I draw three lessons:

1. When someone accuses you of infringement and asks why you did not obtain legally required licenses, do not reply, as Ellen’s producers allegedly did, that you don’t “roll that way.” Otherwise, your adversary will reply:

“As sophisticated consumers of music, Defendants knew full well that, regardless of the way they rolled, under the Copyright Act, and under state law for the pre-1972 recordings, they needed a license to use the sound recordings lawfully”

2. Despite the usual tendency of major TV and film studios to be much more cautious than necessary about IP clearance and licensing, sometimes they screw up too. Even when they are, as the plaintiffs point out, “sophisticated consumers of music.”

3. I always wondered if there were some legal remedy for those dances. Once again, IP comes to the rescue.

1. Judge Gertner will reduce the damages somewhat.
2. She will find that the statutory damages provisions of the Copyright Act do not contravene constitutional protections under the Gore line of cases.
3. The First Circuit will affirm.
4. The Supreme Court will deny cert.

I think the damages provision might be vulnerable in a specific defendant’s case (though Ms. Thomas-Rasset would be a better test than Mr. Tenenbaum here), but is safe on its face. In lawyerspeak, it’ll survive a facial challenge, but might fail as-applied.

The Gore limits depend in part on the concept of notice: defendants should know ahead of time how much they’d be liable for if they violate the law. No one expects punitive damages of 500:1 (Gore) or 145:1 (State Farm). But predicting liability – at least at its minimum / maximum amounts – is easy for copyright law. That’s a key difference between a statutory damages scheme, with a range specified by the legislature, and a common-law one where juries pick a number from a hat.

Second, the range of damages in the Copyright Act looks reasonable on its face. $30,000 per work (and up to $150,000 for willful infringement) is a lot, especially if it’s just to deter (or compensate for harm by) a single defendant. (General deterrence is out under Philip Morris v. Williams, which is sad for law & econ thinkers.) Imagine a business that runs off copies of “Harry Potter and the Deathly Hallows” in its basement and sells them. Copyright infringement of this one work is clear, but the business carefully shreds all evidence of sales. So, it’s impossible to prove actual damages; businesses are often risk-averse, meaning that higher awards of damages are needed to deter; and there’s only 1 copyrighted work at issue. Statutory damages are important to provide any deterrence – since proof of harm is under the infringer’s control – and since the infringement might be quite profitable, an award might need to be high (even $150K). Hence, the damages scheme is clearly rational in at least some cases.

The harder question is whether the unconstrained jury discretion for statutory damages could run afoul of due process protections. Individual downloaders tend to be pretty similar if you think about it: there’s not much difference between Thomas and Tenenbaum. So why is her penalty almost 4 times more per work than his, for the same type of infringement? Neither has much in the way of monetary resources, so they’re either undeterrable, or able to be deterred at a fairly low amount (marginal value of a dollar and all that). Here is where the damages scheme seems like it might be vulnerable: it does get hard to predict liability in some individual cases, and the wide range of damages looks a bit too much like absolute discretion. (Thought exercise: what if a jury could award any amount of damages per infringement? Would that improve deterrence against Tenenbaum and Thomas? Would it be significantly less accurate than the actual damages, which everyone agrees are pretty low in real terms? But such a framework would likely run afoul of constitutional limits.)

If this is right, it means that both sides should worry – as should Congress. Getting damages right is important, but preserving both procedural and substantive protections for defendants is just as much so. Comments and disagreement welcomed…

I wasn’t at the courtroom, so I’m relying on reporting / blogs, but I think they’re wrong. Here’s why.

First, Tenenbaum’s attorneys failed to object to the liability question. So, it’s not preserved for appeal. That’s bad, unless the First Circuit decides to tackle it sua sponte, which they won’t.

Second, look at Joel’s actual testimony (quotes from Ray Beckerman’s helpful site, emphasis mine):

“He also testified that he had used the sublimeguy14 username, admitted that he had used KaZaA, and that the KaZaA shared folder in the screenshots from MediaSentry were his. He also testified that it was not uncommon for him to see other people uploading files from him on the KaZaA traffic tab.”

“He testified that he had burned CDs of the music in his shared, and testified that he had ripped CDs to his computer.”

“He testified that he had listened to, talked about, made mixes of, and made available for distribution all of the music in his shared folder.” [ignore the distribution part]

“The redirect was very short… He was asked if he was now admitting liability, to which he said yes. ”

Even throwing out the redirect, if Beckerman is reporting this accurately (I trust him), Tenenbaum has admitted to facts that constitute violations of 17 USC 106(1), 106(2), and 106(3). The liability bit came on redirect and can be ignored without affecting the outcome. The plaintiffs thus clearly made out their case on chief on infringement, and since Tenenbaum’s fair use defense was shot down ahead of time, it was all over but the shouting (and the damages calculation)…

So it’s interesting to speculate what the judge might make of the damages a jury just awarded to the record label plaintiffs in the Jammie Thomas retrial: $1.92 million — more than 8 times the amount awarded in the first trial, or $80,000 (based on the jury’s finding of willfulness) for each of the 24 works Thomas infringed.  Remittitur motion, anyone?

With a seemingly impecunious litigant like Thomas, it probably makes little difference whether the jury awarded $1,920,000, or $222,000, or “a bazillion kajillion dollars”; I know of nobody who seriously expects the record labels to see more than a tiny fraction of the recompense from Thomas they claim they are owed.  And, as I’ve noted before, if you want to campaign for reducing the maximum statutory damage awards for copyright infringement, Thomas is probably not the most sympathetic candidate to make that argument.

It’s certainly an extraordinary number, however, and it just shows how quickly individual acts of file-sharing can pile up into multi-million-dollar liability under current law.

UPDATE: EFF’s Fred von Lohmann ably tees up the constitutional excessiveness issue also raised by Derek following my earlier post. Not a topic upon which I feel qualified to opine, but see Fred’s post and judge for yourself.

In some ways this is better, and in some ways worse. Fingerprinting can catch a lot of infringement, but it’s a rule rather than a standard: there’s no way for a content ID system to figure out fair use. (Heck, lawyers are bad at it.) The dispute resolution form improves things, but I wish the system notified the poster first (with, say, a deadline for response) before taking down the allegedly infringing content.

This is a nice reminder that the Internet is a world of private power. There’s no right to post to YouTube, and posters get whatever process Google decides to afford them when content appears to be infringing. In many ways, this recapitulates the standard public choice problems of copyright law: copyright-owning interests are concentrated and powerful, and copyright-using interests tend to be dispersed and weaker. In that sense, we’re probably fortunate that the content ID take-down system is as thoughtful as it apparently is. /Update

Warner Music has issued a take-down notice under the Digital Millennium Copyright Act for the video slideshow of Larry Lessig’s keynote talk at the OFC Conference in San Diego in March 2009. This can only be viewed as 1) a sad commentary on automated detection of copyrighted material, 2) a serious error in judgment, or 2) a deliberate provocation. Naturally, Lessig is going to fight, which I assume begins with a counter-notification under Section 512(g)(3) of the DMCA. I’ve been going over the myriad of clips in Lessig’s talk, trying to figure out which might be the source of Warner’s notice. It’s an interesting question whether Warner might be liable under 512(f) of the DMCA, along the lines of Lenz v. Universal’s claims, for failure to issue this notice in good faith. (Is it sufficient to believe in good faith that all uses require permission, or to hold a completely unreasonable yet devoutly believed view on the topic? I’m dubious – most subjective standards have some objective grip at bottom.) It’s hard to see Lessig’s utilization of the clips as anything but fair use: they’re quite abbreviated, the use is in a non-traditional educational setting, and there’s no market displacement of the originals. Not sure, in other words, how this fight started, but I have a sense of how it’ll end…

One of those still-pending cases settled today for $7,000 (which seems generally in line with prior settlements, in view of the fact that there were two accused downloaders).  I found this comment from the story particularly striking (emphasis mine):

“We are pleased to have reached an agreement with the Santangelos,” Cara Duckworth, a spokeswoman for the R.I.A.A., said in an e-mailed statement. …

She said the lawsuit had succeeded in showing that breaking the law has consequences and in steering music fans toward legal online services “that fairly compensate musicians and labels.”

The really interesting phenomenon, however, is just how resistant P2P users’ norms proved to be to the RIAA’s litigation campaign. Indeed, if the campaign was actually “steering fans toward legal online services,” it would doubtless still be underway.  If P2P users now understood “that breaking the law has consequences,” the entertainment industry would not be complaining so loudly about piracy.  What actually happened was, the RIAA sued a bunch of people, and users apparently reacted with a shrug, so they stopped suing.  Why did the RIAA’s years-long (and expensive) litigation campaign fail to sway hearts and minds?  More generally, has suing people ever been an effective way to alter their opinions?  Those are the questions that the content industries ought to be trying to answer.

As Tim and I explained in those past years, showing a broadcast of the game on a big-screen television “of a kind commonly used in private homes,” without charging admission, is almost surely allowed under the “homestyle exception” to copyright law. The trademark claim is even weaker, because simply holding a “Super Bowl viewing party” is a reference to the mark surely covered by various trademark fair use theories (notwithstanding their other flaws).

So, it’s no surprise that this year — wait, what’s this? STOP THE PRESSES! Ladies and gentlemen, can you believe it?!? The NFL now says that it will not stop churches from doing what the law allows and holding the parties. As Tim and this fellow IP blogger suggest, proposed legislation by Senator Specter may have played a role in the league’s change of heart. But my whole view of the universe is shifting nonetheless. The RIAA has stopped its lawsuits, iTunes has abandoned DRM, the TV networks now put episodes up online for free, and there is a pig sailing through the air outside my window.

Don’t worry, though. We can still count on the annual “Network Bans Sexy Super Bowl Ad” controversy. (Concerns included “licking eggplant” and other steamier issues…)

1. Public choice problem – IP profs know they just aren’t that influential with Congress, compared to industries (such as movies and music) with a significant economic stake. Tilting at windmills isn’t a good use of time.
2. No big deal – the Act doesn’t make enough substantive changes to IP law (focusing mostly on counterfeit trademarks and setting up a new IP bureaucracy in the federal government) to rally profs to the barricades. It’s not much of a windmill, anyway.
3. Advancing the ball – the Act is actually a good idea. Windmills can be useful things.

I’d be interested to know what you think.

Update (Oct. 15, 1:30PM): Susan Scafidi has an insightful post at Concurring Opinions. My only quibble: do we want an IP czar generally? Larry Lessig would be terrific, but administrations come and go; the structure remains. (Particularly given the bipartisan consensus on this bill.)

I was thinking of Cecilia Gonzalez’s case when reading the news that the trial judge in the Jammie Thomas file-sharing case (previously covered here) has ordered a new trial. The core legal issue on which the new trial order turned isn’t the size of the damage award ($222,000 based on 24 infringed works). Rather, it’s whether a copyright plaintiff must prove, in order to show a violation of its exclusive right under § 106(3) “to distribute copies or phonorecords of the copyrighted work to the public,” that an actual distribution or transfer occurred. The court initially answered that question in the negative; now, apparently moved by more recent cases holding that merely “making available” a copyrighted work for download isn’t enough, it has changed its mind. But even though the new trial in Thomas’s case rests on the court’s evolving understanding of the distribution right, the court’s opinion concluded with a lengthy passage criticizing the same court’s own damage award against Thomas as irrational. From pp. 42–43 of the court’s new trial order:

While the Court does not discount Plaintiffs’ claim that, cumulatively, illegal downloading has far‐reaching effects on their businesses, the damages awarded in this case are wholly disproportionate to the damages suffered by Plaintiffs. Thomas allegedly infringed on the copyrights of 24 songs — the equivalent of approximately three CDs, costing less than $54, and yet the total damages awarded is $222,000 — more than five hundred times the cost of buying 24 separate CDs and more than four thousand times the cost of three CDs. While the Copyright Act was intended to permit statutory damages that are larger than the simple cost of the infringed works in order to make infringing a far less attractive alternative than legitimately purchasing the songs, surely damages that are more than one hundred times the cost of the works would serve as a sufficient deterrent.

I’m inclined to agree that awarding $750 for infringing a 99¢ copyright is a pretty stiff punishment, and that the magnitude of possible injustice increases rapidly as we begin to multiply each of those figures times greater and greater numbers of infringing acts. And let’s not forget how easy it is, in the file-sharing context, to cross the line into criminal liability: under § 506(a)(1) as amended by the No Electronic Theft Act, all that is required is “reproduction or distribution … during any 180–day period, of … 1 or more copyrighted works, which have a total retail value of more than $1,000.”

If you want a poster child for the excesses of copyright damage awards, however, it’s probably best not to pin your hopes on Jammie Thomas. The facts of the case, at least as I understand them (and I never followed the trial as closely as Bill did) seem to suggest that Thomas in fact did share copyrighted songs with unknown numbers of other persons. (Fred Yen, for one, seems to think the RIAA is not going to have much trouble proving actual distribution if the retrial happens.) The apparent unfairness of such high multipliers between the damage award and the underlying actual harm may abate somewhat if the court is persuaded that the underlying harm is actually greater, due to the possibility that many other file-sharing users may have downloaded songs from Thomas.

But there’s a better candidate out there for illustrating the excesses of the existing statutory damage award scheme, and that’s Cecilia Gonzalez, who is not alleged to have shared a single song with anyone else. (In technical terms: she’s a downloader, not an uploader.) By the time we’ve multiplied $750 times 30 to reach Gonzalez’s damage award of $22,500, the question whether that award is excessive in view of the $15 in actual harm to the plaintiffs practically answers itself.

More on the new trial order in the Thomas case via Ars Technica, EFF, The Register, and Slashdot.