is highly questionable.

Moreover, while we agree that many smoking scenes and characters are likely to contribute to a net positive view of smoking, some explanation is needed for the role of entertainment media in also contributing to the major cultural denormalisation of smoking(6) that has been occurring in recent decades. Part of this may be the reduction of smoking in movies(7), but there are also many overtly and subtly negative treatments of smoking in movies and television that probably contribute to the decay of smoking’s former status. This compilation from the globally massively popular Friends series is illustrative.http://www.y... If Sargent et al had their way, no adolescent should ever be exposed to such programs.

Sargent et al argue for the virtues of a “parsimonious” theory of smoking in movies and its relationship to subsequent smoking: social modeling. The idea of parsimony being virtuous in science centres on the central tenant of Occam’s razor that the simplest explanation is most likely the best. However, simple theories are not inviolable. Media theorists examining the polysemic nature of media content and the often radically different ways in which different audiences deconstruct meaning of the same programs have for decades rejected as simplistic the “hypodermic” model of media effects (8). They instead favour social constructivist accounts of how audiences actively mediate the meaning of complex media. This is not to say that media has no “effects”, but that the relationship is far from linear.

By contrast, Sargent et al inhabit a deeply reductionist world view where complex human behavior can all be explained by models derived from quantitative data. To them, nothing is beyond the reach of such reductionism and presumably no behavior cannot eventually be reduced to an equation. They concede that movies are “complex stimuli” but are untroubled by the progression of reasoning that quickly takes us to conclusions that youth smoking would be 44% lower if children never saw smoking in movies(9). In the history of tobacco control, we are unaware of any policy or intervention which has ever been remotely declared to have such power. Such extrapolations seem instinctively silly and deserve forensic examination in order to understand how they get off the leash of commonsense.

Finally, there is actually a fifth major problem with the adult-rating proposal. Studies in this field include R-rating movies in their exposure assessments. For example, in Sargent et al’s 2007 paper, 40% of the films were R-rated(10). Sargent et al have also shown that between 68-81% of US adolescents are allowed to watch R-rated movies(11-13). Many more watch without parental approval via downloads and file-sharing. Furthermore, 88.2% of youth-rated movies in the US now have no tobacco scenes (http://www.cdc.gov/mmwr/p...).
Putting these together, estimates of the effect of movie smoking exposure already include the impact of the R-rated solution being proposed to reduce that exposure. If youth who allegedly start smoking because of exposure to smoking in movies are already watching lots of R-rated movies, how would an R-rating reduce such exposure? Moving nearly all movies with smoking to R-rating would put the onus on parents to regulate their children’s viewing. Few would disagree with that. But why would parents regulate their children more because of concern about smoking than they do now with because of concerns about exposure to strong violence and explicit sex in R-rated movies?


1. Chapman S, Farrelly M. Four arguments against the adult-rating of movies with smoking scenes. PLoS Med. 2011;8(8):e1001078. doi:10.1371/journal.pmed.
2. Chapman S. What should be done about smoking in movies? Tob Control. 2008 Dec;17(6):363-7.
3. Chapman S. With youth smoking at historical lows, how influential is movie smoking on uptake? Addiction. 2009 May;104(5):824-5; discussion 5-7.
4. Gostin LO. Corporate speech and the Constitution: the deregulation of tobacco advertising. Am J Public Health. 2002 Mar;92(3):352-5.
5. Farrelly MC, Kamyab K, Nonnemaker J, E C. Movie smoking and youth initiation: parsing smoking imagery and other adult content. Social Science Research Network. . 2011; Available from: http://papers.ssrn.com/so....
6. Chapman S, Freeman B. Markers of the denormalisation of smoking and the tobacco industry. Tob Control. 2008 Feb;17(1):25-31.
7. Jamieson PE, Romer D. Trends in US movie tobacco portrayal since 1950: a historical analysis. Tob Control. 2010 Jun;19(3):179-84.
8. Livingstone S. On the continuing problem of media effects research. In: Eds Curran J, Gurevitch M, editors. Mass media and society. London: Edward Arnold; 1996. p. 305-24.
9. Millett C, Glantz SA. Assigning an '18' rating to movies with tobacco imagery is essential to reduce youth smoking. Thorax. 2010 May;65(5):377-8.
10. Sargent JD, Tanski SE, Gibson J. Exposure to movie smoking among US adolescents aged 10 to 14 years: a population estimate. Pediatrics. 2007 May;119(5):e1167-76.
11. Sargent JD, Beach ML, Dalton MA, Ernstoff LT, Gibson JJ, Tickle JJ, et al. Effect of parental R-rated movie restriction on adolescent smoking initiation: a prospective study. Pediatrics. 2004 Jul;114(1):149-56.
12. Farrelly M, Kamyab K, Nonnemaker J, E. C. Movie smoking and youth initiation: parsing smoking imagery and other adult content. Social Science Research Network. Social Science Research Network; 2011 [31 Mar 2011]; Available from: http://papers.ssrn.com/so....
13. de Leeuw RN, Sargent JD, Stoolmiller M, Scholte RH, Engels RC, Tanski SE. Association of smoking onset with R-rated movie restrictions and adolescent sensation seeking. Pediatrics. 2011 Jan;127(1):e96-e105.