Dispute Over File Sharing’s Harm to Music Sales Plays Again

Last week in Vienna, where Beethoven, Haydn, and Mahler once walked, scholars came together to argue about Radiohead, file sharing, and the economics of music.

At a conference known as Vienna Music Business Research Days, two American economists renewed their long-running dispute about whether or not peer-to-peer file sharing is responsible for the worldwide decline in CD sales.

The quarrel centers on a widely cited paper by Felix Oberholzer-Gee, a professor at the Harvard Business School, and Koleman Strumpf, who now teaches at the University of Kansas. Mr. Oberholzer-Gee and Mr. Strumpf argued that file sharing does not have a net negative effect on the recorded-music industry. They arrived at that conclusion by examining the relationships among American record sales, American file sharing, and school holidays in Germany during the last quarter of 2002. (If file sharing injures CD sales, the paper’s reasoning goes, then American CD sales should suffer especially during weeks when young Germans are home from school because Germany is a major source of files traded on peer-to-peer networks.)

From the time drafts of the paper first circulated in 2005, it has been attacked by Stan J. Liebowitz, a professor of economics at the University of Texas at Dallas. (Some of Mr. Liebowitz’s work has been financially supported by a music-industry trade group.) In Vienna last week, Mr. Liebowitz reopened the argument with a new line of criticism.

In his Vienna paper, Mr. Liebowitz argued that Mr. Oberholzer-Gee and Mr. Strumpf’s central statistical model seems badly wrong in a way that he had not previously noticed. At one point their paper reports that a one-standard-deviation increase in the number of German students on vacation raised American file-sharing rates by half of their mean level.

But common sense suggests that the effect cannot possibly be so strong, Mr. Liebowitz’s Vienna paper says. If that were true, then “a power failure in a portion of Germany, or any event that caused German students to turn off their computers, would completely eliminate American file sharing. How realistic is that?” (Mr. Liebowitz and other skeptics have not been able to directly replicate Mr. Oberholzer-Gee and Mr. Strumpf’s central analysis, because it is based on a confidential and proprietary data set from a file-sharing company.)

Mr. Olberholzer-Gee did not reply to a request for comment from The Chronicle. But he and Mr. Liebowitz had some heated words during the conference panel, according to an account by John P. Palmer, of the University of Western Ontario.

Later in the conference, Mr. Oberholzer-Gee presented a recent paper by himself and Mr. Strumpf. The new paper makes arguments that are less controversial than the previous one’s were. First, they point out that file sharing has generated huge new complementary industries for MP3 players and other products. Second, they note that the amount of new music does not seem to have declined–so there is no strong evidence that file sharing has destroyed muscians’ incentives to create.

But even though they seem to be de-emphasizing the idea that file sharing does not hurt sales, Mr. Liebowitz did not give his colleagues any peace. In his Vienna presentation, he said that their new paper seemed to mischaracterize at least two recent studies of file sharing by other scholars.

On Page 16 of their Vienna paper, for example, Mr. Oberholzer-Gee and Mr. Strumpf cite a 2007 paper in Management Science as one of several that found that “file sharing does not hurt sales at all.” But that phrase is badly misleading, Mr. Liebowitz said, because the paper actually concluded that file sharing reduces the amount of time that many albums spend on the music charts.

“It is not correct to say that our work shows file sharing is unrelated to changes in sales,” said the Management Science paper’s lead author, Sudip Bhattacharjee, in an e-mail message to The Chronicle.

The paper did not look directly at sales, only at chart longevity, also known as chart survival. And “we did report a decrease in survival over all,” said Mr. Bhattacharjee, who is an associate professor of operations and information management at the University of Connecticut.