Opinion: Compensating the record companies for piracy


This article was originally on a blog post platform and may be missing photos, graphics or links. See About archive blog posts.

A federal judge in Boston took the unusual step last week of slashing damages that a jury had ordered a file-sharer to pay the major record companies for infringing their copyrights. Err, well, perhaps that’s not so unusual -- not many of these cases have actually reached juries, and judges have tossed out the damage awards in the two most widely publicized cases so far. What made District Court Judge Nancy Gertner’s ruling unusual was her finding that the amount of damages had violated the Constitution, even though the jury had awarded damages in the lower range of the amounts approved by Congress.

The Times’ editorial board, which has complained in the past about the magnitude of statutory damages in copyright law, will probably weigh in on the ruling within a few days. In the meantime, I asked the folks at BigChampagne to help me with a thought experiment about the right level of compensation for file-sharing infringements.


In the case of file-sharer Joel Tenenbaum, jurors awarded the major record companies $22,500 per song. I wondered what that would work out to if the Recording Industry Assn. of America managed to collect such a sum from everyone sharing those songs. Granted, that’s not going to happen, especially with the industry no longer suing file-sharers en masse. But Tenenbaum was just one person sharing those files -- one of many sources for others to download from. And on a file-sharing network, there’s not much difference between Source A and every other person offering a particular song.

I sent the titles of nine of the songs Tenenbaum had shared to BigChampagne. Diving into the data it collects from file-sharing networks, it reported that those songs -- all rock and rap hits by the likes of Eminem and Nirvana -- were being offered by nearly 6.8 million file-sharing computers. At $22,500 per song, that’s $152 billion -- many orders of magnitude more revenue than those songs could have generated even in a world with no infringement. Assuming those nine songs were representative of all 30 of the tracks Tenenbaum infringed, the amount increases to half a trillion dollars.

Even at the reduced level set by Gertner -- $2,250 per song -- the extrapolated amount is staggering. If all 6.8 million file-sharers with those nine songs were dunned that amount, the labels would collect $15 billion. That’s more than the total amount of music sales in the U.S. each year. Not bad for nine songs.

Updated, 4:32 p.m.: Eric Garland of BigChampagne points out that the numbers his company provided are just for file-sharers in the U.S. If I’m not mistaken, there’s some of that behavior overseas as well.

The point here isn’t to defend Tenenbaum, who first lied about his file-sharing and then tried to defend it in the most self-serving way possible. ‘Art is meant to be shared’? You might get away with saying that, Joel, if you ever paid for it. I’m just noting here that $22,500 per song makes sense only in a world where one infringer is held liable for the acts of millions. It’s true that the Internet enables a single person to become a global infringer. But does that mean juries should act as if each pirate they encounter is devastating the industry singlehandedly?

Updated, 10:37 a.m. Friday: Cary Sherman, president of the Recording Industry Assn. of America, sent in a lengthy reply this morning. Here is it, unedited:

Your analysis (about how much money record companies would earn if all file-sharers paid the statutory damages award) implicitly presumes that statutory damages should reflect actual damages (i.e. the sales revenue that was actually lost). But that’s not what stat damages are about.

Statutory damages were specifically created by Congress as an alternative to actual damages -- in part because actual damages are often difficult to calculate, but even more importantly, to provide meaningful deterrence. How much of a deterrent effect would you have if you reduced statutory damages to some modest amount on the ground that everybody else was infringing, too? The greater the illegal conduct, the less each violator would have to pay, reducing the risk to the point that the deterrent effect is lost. How, then, would you deter the theft?

Congress chose to impose meaningful risk on infringers that they could have to pay substantial damages even if it couldn’t be proven that the actual damages were significant. That’s a sensible way of achieving a policy of deterring infringement. Your ‘experiment’ yields precisely the opposite effect of what Congress intended.

As you may recall, we didn’t ask the jury for any particular level of damages. We left the amount completely up to the jury. I presume that they took into account the fact that Tenenbaum forced us to spend a lot of time and money litigating the facts of a case in which the facts were completely clear except that he lied about them repeatedly, and under oath; that he tried to pin the blame on others; that he continued stealing music even after he was sued; that he never showed any remorse; all that kind of stuff, beyond just the impact on the industry. Isn’t that what you’d want a jury to consider in picking a damage award?

Isn’t that precisely how a jury deters misconduct? Should such egregious misconduct (and disrespect for the court system) be beyond the scope of the jury’s damage award because, if every infringer were also sued, we’d recover too much?

Statutory damages also reflects another basic practicality -- that every infringer is NOT going to be sued. Just like every speeder is not going to be caught. Or every looter captured. So you provide real consequences for the unlucky few who get caught to deter all the others.

Should a looter’s liability be reduced because, if he hadn’t stolen the TV, another looter would have anyway?

I fear your ‘experiment’ proves too much. And your embracing it would give it credibility it doesn’t deserve.

He makes a good point about damages needing to deter other infringers. Nevertheless, I don’t think a jury needs to go to $22,500 per song, or $675,000 in all, to offer an appropriate deterrent in addition to compensating the labels for their losses. Gertner’s formulation -- $67,500 for 30 songs -- strikes me as a whopping deterrent.

Battling file-sharing is a game of whack-a-mole, and copyright holders won’t increase their score by using a bigger hammer.

-- Jon Healey

Illustration by M. Ryder / Tribune Media Services