About two years ago, a young economist named Koleman Strumpf took on the music industry's conventional wisdom and created a little stir.
By way of justifying its lawsuits against music fans as young as 12, the Recording Industry Association of America, a powerful trade group that speaks for the major labels, asserted that rampant file sharing was responsible for the 20 percent decline in record sales between 1998 and 2003.
Strumpf, a professor of economics at UNC-Chapel Hill, saw an opportunity to find out if the claim was true: Did illegal music swapping through file-sharing networks really make a dent in the industry's bottom line?
While much of the existing research into online habits was based on surveying people about their behavior, Strumpf decided that asking people to give honest answers about an essentially illegal activity wasn't the most empirical way to find things out. He and co-author Felix Oberholzer-Gee of Harvard gathered data from a popular peer-to-peer network and proceeded to figure out which songs were downloaded and how many times, and compared those numbers with sales figures for the albums from which the songs were lifted.
They found that there was little to no economic damage done by file sharing. Most songs are never downloaded, and those that are correspond to the music industry's top sellers.
Reaction to the study was extremely mixed. The music industry vehemently tried to discredit these findings, but its rhetoric seems to have cooled off a bit. Last August, RIAA CEO Mitch Bainwol told a gathering of music retailers that burning--making illegal copies of CDs to share with or sell to others, offline--"is becoming a bigger problem than P2P."
Now Strumpf and Oberholzer-Gee are investigating whether movies traded illegally on file-sharing networks see a loss at the box office and the video store as a result. For a draft of the paper, see Strumpf's Web site.
I caught up with Strumpf at his office at UNC last week.
INDEPENDENT: How did you get started looking at movies?
KOLEMAN STRUMPF: About a year ago, we met this Dutch computer scientist who was collecting lots of information about BitTorrent, which is what kids like to use to download movies these days. He had this huge dataset--basically a snapshot taken every hour of all downloads in progress on BitTorrent sites from the fall of 2003 to the present. We're still in the process of getting this data together.
What have you found out so far?
We want to see if it's true that when there's an increase in downloads, these various measures of sales suffer. We've done it for a very small sample of films so far and we've found some evidence of a crowd-out for the box office, so there's a small negative effect. But so far it's not something we would call economically large. We've found maybe in the worst case that it will lower box office by a couple of percent. This is a very small subset of our data, so I don't read a lot into this yet. My hope is that we'll have more results this spring.
So, what movies are people downloading?
We started looking at Web pages like piratebay.org, which has a list of the top downloaded movies at any one time. This list is from about two weeks ago: Memoirs of a Geisha, King Kong, Syriana, and some things that had just come out on DVD like Wedding Crashers.
At first pass, it struck us that if you can get a movie the weekend the movie comes out, or even before, that does sort of sound like bad news for the movie business. Of course, this puts aside the question that you might really want to see King Kong on the big screen with nice speakers. But there's actually something more important than that. The question is, what do you get when you download one of these movies?
For instance, War of the Worlds was released in New York on June 23 and everywhere else on June 29. If you went to a file-sharing network the weekend of the 23rd, what would you find? A cam copy, which isn't very good.
In other words, someone went into the theater with a camcorder and just filmed the screen?
Right. Besides hearing the audience laughing and seeing the movement of the camera, the image is distorted. I looked at a still from a cam version of a scene in which Tom Cruise is staring at his hand while blood is misting in the air. It's no good--you can't see anything.
Then I compared it to a telecine version, which is a digital scan of the original physical print. It's much better. It basically takes a week before that version or the screener DVD copies are available. The reason that's relevant is that for a typical movie--and it's especially true of a big box-office action movie like this--about 40 percent of revenue from the box office comes the first week. So if there's a week where the competitor is a piece of junk, it's unlikely to be hurting revenue much.
I thought it was important, if I was going to look at the quality, to look at a pirated telecine copy, and I also went out and bought the real DVD. The quality of those two are virtually identical.
Another example is Lord of the Rings III from 2003. The movie was released in theaters on the 17th, and it got out on the file-sharing networks on the 21st. You can see a lot of people are trying to get a copy of this thing because it's really hot. But it took a week before anyone actually downloaded a full version of this movie.
What does Hollywood say about how downloading is affecting the movie business?
Last year was the first year things started to go south for Hollywood. Box office revenues were down 5 percent, which is the first time that's happened in a while. Even home DVD sales and rentals were flat--they dropped for the first time in 25 years.
The movie industry's a lot different from the music industry in that, to the best of my knowledge, they have never said this is due to file sharing. But I'm anticipating that time will come in the near future.
Hollywood's having tough times and file sharing's growing, but the good stuff doesn't leak quite so fast. And even when it does, there's a huge bottleneck.
Now, Hollywood gets only between 20 percent and a third of its revenues from box office--box office is basically an advertisement for DVD. So it's important to focus on other stuff, like what's happening with DVD sales and rentals.
What have you found with DVD sales so far?
The home video effect has been even smaller, which is the opposite of my intuition. My intuition would suggest that downloading would have more of an effect on DVD sales and rentals--one, because of the windowing system that Hollywood uses, this three-month lag before stuff is available on video, when it's easy to get on file-sharing networks. Also it's the nature of the product. You might think, I'm going to spend the $8 to go see this on the big screen, but if you're going to watch it at home, if you download it or buy it, it's pretty much the same thing.
On the other hand, it strikes me that potentially this could be a goldmine for these guys. The one piece of information we have so far is that these TV shows on iTunes are selling really, really well.
Hollywood has a bigger advantage over the music guys in that they have a really valuable inventory. Pretty much every movie ever made from 1901 to the present is sellable in some form. There's somebody out there who wants it--more so than music, especially popular music. So basically by putting this up on iTunes, there are no extra costs and it's all profit.
Read Godfrey Cheshire's review of Bubble, the first film to be released simultaneously in theaters, on DVD and on TV.