SIVACRACY.NET

Chronicle.com:

Dispute Over the Economics of File Sharing Intensifies

By DAVID GLENN

In 2004, two economists released a working paper with a contrarian message: Sharing music files over computer networks isn't to blame for the recent collapse in CD sales.

It was a provocative attack on an assertion that has loudly been made by the music industry. A revised version of the paper ran last year in the Journal of Political Economy, which by many measures is one of the top five economics journals.

But now a rival scholar says that some of the paper's arguments are transparently shoddy and that the journal has failed to alert its readers to those flaws. This month he has taken his campaign to the Web, posting (among other things) e-mail messages and internal referee reports from the paper's authors and one of the journal's former editors.

The paper's authors—Felix Oberholzer-Gee, now an associate professor at Harvard Business School, and Koleman S. Strumpf, now a professor of economics at the University of Kansas—had persuaded an open-source file-sharing network known as OpenNap to give them a huge cache of raw user data from the last several months of 2002. They then counted songs that were illegally downloaded and the sales of CD's that contained those songs. They looked especially at how CD sales varied during weeks when songs were more easily available on the file-sharing network (because European students were on vacation and therefore spending more time online). The paper's bottom line: File sharing had no net effect on sales.

The paper seemed like a model piece of empirical social science for the Freakonomics era. Unusual data source analyzed with "instrumental variables"? Check. Counterintuitive conclusion? Check. Implications for hot-button policy debate? Check. The scholars filed an amicus brief in defense of file-sharing companies in the U.S. Supreme Court's Metro-Goldwyn-Mayer Studios Inc. v. Grokster case in 2005. When a revised version of their working paper appeared in the February 2007 issue of the Journal of Political Economy, it was the lead article.

But ever since the working paper first appeared, Mr. Oberholzer-Gee and Mr. Strumpf have been hounded by a critic: Stan J. Liebowitz, a professor of economics at the University of Texas at Dallas. Mr. Liebowitz, whose research has often been financially supported by the record industry, faults the scholars for not sharing their OpenNap data, which would allow colleagues to replicate their analyses. Insofar as he can check the paper, he says, he finds its arguments to be weak and misleading.

Until recently, the three scholars' quarrel had been on a low simmer—a testy e-mail exchange here, an awkward conference panel there. But in late May, the Journal of Political Economy rejected a critique that Mr. Liebowitz had submitted of Mr. Oberholzer-Gee and Mr. Strumpf's paper. (In economics parlance, critiques of previously published papers are known as comments.) The editor who sent the rejection notice was Steven D. Levitt, a professor of economics at the University of Chicago and the co-author of the best-selling Freakonomics: A Rogue Economist Explores the Hidden Side of Everything. Mr. Levitt's term as editor ended on June 1.

The rejection enraged Mr. Liebowitz for two reasons: First, he believes that he has demonstrated serious flaws in the file-sharing paper and that the journal owes it to its readers to discuss those flaws. Second, he discovered that the journal had used Mr. Strumpf himself as one of the two anonymous referees for his comment. (The second referee gave a positive report and advised that the comment be published.) Mr. Strumpf's role as referee was revealed when a reporter for the German newspaper Handelsblatt asked Mr. Strumpf to reply to Mr. Liebowitz's criticisms. He e-mailed to the reporter a response that was substantially identical to the negative referee report.

In recent weeks, Mr. Liebowitz has stepped up his campaign by posting new critiques of the file-sharing paper on the Social Science Research Network, along with the referee reports and other correspondence with Mr. Levitt and the two authors. ...