The argument went like this. Like the VCR, the Napster service had substantial noninfringing uses. It allowed bands to expose their music to the world through the "New Artists" program. It made it easy to share music which was no longer under copyright. These uses clearly do not infringe copyright. There were also the claims that it permitted "space-shifting" by consumers who already owned the music or "sampling" of music by listeners as they decided whether or not to buy. One could argue that space- shifting and sampling were fair use (though in the end the court disagreed). But since we have two clear noninfringing uses, the technology obviously does have substantial uses that do not violate copyright. Thus, Napster cannot be liable as a contributory infringer, just as Sony could not be liable for the Betamax. Supreme Court precedent covers this case. The Ninth Circuit is bound by that precedent. All the judges can do, goes the argument, is to apply the words of the rule laid down in Sony, say that Napster wins, and move on to the next case. If Congress wants to make services like Napster illegal, it is going to have to pass a new law. The boundaries of the Sony rule are clear and Napster fits within them. (Of course, the last point is subject to argument, but the argument for Napster on this issue was a good one. Not overwhelming—there were more noninfringing uses in the Sony case because the normal way consumers used the technology in question was found to be a fair use—but certainly powerful.) 67
A more daring strategy was to suggest that all the copying done over Napster was fair use, or at least presumptively fair. In Sony, the Supreme Court had said that the law presumes that noncommercial private copying—such as taping a show at home for future viewing—is a fair use. This presumption shifts the burden to the copyright holder to prove that the practice caused harm to them. Copying on Napster was done by private individuals. No money was exchanged. Does this mean we must presume it was fair use and require the music companies and songwriters to show clear evidence of "market harm" if they want to convince us otherwise? 68
It sounds as though proving market harm would be pretty easy. How could millions of people exchanging hundreds of millions of songs not be causing harm? But it is more complicated. Remember the Jefferson Warning. We are not talking about swiping shoes from a shoe store. There one merely has to show the theft to prove the loss. By contrast, music files are copied without being "taken" from their owner. The record companies would have to show harm to their market—the people downloading who do not purchase music because it is available for free. Those who download, but would not have purchased, do not count. And we have to balance those who are deterred from purchasing against those who purchase a whole CD because they are exposed to new music through Napster. One very interesting empirical study on the subject indicates that the result is a wash, with hardly any measurable effect on sales; the overall drop in CD purchases results from larger macroeconomic issues.26 This study, however, has been subject to detailed methodological criticism.27 Another study shows a weak effect on sales, though rather woundingly it seems to suggest that the result is economically efficient—fewer people end up with music they do not like.28 Other studies, by contrast, support the record company position—suggesting that illicit file sharing does indeed undercut sales of both CDs and authorized digital downloads.29 Given the complexities of the issue, the record companies did not want to engage in a war of dueling empirical studies. 69
So, if Napster's users were not infringing copyright law in the first place—at least until the record companies came up with convincing evidence of market harm—because their copying was noncommercial, then Napster could hardly be guilty of contributory infringement, could it? There would be no infringement at all! 70
You could see Mr. Boies's arguments as simple equations between the cases. 71
* Noninfringing uses such as recording public domain films and "time-shifting" programs are equivalent to noninfringing uses such as the New Artists program or sharing public domain music (and maybe "space-shifting" one's own music?); or * Private noncommercial videotaping is equivalent to private noncommercial file sharing. Both are presumptively fair uses. * Either way, Sony=Napster and Napster wins. 72
Napster did not win, of course, though when the judges handed down their decision it was clear they had been paying attention to Mr. Boies, at least enough to make them very wary of tampering with Sony. They claimed that they were upholding that case, but that Napster could be liable anyway. How? Because there was enough evidence here to show that the controllers of Napster had "actual knowledge that specific infringing material is available using its system, that it could block access to the system by suppliers of the infringing material, and that it failed to remove the material." There was indeed evidence that Napster knew how its system was being used—an embarrassing amount of it, including early memos saying that users will want anonymity because they are trading in "pirated music." Then there were nasty circumstantial details, like the thousands of infringing songs on the hard drive of one particular Napster employee—the compliance officer tasked with enforcing the Digital Millennium Copyright Act! (The recording company lawyers waxed wonderfully sarcastic about that.) 73
But despite the ludicrously dirty hands of Napster as a company, lawyers could see that the appeals court was making a lot of new law as it struggled to find a way to uphold Sony while still making Napster liable. The court's ruling sounded reasonable and clear, something that would only strike at bad actors while paying heed to the Sony Axiom and the assurance of safety that the rule in Sony had provided to technology developers for the previous twenty years. But hard cases make bad law. In order to accomplish this piece of legal legerdemain, the court had to alter or reinterpret the law in ways that are disturbing. 74
The first thing the court did was to reject the argument that the "sharing" was private and noncommercial. As to the idea that it is not private, fair enough. Sharing one's music with fifty- four million people does not sound that private, even if it is done for private ends, in private spaces. What about noncommercial? Embracing some earlier rulings on the subject, the court said a use was "commercial" if you got for nothing something for which you would otherwise have to pay. On the surface this sounds both clever and reasonable—a way to differentiate home taping from global file sharing—but the argument quickly begins to unravel. True, the Betamax owners could get TV shows for free just by watching at the regular time. But they could not get a copy of the show for free at the moment they wanted to watch it. That was why they taped. One could even argue that Napster users would have access to most songs over the radio for free. But lawyers' quibbling about which way the rule cuts in this case is not the point. Instead, we need to focus on the change in the definition of "commercial," because it illustrates a wider shift. 75
Remember, a finding that a use is "noncommercial" makes it more likely that a court will find it to be legal—to be a fair use. The old test focused mainly on whether the motive for the copying was to make money. (A different stage of the inquiry concerned whether there was harm to the copyright holder's market.) The Napster court's test concentrates on whether the person consuming the copy got something for free. Instead of focusing on the fact that the person making the copy is not making money out of it—think of a professor making electronic copies of articles for his students to download—it focuses on the presumptively dirty hands of those who are "getting something for nothing." But lots of copyright law is about "getting something for nothing." 76