BoingBoing announced with a flourish this week the arrival of “a study showing that feature films’ US box office returns are not correlated to BitTorrent sharing.” That is to say, piracy wasn’t hurting ticket sales. This finding gibed nicely with the view of the author, Cory Doctorow, that efforts by studios to clamp down on filesharing are practically pointless and a violation of online liberty, to boot.
“Contrary to what’s often claimed by the movie industry,” crowed TorrentFreak —which admittedly doesn’t sound like the most unbiased source — “the researchers conclude that there is no evidence that BitTorrent piracy hurts US box office returns.”
Trouble is, as Joel Waldfogel, one of the authors of the paper in question, later wrote with some exasperation at Digitopoly, those summaries didn’t reflect the paper’s central finding: “We think our marquee result is the opposite: we do find evidence that piracy depresses international sales,” Waldfogel wrote. In fact, the finding about the U.S. box office was presented deeper in the paper and hedged by statements about the limitations of the methodology.
The meaty part of the study used the lag between domestic and foreign releases of Hollywood films to explore the impact on ticket sales of the arrival of BitTorrent, a platform that made sharing movies much faster, and therefore feasible. Hollywood films typically appear in foreign theaters several weeks after they do so in the U.S, creating a tempting window for illicit downloaders. (The reasons include a scarcity of theaters abroad, the price of making many copies of films, and the desire to have stars do publicity tours.) If filmgoers started acquiring films using BitTorrent instead of waiting for them, you’d expect the effect of a delayed opening to become greater.
BitTorrent arrived in 2003 but was far more deeply entrenched in 2004. The study therefore compared the box office returns of Hollywood films (top 10 each week), in 17 foreign countries and the U.S., in 2003 and later years. They looked at the top 10 films each week, and the study controlled for the popularity of individual films.
In fact, the data showed that the penalty of a delayed opening increased markedly after 2003. In 2003, each additional week between a U.S. and foreign release caused a 2 percentage-point drop in sales; by 2005, each week of lag caused a 3.1 percentage-point drop.
Waldfogel, an economics professor at the University of Minnesota, and his co-author, Brett Danaher, a professor at Wellesley, then explored whether the post-BitTorrent penalty was greater for movies popular with young males, who are the most prodigious downloaders.
Indeed, release lags, in the BitTorrent era, hurt sci-fi and action-movie box-office receipts significantly more than they did romantic comedies or artier films.
Looking at 2005, the authors estimate that movie studios lost an estimated $240 million in foreign ticket sales to illegal downloads — a 7% hit. $3.52 billion in returns shrank to $3.28.
Of course, one implication of the study is that shortening release delays would cut down on piracy. And that’s something that studios are already doing, for a variety of reasons, piracy-fighting surely among them. (Digital cinema is also easier to distribute.) What’s more, how much you choose to get worked up over a 7% loss to movie studios is a personal question. I, for example, have some sympathy for the French Batman fanatic who simply can’t bear reading about a new blockbuster that he can’t see legally.
And the U.S. part of the study? What the authors did there was compare the curve of box-office receipts from the early days of BitTorrent to later years. In the later period, they found no evidence of a steeper falloff in ticket sales after a film was released. “In short, we do not see much evidence that piracy displaces US box office sales in our data,” they wrote, “although this result should be taken cautiously as the ‘experiment’ for examining US piracy is less clean than for international piracy.” One obvious shortcoming: if most piracy took place before a movie’s release, hurting mainly a film’s opening weekend, that wouldn’t be picked up by this approach.
“[W]e have been surprised at blog posts saying that researchers find that piracy doesn’t depress movie sales,” Waldfogel wrote, at Digitopoly. “… So much for the wisdom of crowds.” But given how contentious the subject of filesharing is, it’s actually not too surprising that commentators would read a new study on the topic through an ideological lens.