Reforming Copyright Law of Damages to Avoid Results Like “Blurred Lines” and “Dark Horse”

How Can Copyright Law Be Reformed to Avoid Results Like “Blurred Lines” and “Dark Horse”? Part 2.

Damn the Damages! Full Speed Ahead!

This is the third in a series of posts about the dread supposedly wrought by recent music copyright cases (all of which were filed in the same U.S. District Court), like the “Blurred Lines” and “Dark Horse” cases. In the first post, I questioned whether these cases really were hurting songwriting, at least when viewed from copyright law’s main policy goal, but admitted that these cases seemed to expose real problems with copyright law. In the next post, I suggested two areas for reform: (1) a better grasp of the key concept of scènes à faires; and (2) adoption of the Laureyssens test for substantial similarity, which is in crisis in many courts.

In this post, I’ll continue suggesting areas for reform, focusing on damages.

“Yes, Lord, but even dogs eat the scraps that fall from their master’s table.” Not just a nervy comeback, but also a theory of damages.

American Generosity

Damages are chronically overlooked by everyone. They’re an afterthought. The reason is easy to understand. If you’re the defendant, focusing on damages looks like you’re admitting liability. A “win” on damages is only a partial win. Better to put your resources into fighting liability and/or proving your affirmative defenses. This strategy isn’t completely wrong, either. There’s reason to believe that juries take the closeness of its liability determination into account when setting damages.¹The basis for this belief is the idea that juries are usually harsher when they are empaneled solely to determine damages, such as when there’s a retrial or remand just on damages, e.g., the Thomas-Rasset case.

All the same, bad damages awards have pernicious knock-on effects. They distort the market for not only the settlement of disputed claims, naturally, but also license royalties, which are at least influenced by the prospect of big damage awards. They distort the market for insurance. They also distort justice. Just the thought of huge damage awards can and will cause innocent parties to settle, just to be free of the threat. At the same time, they encourage weak lawsuits because the upside starts to make even low-chance suits look financially viable (which also makes those suits harder to settle). Their effect on settlement is to raise the stakes. This will cause some parties to capitulate (and pay more than they should), but others will see the suit as an existential threat that they have to win.

For whatever reason, American juries are just very, very generous. This can seem paradoxical. On the one hand, we respect and admire large corporations as important components of our economy. We hope they bring employment to ourselves or our neighbors—e.g., please, Amazon, come to our town and employ us! On the other hand, juries eviscerate large corporations almost every chance they get. My own thinking is that we are grateful for what we get, but we think we might deserve more. So, if we sometimes feel like we’re the dogs under the table happily eating scraps, we’re not above giving the table leg a little push to knock a few more scraps onto the floor. Others, I’m sure, have their own theories.

Regardless, this problem goes well beyond copyright law and would require a legal sea-change to affect it.

Copyright Damages: An Introduction

Still, copyright has its own particular problems with damages. To explain, let me first just summarize how damages work in copyright law. First, copyright has two broad categories of damages: (1) statutory damages, where the jury picks a number from a range with very little guidance; and (2) what we might call traditional damages, which are very much like “actual damages,” except that “actual damages” are a subset of these damages, so we can’t call them “actual damages.” So let’s just call them “traditional damages.”

In these music cases, statutory damages are irrelevant. They only get significant²For a certain value of “significant.” if there are many copyrighted works involved. Usually with these music cases, there’s just the one. So set those aside for now.

Traditional damages are comprised of two overlapping but distinct types of damages: actual damages (see?) and disgorgement of (the defendant’s) profits.

Well, Actually….

Actual damages are … kind of hard to define precisely. They’re whatever damages that flow foreseeably from the infringement. That’s a general principle of law, and not anything unique to copyright law. Some examples might help.

• The infringement consists of an exact duplicate of the copyrighted song, which undercuts the price for legitimate copies of the song. Sales of infringing copies thus displace sales of legitimate copies. The copyright owner would be entitled to profits it would have made (“lost profits”) on displaced sales of legitimate copies.³Counter argument: maybe buyers of the infringing copies wouldn’t have paid what the legitimate sellers were asking for, so not all such sales displaced legitimate sales.
• The song is being used in a commercial without permission. Ordinarily, one would pay a license for this use, but the infringer didn’t. The copyright owner is entitled to the fair market value of the license it would have granted had it been asked (a “reasonable royalty”).
• The infringer put infringing copies in a place where they can never be removed, and legitimate sales fall off a cliff. The copyright owner might be entitled to the fair market value of the copyright itself, or some portion thereof (“diminution of value”).

Note that these examples assume exact or near-exact duplication of the copyrighted song (or “literal infringement”). These theories don’t work as well for the types of partial or non-literal infringement that we saw in the “Blurred Lines,” “Dark Horse” and “Stairway to Heaven Cases.” Reasonable royalty is flexible enough to apply to most situations, but it requires a lot of heavy lifting by experts.

Gorging on Profits

Disgorgement is much easier to understand and much, much easier to prove. The theory is that the infringer isn’t entitled to the profit that it earned from infringing the copyrighted work, and that profit should instead be given to the copyright owner. The burden of proof on the copyright owner is fairly mild. It just needs to prove (1) a connection (“nexus”) between the profits and the infringement (2) the gross revenue earned from those acts of infringement. Sometimes the “nexus” is hard to prove. Consider, for example, an infringing jingle used in a commercial to sell cars: is the copyright holder entitled to the profits of the cars sold during the time the commercial aired? But in these music cases, the nexus isn’t hard to prove. Indeed, in the “Blurred Lines” and “Dark Horse” cases, the parties mostly agreed on the numbers.

Expenses and other reductions must be proven by the infringer, on the theory that the information is already in its hands and it has the motivation to prove them up. In literal-infringement cases, this isn’t hard. But in non-literal infringement cases, like “Blurred Lines” and “Dark Horse,” the defendant gets the opportunity to further reduce the profits through “apportionment,” and this lies at the heart of the damage awards in these cases.

Stumbling Blocks

I’ll discuss apportionment in more detail below. Before I do, I want to briefly discuss two other significant stumbling blocks:

First, actual damages and disgorgement usually overlap, and you’re not allowed to double-count damages. For example, if you prove that your sales of your copyrighted song were diverted to the defendants, then your lost profits will overlap the defendants’ profits. If the defendants are more efficient at infringing your song than you are at selling it, your profits will likely be entirely subsumed by the defendants’ profits, so you’d only recover defendants’ profits. But if the defendants so undercut your sales that they barely made any money, your lost profits would subsume their profits.⁴There are some difficult cases, too. What if you prove you’re owed a reasonable royalty, say 5% of the defendants’ revenues? Should that amount be deducted from your award of the defendants’, since their profits would have been reduced by the they would’ve paid you in royalties?

Second, reasonable royalty awards can be unreasonably high, because they’re calculated in isolation and with hindsight. In theory, calculating a reasonable royalty requires an exercise in imagination: if the two parties were negotiating the royalty rate without rancor, long before a lawsuit was even contemplated, what would they have arrived at? This exercise is impossible. The infringing song has already been written. Without the infringing elements, it would be a different song. And with a lawsuit, the copyright owner might have stopped the release of the song altogether, no matter how significant or insignificant the contribution of infringing elements to its success.

Watch Your Portions!

OK, back to apportionment. Recall that the profits must be reasonably connected to the infringement. Now apply that to non-literal infringement. We do this by asking how much the bits taken from the copyrighted work drove sales of the infringing work.

There’s easy apportionment and difficult apportionment. Easy apportionment is required when the infringing song is bundled and sold with other non-infringing songs, say in an album. The analysis isn’t quite as straightforward as dividing the album’s profits by the number of songs. Some songs drive sales of the album more than others, e.g., hit songs versus deep cuts. Consider a song that not only was a hit but was released in advance of the album, creating (in theory) demand for the album. Even “easy” apportionment requires some approximating.

Of course, albums aren’t as important as they used to be, so let’s move onto hard apportionment. Let’s say we can confidently calculate the profits for the infringing song, but only a small portion⁵The word “portion” is doing a lot of work here because it means more than just a discrete bit of the copyrighted song. It would also include the “constellation” of elements at issue in the “Blurred Lines” case, which isn’t so much a collection of bits as the way the bits were combined. of the infringing song was taken from the copyrighted song. How much did that small portion drive sales of the song? Do you try to imagine the song with a different element and decide how important the infringing element is (aesthetically)? If so, is the hypothetical element equally as good (aesthetically) as the actual infringing one, or do you use a kind of “replacement-level” element as a kind of base line?

Hit songs aren’t granular, unfortunately. Hit songs are subject to tipping points. (Nowadays, we might say “virality.”) What I mean is: if you replaced one element with another, you don’t necessarily just reduce its profitability by some proportional amount. It might mean the song is never a hit. That will probably be true for more than one element. For popular music these days, you might say the “beats” (meaning a combination of rhythm and semi-melodic bass) is crucial, and the top-line (lead melody) is crucial. What about the lyrics? Nearly all popular songs have lyrics, but they’re often denigrated as interchangeable, but yet again, lyrics don’t just write themselves. Let’s just assume they’re crucial, too.

So if the infringing song took (as in the case of “Dark Horse”) the “beats,” and you conclude the song would not have been a success without the “beats,” do you apportion 100% of the profits to the infringement? That makes sense, but can’t be right, either, because you would have to assume the crucial elements of the song add up to 300% of its value.

OK, so it’s got to add up to 100%. Do you apportion one-third of the profits? Maybe one-quarter to account for all the minor elements? Not so fast.

No, Really Watch Your Portions!

Did you buy “Dark Horse” solely because of the “beats,” or the combination of “beats,” “top line” and melody? Don’t you like the way Katy Perry sings? What if the sound engineering stunk, and the song sounded like a reverb-party? What if the rapping was just embarrassing? You assume a certain level of quality in popular music, but that quality doesn’t come from nowhere. What’s more, that quality you’re enjoying is protected by a separate copyright that belongs to the singers, musicians, and possibly sound engineers and producers—the people who make the music sound good. The copyright holder in the underlying musical composition can lay no claim to that.

We’re not done. How did you even learn about “Dark Horse”? It wasn’t just the quality of the song and the quality of the sound recording. Do you think that, when you listen to the radio, the radio station has performed some massive ongoing curation project to find the very best songs, recorded in the very best way? In large part, you learned about it because of Katy Perry. Not her singing, but her fame. She’s popular independently of her music, and her popularity makes it much more likely that a radio station will play it, and that it will show up in your Spotify feed. Heck, you might be a huge fan and seek out her music on your own.

We’re still not done. Behind Katy Perry is a label and its marketing department. Labels might not develop new talent much any more, but they promote the heck out of established, bankable stars like Katy Perry. They also trust producers like Dr. Luke to produce quality content. Not groundbreaking, heartbreaking, genre-breaking content, but content that definitely doesn’t suck. And if you didn’t learn about “Dark Horse” through radio (which itself has been subject to this promotion), Spotify, or your friends, you must’ve learned about it through some other promotion.

If you’re sitting at 25% apportionment based solely on marketable aesthetics, how much to you cut that amount to take into account everything else that makes a hit song a hit song?

You Mean It Doesn’t Matter Who Sings the Song?

If you’re a juror, you might not cut it at all. Sticking with the “Dark Horse” case, the jury decided that 22.5% of the song’s profits were attributable to the infringement. From one perspective, that seems impossible because it requires us to conclude that everything else—the rest of the song’s elements, the various performances, the engineering & production, the star power, the labels’ marketing machine—amounts to only 81.8% of the song’s profits. Get rid of Katy Perry (i.e., replace her with a replacement-level talent), and what do you have? Get rid of the whole rap in the middle, and what do you have? Get rid of radio play, Spotify playlists, Katy Perry’s promotion of the song, and how much money do you “Dark Horse” would have made?

If I’m right about the psychology of the American juror, this decision is understandable, though. All those external factors, even including the artistry of the performances, are undeserved, from this way of thinking. The song sold a lot, and its sound recording was of impeccable quality, but only because everyone involved had a lot of money and power. They don’t deserve to profit from it. They already have enough.

Jury Destructions

There might be a different (or additional) culprit: our old frenemy jury instructions. Just how was the jury instructed on this issue? Here you go:

After determining each defendant’s profit, you must then determine whether any portion of the profit from the sale of [“Dark Horse”] is attributable to factors other the use of ostinato #2 [the “beat” at issue] in “Dark Horse.” The defendants have the burden of proving the percentage of their profits, if any, attributable to factors other than infringing the copyrighted work.

Hey, this isn’t terrible! It’s only merely bad. I’ll show you terrible in a moment. Like all bad jury instructions, it’s a correct statement of the law, but doesn’t tell the jury what it needs to know. My main criticism is that the instruction starts with the assumption that “ostinato #2” accounts for 100% of the profits, and as anyone who’s studied negotiations theory would guess, that 100% figure will act as magnet pulling the jury’s number upward. But, then again, it’s an accurate statement of the law: you start with the full amount of revenues, then start paring down. Naturally, the results will be on the high side.

At a minimum, I think the jury instruction should specify the categories of “factors” that could reduce the award, so we at least know they were considered (and rejected). But at least the defense counsel has something to work with when trying to make a case for apportionment—which the defense attorneys totally did. “The reason people buy a Katy Perry song or album is because it’s Katy Perry.” And also: “[The plaintiffs] are discounting efforts of artists and songwriters to make ‘Dark Horse’ what it is.” (But, again, maybe the jury just didn’t want to further reward Katy Perry for being Katy Perry.)⁶Apparently, the plaintiffs argued that the apportionment should be 45% because “ostinato #2” appeared in 45% of the song, which is just an awful argument. But also, why even concede that much? A terrific chorus might account for 20% of a song but might be the main thing listeners remember of like about the song.

Would you like to see a terrible jury instruction. Here’s the one from the “Blurred Lines” case:

Unless you find that a portion of the profit from the use or exploitation of [“Got to Give It Up”] is attributable to factors other than the use or exploitation of [“Got to Give It Up”], all of the profit is to be attributed to the infringement. The Thicke Parties have the burden of proving the portion of the profit, if any, attributable to factors other than infringing [“Got to Give It Up”].

Again, 100% accurate. But this one is much more misleading than the one in the “Dark Horse” case. The takeaway is that apportioning at all is highly unusual. You’re not just starting at 100% apportionment. The defendant has work to do before you’ll even consider moving off 100%.

For what it’s worth, the jury awarded $4 million in actual damages (later reduced to $3.2 million) and about $3.4 million (later reduced to about $2.1 million) in profits. The verdict form is here. It’s hard to know what the jury was thinking in making any award for actual damages and whether those are just the defendants’ profits in disguise. The parties agreed that the song made $16,675,690 in profits. If you look at just the profits awarded by the jury, you get an apportionment of about 20% (though that gets reduced when the award is reduced).⁷If you combine the two awards on grounds that the jury was confused, then it’s closer to 45%. It’s probably a coincidence that two different juries came up with roughly the same apportionment for two different acts of infringement involving different songs. Regardless, it’s a sign that the jury was unwilling to reduce the award base on “elements of profits attributable to factors other than the copyrighted work.”

First Suggested Reform: The Jury Instructions

Looking at avenues for reform, let’s start with those jury instructions. Here’s the statutory language of § 504(b) of the Copyright Act, which governs non-statutory damages:

The copyright owner is entitled to recover the actual damages suffered by him or her as a result of the infringement, and any profits of the infringer that are attributable to the infringement and are not taken into account in computing the actual damages. In establishing the infringer’s profits, the copyright owner is required to present proof only of the infringer’s gross revenue, and the infringer is required to prove his or her deductible expenses and the elements of profit attributable to factors other than the copyrighted work.

As is usually the case, the jury instructions in both the “Blurred Lines” and “Dark Horse” cases don’t stray too far from this language. If anything, the “Blurred Lines” was the more daring of the two, and yet it was the worse. Although the various elements of damages are contained here, they’re not very clear to non-lawyers. It seems to me a good way to go about a jury instruction is to imagine you’re writing it for a law student about to take an exam without studying. Let’s unpack the statutory language, like so (italics are to be replaced by the names of the parties as appropriate):⁸Apologies if this makes them look like they came out of the King James Bible.

• Instead of statutory damages, the copyright holder may elect to recover “actual damages” AND the infringer’s profits.
• Because the copyright holder doesn’t need to choose between statutory damages and these other types of damages, please calculate both statutory damages and these other types of damages.
• Try to be as precise as you can, but precision isn’t necessary. Do not avoid awarding damages just because you cannot calculate them with precision. The only time you may avoid awarding damages for lack of precision is if calculating them requires speculation—not mere inferences—above and beyond what has been proven.
• Both actual damages and the infringer’s profits are subject to an important limitation: the copyright holder may recover those types of damages only to the extent they are “attributable” to the acts of infringement that you’ve found.
• In the case of actual damages, they are “attributable” only if the acts of infringement you’ve found caused the copyright holder damage. See Jury Instruction No. [##] on “causation” for what it means for one action to cause another.
• In the case of the infringer’s profits, profits are only “attributable” if you can draw a reasonable connection between the acts of infringement you’ve found and revenues the infringer made from the infringing work.
• The burden of proving what damages and profits are “attributable” to the infringement rests initially with the copyright owner.
• Actual damages and the infringer’s profits can overlap. This is to say, the same dollar in damage that the copyright owner suffered might also be part of the infringer’s profit. If there is overlap, it’s imperative that you do not count overlapping damages twice.
• Some types of actual damages are profits the copyright holder would have made if not for the infringement, the royalty the copyright holder could have reasonably collected from the infringer had they willingly negotiated a royalty for the infringing uses you’ve found, and the amount by which the value of the copyright was diminished by the infringing acts you’ve found. These are not the only types of actual damages. If any of these types overlap, be sure not to count them twice.
• Profits are revenues less expenses.
• To prove the amount of the infringer’s profits, all the copyright holder need prove is the amount of revenue the infringer earned that are attributable to the acts of infringement you’ve found.
• The infringer then bears the burden of proving the expenses.
• Insert instruction about overhead expenses.⁹Ugh, this are of the law is unsettled and difficult, and it’s not central to my project, so…
• The infringer may also seek to reduce the amount of profits for which it is liable by proving that one or more portions of the profits are actually attributable to factors other than the acts of infringement you found.
• In a case where the infringing work itself used only a part of the copyrighted work, reduce the award of profits by the other elements in infringing work that weren’t taken from the copyrighted work.
• In a case where the infringing work was sold as part of a larger product, including a compilation, reduce the award of profits by the amount that you attribute to the rest of the larger product.
• When making these reductions, consider how much the infringing portions and other elements drove sales of infringing products. The infringing use might comprise a small portion of the infringing products, but it might still be responsible for a much larger portion of sales of those products. Conversely, the infringing use might comprise a large portion of the infringing products but be responsible for a much small portion of sales of those products.
• Further reduce the award of profits by factors that drove sales of the infringing products that are unrelated to the infringement. For example, take into account the fame of the infringer in driving sales, the marketing efforts made to drive sales, the pre-existing sales networks and channels that the infringer had available and used.

OK, obviously this is a work in progress…

Second Suggested Reform: A Significant Statutory Tweak

My other major suggestion for reform will be harder to pull off, because it will require a revision of § 504(b). You’ll notice that § 504(b) uses “attributable” twice as a way of limiting what profits the copyright owner can be awarded. Because under general law, the burden of proving damages is normally on the plaintiff, the copyright holder bears the burden of proving up the profits. But then the statute shifts part of the burden to the infringers to prove down the profits not only by making them prove their expenses (which makes sense) but by also proving those “elements of profit attributable to factors other than the copyrighted work.”

This is how we end up with jury instructions that make apportionment sound unusual. But in two very common types of cases, apportionment is the norm: non-literal cases (as in the “Blurred Lines” and “Dark Horse” cases) and part-of-a-larger-work cases (e.g., a single infringing song on an album of ten songs).

So my suggested reform is simply to remove the final phrase from §504(b): “and the elements of profit attributable to factors other than the copyrighted work.”¹⁰To be clear, the infringers would still bear the burden of proving expenses. The infringers can still try to reduce the amount of awardable profits by pointing to those other factors, but the jury will no longer be told that’s solely the infringer’s burden. Instead, they’ll just be told that the copyright owner bears the burden of proving the revenues attributable to the acts of infringement. Thus, the jury instructions could just be:

• The copyright owner is entitled to profits made by the infringers that the copyright owner proves are attributable to the acts of infringement you’ve found. The copyright owner bears the burden of proving a reasonable relationship between those acts of infringement and the infringers’ profits.
• When determining which profits and what portion of those profits are attributable to the acts of infringement you’ve found, you should consider primarily how the infringement drove sales of the infringing work or otherwise caused the infringers to earn revenue. Revenue attributable to other factors, such as the infringer’s fame or marketing efforts, should be excluded.
• With respect to calculating the profits that are attributable to the acts of infringement you’ve round, the copyright owner need only prove the revenues earned by the infringers. The infringers bear the burden of proving any expenses.

One final thought: Maybe the American jury is right to ignore the “other factors”? If you combined my song with your marketing power and fame to make far more money than little ol’ me could’ve ever earned—and let’s assume infringement far more, um, obvious than in the “Blurred Lines” and “Dark Horse” cases, is it really that unfair that I should benefit from your power and fame? If you can make money with just power and fame, what do you need my music for?

Thanks for reading.

Rick Sanders

Rick is currently General Counsel for Software Freedom Conservancy. Previously, he has been practicing law as an intellectual-property litigator since 2000.