How do we calculate each studio’s share of kids delivered to the tobacco industry?

We carefully analyze several sets of data to produce reliable estimates. First we establish how many tobacco impressions the major studios are responsible for, combine the results with data on the number of teens who see films of various ratings. Second, we apply those numbers to the results of research on movie smoking’s effect on teens. Here’s how, step-by-step. For a brief discussion of more technical statistical issues, see the full report on the number of smoking images in each studio's movies and the scientific papers on the effects of smoking in the movies on adolescents starting to smoke. Questions? Contact us.

STEP 1. Pick a time period for which data is readily available. In this case, we researched the six years 1999-2006, because reliable data on movies’ tobacco content was almost perfectly complete for those years. (Content analysis for the top box-office films is available for most of the 1990s.)

STEP 2: Create a database of films. We used IdbPro, a database that can be power-searched on a subscription basis, to produce a list of all live-action, English-language, U.S.-produced feature-length films grossing over $500,000 released to U.S. theaters between Dec. 25, 1998 and Dec. 24, 2006. Each record included the film’s title, year of release, MPAA rating, production and distribution credits, director, producers, and domestic box office (theater ticket sales figures) while in first-run. The film info was filtered by several different methods to make sure it was complete; data conflicts and lacunae were resolved by consulting other industry resources.

STEP 3: Assign each film to a major studio (and parent corporation, if any). Films from subsidiaries, from production companies in which a major studio had an ownership interest, and from production companies with multi-picture deals with a major studio were credited to the major studio. So were films independently produced but distributed in the U.S. by the distribution arm of a major studio. Where a film was produced by one major studio and distributed by another studio, or when a film was co-produced by two major studios, the movie and its smoking content were co-credited. Analyses were adjusted so as not to double-count co-credited films.

STEP 4: Calculate the number of viewings each film received in theaters. The National Association of Theatre Owners (NATO) publishes average ticket prices for each year we surveyed. We divided the film’s box office total by the average ticket price to find the number of tickets sold — the number of times the film was seen by a paying customer.

STEP 5: Estimate the number of tobacco incidents in each film. Dedicated researchers have trained teams of monitors to keep accurate track of all tobacco uses, references, and visuals (such as billboards) in top box-office films. But we needed data for all movies, not just the most popular. First, we established which films were tobacco-free. Then we surveyed the tobacco content of the rest. Fortunately, an online parental screening service reviews almost every movie that comes out for smoking and other content. We compared its five-point scores for tobacco content to the actual incidents recorded by the researchers in a large enough sample of films to establish that the five-point scale could be translated into a per-film tobacco count. While not reportable for individual films, results are reliable when combined by studio, by rating or by year — our levels of analysis.

STEP 6: Calculate the number of tobacco impressions. To estimate the number of tobacco impressions films delivered to the audience, simply multiply the number of tobacco incidents in a film (Step 5) by the number of viewings (Step 4). A movie with a small number of smoking incidents, but a huge audience, might deliver as many tobacco impressions as a film with a great deal of smoking but a smaller audience.

SO…WHAT DO WE KNOW SO FAR? We know how many films of different ratings, credited to major studios and to a few smaller independents, showed tobacco use. We can reliably estimate how many tobacco incidents each studio produced and distributed, by rating and by year. We can reliably estimate how many tobacco impressions each studio delivered to moviegoers, by rating and by year. We can also estimate the total number of tobacco impressions delivered in a given year and over the five years. And we can estimate what percentage of these tobacco impressions each studio delivered. But we’re most interested in how many tobacco impressions were delivered to kids. That means…

STEP 7: Find out the age make-up of audiences for films of different ratings. Studio marketing data on the audiences for their releases was not accessible. We did, however, locate hard data on the percentages of the audiences for G, PG, PG-13, and R movies who are ages 6-11, 12-17, 18-24, and so on. This data also showed the audience size for films of different ratings, and the absolute number of people in each age category who see films of different ratings.

STEP 8: Estimate the tobacco impressions delivered to kids 6-11 and teens 12-17. Equipped with the age-breakout of movies rated G, PG, PG-13, and R, we can determine how many tobacco impressions from PG-13 movies were delivered to kids or teens, for example. Take the total number of tobacco impressions delivered by PG-13 movies. Multiply that number by the fraction of PG-13 audience members aged 6-11. Follow the same procedure for films of other ratings for kids 6-11 and teens 12-17.

STEP 9: Estimate the tobacco impressions delivered to kids by each studio. We know each studio’s fraction of tobacco impressions delivered by films of each rating. (For example, 26% of tobacco impressions in G/PG films 1999-2004 came from Disney films.) We can also calculate the fraction of teens’ tobacco impressions that come from G/PG movies: divide the G/PG tobacco impressions delivered to teens by the total tobacco impressions delivered to them by all ratings. To arrive at studio share, multiply the two fractions together and sum the results for films of all ratings. Example:

STEP 10. Establish the effect on-screen smoking has on young audiences. Based on landmark research published in the summer of 2003 — supported by a decade of other research findings — we know that real-life exposure to on screen smoking recruits half of all new teen smokers. (90% of all U.S. smokers start younger than 20). Equally important, we also know the effect is very straightforward — a teen who has seen twice as much smoking on-screen is twice as likely to start smoking as one who has seen half as much. From on-going surveys, we know how many teens start smoking each year. Half that number, the half recruited by smoking in movies, equals 390,000.

STEP 11. Estimate the number of teens delivered to the tobacco industry by each major studio. Because we know that exposure is dose-related — twice as much produces twice the effect — we know that each studio’s contribution to teen’s exposure boosts the number of teens who start smoking, in direct proportion. If a studio reduced the number of tobacco impressions it delivered, on the other hand, the number of teens starting to smoke would go down. Multiply the number of kids starting to smoke because of their exposure to smoking on screen (390,000 annually) by the share of tobacco impressions delivered to teens 12-17 by each studio to estimate how many kids they deliver to the tobacco companies. In The Disney Company's case (18% of 390,000), the stunning answer is that 70,000 teens start to smoke every year because of smoking scenes in Disney’s movies.

STEP 12: Estimate each studio’s contribution to tobacco companies’ sales revenue and profit. How much is a studio’s tobacco imagery worth to the tobaco industry? We have already estimated the studio’s share of the 390,000 annual new young smokers recruited by their on-screen. The same fraction can be applied to that cohort’s lifetime tobacco sales revenue ($4.1 billion) captured by the tobacco manufacturers and the piece of this revenue counted as profit. Based on its share of new young smokers recruited 1999-2004, for example, Disney’s tobacco imagery has been worth an average of $738 million annually in sales revenue to the tobacco industry over that period: $4.1 billion x 0.18 = $738 million (npv). The notation “(npv)” given after the dollar-amount of this multi-year money flow means “net present value;” values over a lifetime are discounted to state their worth here and now.

Because the number and ratings mix of a studio's releases, the box  office earned by its releases, and their tobacco content all change  year-to-year, a studio's share of new smokers recruited will also  change. The computations we describe produce, for each studio,  average annual figures over a defined period of years. The data used  in the illustrations on this page come from 1999-2004. As data are  updated, each studio's average may shift up or down; so the number of  new smokers attributable to each studio changes, as will the studio's  contribution to tobacco sales revenue and profit. For example, based on the average amont of smoking in Disney films from 1999-2004, we estimated that the average value of new smokers to the tobacco industry was $738 million a year; based on the average amount of smoking in Disney films from 1999-2005, this estimate becomes $690 million.

As we’ve noted, the CEOs of the major media companies that own most of Hollywood’s major studios could save 50,000 lives a year just by picking up the phone and telling their studio chiefs to keep smoking out of their kid-rated movies. That’s what the R-rating will do, simply, unintrusively, effectively.