The International Impact of the Day After Tomorrow

��

�� The Day After Tom � � �� orrow � �� The Day After �� Tomorrow ��

� ��

��

�� The International Impact � of The Day After Tomorrow

ith its November issue, Environment been conducted: a Japanese study and two has provided a platform for an inter- British studies.4 PIK and the European Climate W esting study on a rather new type of Forum (ECF) hosted a workshop in late Octo- climate impact: After climate change reached ber with the main authors from all five studies. Hollywood, Hollywood struck back and gave This meeting informs my comments. the world Roland Emmerich’s The Day After Before turning to study results, I would Tomorrow, the top-ranking film in the recently first like to comment on the methodology created film genre “Global Warming Films.”1 Leiserowitz and his colleagues used. As far Anthony Leiserowitz’s article, “Before and as one can tell from the article, the distinc- After The Day After Tomorrow: A U.S. Study tion between “watchers” and “nonwatchers” of Climate Change Risk Perception,”2 is an was based upon the answer to a question like important contribution to a special field that “Have you seen The Day After Tomorrow?” assesses this new type of impact: that of cli- in the second of the two nationwide surveys mate change communication via the media the author performed. Most of the conclusions In addition to the U.S. on the general public. The author has quoted about the film’s impact on the U.S. public is and German studies, some important literature on this issue, and based upon this distinction. Of course, we three other impact his study might help contribute a novel facet learn a lot if we compare both groups. But do to it. The article is also one of several recent we really learn about the impact of the film on studies on the film demonstrations that climate change science, the public? Do watchers display their often sig- have been conducted: largely dominated by the natural sciences, is nificant, distinctive answering patterns due to a Japanese study and an interdisciplinary endeavor that needs social the fact that they have seen the film (in which science support. case their own answering behavior before hav- two British studies. Along with colleagues at the Potsdam Insti- ing seen the film would have been significantly tute for Climate Impact Research (PIK), I different), or do their answers reflect that they have completed a parallel study on the public had more pro-climate or pro-environment impact of The Day After Tomorrow in Ger- attitudes before entering the cinema? Only a many.3 We surveyed about 1,300 people imme- comparison of cinema visitors before and after diately before and after viewing the film and having seen the film (a panel study) would about 150 people 4 weeks later in subsequent be able to uncover the true effect of the film. telephone interviews to control for persistency Leiswerowitz’s study compares watchers to effects. The questionnaire comprised about 20 nonwatchers, but this comparison does not questions, covering issues of climate system necessarily tell us what the film’s impact was and change, climate policy, film attributes, and on its audience. sociodemographic and lifestyle information. The German panel study demonstrates a In addition to the U.S. and German studies, rather strong self-recruitment of better- three other impact studies on the film have educated and more engaged visitors of the

VOLUME 47 NUMBER 3 ENVIRONMENT 41 film: Thirty-six percent of respondents said ing that climate change is unpredictable and that before watching The Day After Tomorrow, that scientists have not agreed that it exists. In climate change was an issue they were interest- Europe and, to a lesser extent, Japan, the pub- ed in, one that had drawn them to the cinema lic discourse is much more influenced by the to see the film. This significant self-selection assumption that climate change is predictable effect should prevent the scientist and the reader or that it already happens. from drawing too strong a conclusion from com- Model A (“Climate is stable within limits”) parisons between watchers and nonwatchers. is the dominant choice in all three countries Despite this general caveat, the study is very (except for the nonwatchers in the United useful and informative, especially in its demon- States, who prefer model B). However, the sec- stration of the clear links to climate and general ond peculiarity is the change given the “impact” policy issues and in its comparison of the media of the film. If we assume that Leiserowitz’s impact of The Day After Tomorrow and other article indeed measures the film’s impact, in the Comparing the films or events. This makes the U.S. study rather United States, model A seems to be supported Japanese, German, unique, as none of the other four studies has by viewing the film, as the share rises from 27 taken such a deep look “outside” the climate percent to 39 percent. Although model A is the and U.S. studies with change issue and into the field of policy. dominant model in Japan and in Germany— regard to conceptual From a non-U.S. point of view, one result of before and after viewers have seen the film— models of the climate Leiserowitz’s study is particularly interesting. this model loses credibility due to the film: In He uses the four “myths of nature” from cul- the German case it falls from 44 percent to 42.2 system, three tural theory—an influential theory developed percent, and in Japan from 47.3 percent to only peculiarities arise. by anthropologist Mary Douglas that attempts 35 percent. Cultural theory interprets this model to explain risk perception—to concretize the as part of the worldview of the “hierarchist,” conceptual model of the interviewed. The that is, people who like to know about mecha- Japanese and the German studies have fol- nisms so they can control them (such as scien- lowed this approach as well. Cultural theory tists and bureaucrats). So while U.S viewers assumes four basic “myths”: nature is stable seem to have been led to frame climate change within certain limits, beyond which critical in a “hierarchist” way, this model has lost some developments lead to ecological crisis (model credibility for German and Japanese viewers. A); nature is random, its reactions to human Leiserowitz indicates that model A was the interventions cannot be predicted (model B); model conveyed by the film itself. If we take a nature is ultra-stable, no human intervention look at cross-cultural studies, this is a contest- can destabilize it (model C); and nature shows able statement. a delicate balance that can easily be disturbed For Germans, model D (“Climate shows a by human action (model D). Leiserowitz has delicate balance”) is the film’s clear “winner”: added a fifth conceptual model—“Climate is It went up from 29.1 percent to 37.5 percent. slow to change. Global warming will gradually This still places model D only second to A, but lead to dangerous impacts”—which is not part the trend is remarkable. German viewers of The of the original cultural theory framework and Day After Tomorrow take home the message may make comparison difficult. Nevertheless, of a very vulnerable and complex climate sys- a limited form of comparison is possible—and tem we should possibly not disturb—a model quite revealing. attributed to the worldview of the “egalitarian” Comparing the Japanese, German, and U.S. according to cultural theory. The Day After studies with regard to conceptual models of Tomorrow has especially brought forward the the climate system, three peculiarities arise. role of the oceans in the world’s climate system First, the number of U.S. respondents who and the existence of nonlinear changes, aspects chose model B (“Climate is random”) is very that many viewers did not know of before. This high (34 percent nonwatchers, 29 percent “sensitizing” effect was much weaker in the watchers). In Germany this model has lower Japanese case, where model D gained ground acceptance, and it even dropped after watch- only moderately (from 32.6 to 34.5 percent). ing the film from 26.2 to 19.9 percent. This (Japanese viewers were much more convinced difference might be a consequence of the of model B (“climate is random”), which went public debate and especially the dominant up from 18.7 to 28.6 percent, than they were discourse of the U.S. government, suggest- before watching the film.) But while model D

42 ENVIRONMENT APRIL 2005 rates second in Japan and Germany after view- s the article “Before and After The Day ers have watched the film, it performed poorly After Tomorrow” was going to press, in the United States and—even stranger from A I was very pleased to learn that some- a non-U.S. view—is not affected by the film: what similar studies had been conducted in the About 7 percent of watchers and nonwatchers United Kingdom, Germany, and Japan. Thanks chose it. to the generous hospitality of Fritz Reusswig There is much more to comment and com- and the Potsdam Institute for Climate Impact pare about the studies mentioned, and the par- Research, the primary investigators of all these ticipants of the Potsdam workshop agreed to studies gathered for a workshop in October unite forces to create such a comparison. For 2004 to share our respective findings. This now it is worth noting that the impact studies meeting was quite stimulating and led to the of The Day After Tomorrow have entered a formation of an international research team to new, reflexive area of climate change research: conduct cross-cultural experimental research. the area of the impacts of impacts. Twentieth I thank Reusswig for his comments on the Century Fox Germany has established an initia- paper and would like to take this opportunity tive to facilitate emissions trading rights and to address his primary concern. We conducted reducing CO2 emissions of services, events, three national surveys of the American pub- and traffic (see http://www.climatepartner.de). lic—before, during, and several months after One might take it as image work, but it is also the movie played in theaters. The article an indication that The Day After Tomorrow reported results from the first two waves, in might not be the last of the global warming particular the second, which compared a ran- movies. Thus, it will be helpful for climate domly selected group of movie watchers and Are the significant scientists to continue researching media and nonwatchers from a national sample in June differences observed film representations of climate change and the 2004—several weeks after the movie debuted. public’s response to them. It is doubtful that The first two surveys were not based on a in the U.S. study the creators of the United Nations Framework within-subject (panel) design, so this study between movie Convention on Climate Change had Hollywood was unable to directly measure whether watch- watchers and on their minds when they drafted Article 6, ing the film changed an individual’s attitudes which asks for improved communication and toward climate change. Thus Reusswig raises nonwatchers really education on the issue of climate change. But a legitimate question: Are the significant dif- due to the impact the entertainment industry seems to have done ferences observed in the U.S. study between of the film, or did quite a lot for the public awareness of climate movie watchers and nonwatchers really due to change, and Anthony Leiserowitz gave us a the impact of the film, or did movie watchers movie watchers very useful look at this new domain of climate already have “more pro-climate or pro- already have “more impact research. environment attitudes before entering the cine- pro-climate or ma”? In other words, perhaps moviegoers went Fritz Reusswig to the film because they were already more pro-environment Potsdam Institute concerned about global warming. attitudes before for Climate Impact Research Three streams of convergent evidence sug- entering the cinema”? Germany gest this hypothesis is incorrect. First, our own and other previous national surveys have found 1. See http://www.boxofficemojo.com/movies/?page=main &id=dayaftertomorrow.htm. that climate change is not a highly salient con- 2. A. Leiserowitz, “Before and After The Day After Tomor- cern of the American public, yet by the time of row: A U.S. Study of Climate Risk Perception,” Environment, our second survey, 21 million American adults November 2004, 22–37. 3. F. Reusswig, J. Schwarzkopf, and P. Pohlenz, Double had seen the movie in the theater. Our respon- Impact. The Climate Blockbuster The Day After Tomorrow and dents were randomly selected to represent this its Impact on the German Cinema Public, Potsdam Institute for Climate Impact Research (PIK) Report No. 93 (Potsdam, group. On its face it seems unlikely that 21 Germany: PIK, 2004), http://www.pik-potsdam.de/publications/ million Americans went to the film because pik_reports. 4. M. Aoyagi-Usui,“The Day After Tomorrow: A Study on they were already highly concerned about glob- the Impact of A Global Warming Movie on the Japanese Public,” al warming. It is more likely that most people National Institute for Environmental Studies (NIES) Working Paper (unpublished), October 2004; T. Lowe et al., “Does Tomor- went to see the film because it was a summer- row Ever Come? Disaster Narrative and Public Perceptions of time, blockbuster disaster movie. Climate Change,” Draft Tyndall Working Paper (unpublished), October 2004; and A. Balmford et al., “Hollywood, Climate Nonetheless, we explicitly tested this hypoth- Change, and the Public,” Science, 17 September 2004, 1713. esis in our third and final survey, completed in

VOLUME 47 NUMBER 3 ENVIRONMENT 43 November 2004. In this survey (not reported in reported in the article, we found that even after our article because it had not been conducted controlling for these variables, there remained yet) we re-interviewed the same respondents significant differences between the attitudes of as in wave two, including movie watchers. We watchers and nonwatchers. asked them, “Why did you watch this movie?” Third, as reported in the Environment arti- Of all movie watchers, only 17 percent said cle, we directly asked movie watchers whether they went because they were “interested in the movie made them more worried about global warming.” By contrast, 83 percent global warming. Forty-nine percent of movie- of moviegoers went because they “liked the goers said the film made them somewhat (36 trailer” (29 percent), “like disaster movies” percent) or much more worried (13 percent), (21 percent), “like to see all big films” (21 42 percent said it did not change their level of percent), or “another reason” (12 percent). In worry, and finally, only 1 percent said it made contrast, Reusswig’s team found that among them less worried. These three streams of con- We have only German moviegoers, 36 percent said a prior vergent evidence all suggest that indeed, the scratched the surface interest in climate change led them to watch reported differences in perceived risk between in the effort to the film. As he writes, “The German panel watchers and nonwatchers were due to the study demonstrates a rather strong self- impact of the film. understand the recruitment of . . . more engaged visitors of the During the meeting in Potsdam, the principle role of popular film.” Again, by contrast, only 17 percent of investigators of all five studies identified a representations of risk. American moviegoers said they went because number of other intriguing cross-cultural dif- of a prior interest in global warming. Thus, the ferences in American, British, German, and results on which he bases his conclusion that Japanese responses to the movie, which we “there is a significant self-selection effect” are intend to investigate further with a multination- probably more indicative of very interesting al experimental study, using exactly the same cross-cultural differences between German and research design and instruments in these and American climate change risk perceptions. other cultural contexts. We have only scratched Second, as reported in the article, we deter- the surface, however, in the effort to under- mined that movie watchers were demographi- stand the role of popular representations of risk cally different from the general public—they (such as movies, books, television, fiction, and tended to be slightly younger, male, Hispanic, nonfiction) or of cross-national differences in and politically liberal. We therefore used mul- public risk perception and behavior. tiple regression to control for sociodemographic and political variables, including sex, age, Anthony A. Leiserowitz education, income, race, political party, and Decision Research political liberalism. In almost all cases and as Eugene, Oregon

44 ENVIRONMENT APRIL 2005