Finally, a federal appellate judge rules directly on file-sharing-as-fair-use. From BMG Music v. Gonzalez, issued last Friday; opinion by Easterbrook, J.:
Gonzalez was not engaged in a nonprofit use; she downloaded (and kept) whole copyrighted songs (for which, as with poetry, copying of more than a couplet or two is deemed excessive); and she did this despite the fact that these works often are sold per song as well as per album. This leads her to concentrate on the fourth consideration: “the effect of the use upon the potential market for or value of the copyrighted work.”
*2 As she tells the tale, downloading on a try-before-you-buy basis is good advertising for copyright proprietors, expanding the value of their inventory. The Supreme Court thought otherwise in Grokster, with considerable empirical support. As file sharing has increased over the last four years, the sales of recorded music have dropped by approximately 30%. Perhaps other economic factors contributed, but the events likely are related. Music downloaded for free from the Internet is a close substitute for purchased music; many people are bound to keep the downloaded files without buying originals. That is exactly what Gonzalez did for at least 30 songs. It is no surprise, therefore, that the only appellate decision on point has held that downloading copyrighted songs cannot be defended as fair use, whether or not the recipient plans to buy songs she likes well enough to spring for. See A & M Records, Inc. v. Napster, Inc., 239 F.3d 1004, 1014-19 (9th Cir.2001). See also UMG Recordings, Inc. v. MP3.com, Inc., 92 F.Supp.2d 349 (S.D.N.Y.2000) (holding that downloads are not fair use even if the downloader already owns one purchased copy).
Although BMG Music sought damages for only the 30 songs that Gonzalez concedes she has never purchased, all 1,000+ of her downloads violated the statute. All created copies of an entire work. All undermined the means by which authors seek to profit. Gonzalez proceeds as if the authors’ only interest were in selling compact discs containing collections of works. Not so; there is also a market in ways to introduce potential consumers to music.
Think of radio. Authors and publishers collect royalties on the broadcast of recorded music, even though these broadcasts may boost sales. . . . Downloads from peer-to-peer networks such as KaZaA compete with licensed broadcasts and hence undermine the income available to authors. This is true even if a particular person never buys recorded media. . . . Many radio stations stream their content over the Internet, paying a fee for the right to do so. Gonzalez could have listened to this streaming music to sample songs for purchase; had she done so, the authors would have received royalties from the broadcasters (and reduced the risk that files saved to disk would diminish the urge to pay for the music in the end).
Licensed Internet sellers, such as the iTunes Music Store, offer samples–but again they pay authors a fee for the right to do so, and the teasers are just a portion of the original. Other intermediaries (not only Yahoo! Music Unlimited and Real Rhapsody but also the revived Napster, with a new business model) offer licensed access to large collections of music; customers may rent the whole library by the month or year, sample them all, and purchase any songs they want to keep. New technologies, such as SNOCAP, enable authorized trials over peer-to-peer systems. . . .
Authorized previews share the feature of evanescence: if a listener decides not to buy (or stops paying the rental fee), no copy remains behind. With all of these means available to consumers who want to choose where to spend their money, downloading full copies of copyrighted material without compensation to authors cannot be deemed “fair use.” Copyright law lets authors make their own decisions about how best to promote their works; copiers such as Gonzalez cannot ask courts (and juries) to second-guess the market and call wholesale copying “fair use” if they think that authors err in understanding their own economic interests or that Congress erred in granting authors the rights in the copyright statute. Nor can she defend by observing that other persons were greater offenders; Gonzalez’s theme that she obtained “only 30” (or “only 1,300”) copyrighted songs is no more relevant than a thief’s contention that he shoplifted “only 30” compact discs, planning to listen to them at home and pay later for any he liked.
You can argue with the economics: How much of the drop in music sales is really attributable to file sharing? You can argue with the logic: Are the downloads really substitutes for music that Gonzalez would have purchased? And if downloading is a substitute for radio or advertising, then isn’t consumer free riding OK? But it’s going to be hard to distinguish this case in the future. It sure looks like Gonzalez was trying to get something-for-nothing. With Judge Easterbrook’s imprimatur, Gonzalez may turn out to be the ProCD v. Zeidenberg of copyright law: a case that takes a complex issue and treats it both reductively and persuasively. (Ironists take note: we already have a ProCD v. Zeidenberg of copyright law. It’s ProCD v. Zeidenberg.) As Eric Goldman points out, there should be room for fair use at the metaphoric file sharing table, even if Gonzalez’s “no harm, no foul” argument fares badly, in rhetorical terms, for fair use proponents. But this opinion is going to make it hard for fair use to find a seat.
UPDATE: Bill Patry has some typically insightful analysis of this case. And over at Ideoblog, Geoff Manne questions some of my comments above. I’ve replied to him in the comments over there.