Proving Damages: What are the most effective ways to calculate and apportion damages? Featuring Sarah Burstein (University of Oklahoma College of Law), John Desmarais '88 (Desmarais LLP), and Kristelia Garcia (Colorado Law), moderated by Christopher Sprigman (NYU Law).
Full video of this panel is available on YouTube.
- All right, our final panel today is all about the harms. It's all about proving damages. And this is the last panel today and the importance is obvious, but as you seen throughout today, there are many different views on how to arrive at answers, at numbers, at definitions, at markets, and so we wanted to talk about at the end of the day today is when there is damages, and when we're talking about damages, what do you do to try and calculate it, especially in spaces like IP that can be so abstract. So I will bring it over to our final panel of the day, thank you.
- Michael, thank you. I'm Chris Sprigman, I teach IP Law here at the law school, and I'm very happy that so many of you have showed up for our conference which I have the privilege of chairing a panel that is gonna look at the question of damages, and it's gonna do so across utility patent, design patent, trademark, trade dress, and copyright. We have a panel that we designed to allow us to range widely across that terrain. So with us today, to my right, is Sarah Burstein from the University of Oklahoma College of Law. Sarah's scholarship focuses on design law with a particular focus on design patent. She's gonna talk a bit about trade dress and design patent damages with an emphasis on the differences in what proof is required for disgorgement of profits across those different domains. To her right is John Desmarais, a graduate of this law school, and John is the founding partner of Desmarais LLP, where he tries cases across the various fields of IP. I should mention that John is chair of the Engelberg Center's advisory board and also former assistant U.S. attorney in the Southern District of New York. So John's here today to talk about the many difficult issues that attend to the enterprise of proving patent damages as some form of a reasonable royalty. So finally we have with us today Kristelia Garcia who's associate professor at the University of Colorado School of Law, in addition to teaching copyright, trademark, and property, Professor Garcia serves as director of the content initiative at the Silicon Flatirons Center for Law, Technology, and Entrepreneurship. Kristelia's research is focused on the intersection of law, technology, and economics, with a particular focus on efficiency, competition, private ordering, and distributive justice. Before becoming a law professor, Kristelia worked in the music industry, both on the distribution side and on the content ownership side. And today Kristelia's gonna talk about a couple of key issues in proving copyright damages. First, the current litigation landscape for statutory damages, and second, the new willing buyer/willing seller standard that the 2018 Music Modernization Act has put in place for setting rates for mechanical reproductions of musical compositions. So a wide range of topics, and I'm hoping we'll have plenty of time to have some interaction between the panelists and then questions from the audience. And with that, Sarah, take it away.
- All right, thank you so much. So I'm gonna just hit things at a really high level 'cause I think that's our plan here so we can have time for questions and discussions. So if this is really, really a 5,000 V, that's why. So for trade dress, trademark infringement, right, we know for a product design, you could get both trade dress protection and design patents. You can your damages, right? If you prove trade dress infringement, what did you lose? You can get the infringer's profits from the product and costs. One really interesting thing about the Lanham Act, it points out specifically that it's supposed to be in the nature of compensation and not a penalty, right? That's what we're doing so we give the judge of discretion, potentially trouble damages, circuits have different rules, but, again, it's compensation not a penalty. And one part that's really interesting to our topic today is it specifically says that when we're talking about profits, the plaintiff, quote, shall be required to prove the defendant's sales only. And then the burden shifts to the defendant to prove any offsets, things that are related. So this like almost too good to be true, right? I remember having this conversation with a partner when I was in practice, and he's like, "Related's all we have to do?" It's like, yes, this is what statute says. Also the Seventh Circuit says that's what the statute says. So it's kind of great, at least in comparison to design patents. So under patent law, design patents are one of three types of patents we have, and design patent owners are entitled to all this sort of, quote, unquote, regular profits, sort of injunctions, 289 damages, damages adequate to compensate for infringement but not less than reasonable royalty, you guys are probably familiar with this. So they can get that, but we've also got this special remedy that says you can in the alternative get the total profits from whatever article the defendant applied the patented design to. So what does that mean? Okay, the air play between these sections is pretty important. You have to pick one or the other, you can't have both. The Federal Circuit has said you're entitled to whichever one is bigger, and if you want, you can ask both questions to the jury, and you're entitled to the one that's bigger. That decision actually got vacated, but I'm pretty sure that it's gonna get reinstated when they get another chance. And we've got some questions about this total profit remedy, right? We know for sure that, oh, I forgot one part. So it's the total profit or a minimum statutory penalty of, wait for it, $250. That amount has not changed since 1887, if you were curious. You do have to mark your patents to get this special statutory remedy, but $250 is per patent, not per infringed article. So if I sell a million infringing sculpture, I'm only liable for $250. The profits are post-tax not pre-tax, we've ligated that, but what we don't know, I just told you that you get the profits for the article, but what's the article? Take a case like Apple v. Samsung where one of the design patents covered just the screen of the smartphone. Has the design for the screen been applied to just the screen or to the whole smartphone? The Federal Circuit said, well, the article's whatever the defendant sells. Period, full stop. I think a lot of people assume this was the rule pre Apple v. Samsung, but no court had actually ever said that, and a lot of people in patent were going, yeah, that's obvious that's what things say, and the Supreme Court was not impressed. They got up there and I actually went for the arguments, and it was amazing to just watch the justice's faces, I really wish there had been cameras 'cause basically the overwhelming impression I saw was you gotta be kidding me, right? If you make a patent for one little tiny part of the phone, you get all the profits for the phone? That's insane! And so they said, yeah, we don't know. The article could be the phone or could be the screen, but we're not gonna tell you which and we're not gonna come up with a test, so lower courts have fun with that. That was a couple years ago. We're still kind of wrestling with the aftermath. We've had a couple cases, some of which, shockingly, have acted as if Apple v. Samsung never existed. They gave old model jury instructions from like 2010. Please don't do that, if there's District Court clerks or judges in the audience. So we don't really know. There's one case right now before the Federal Circuit where this issue might come up, but I think it probably won't. So we've got a lot of questions. So I don't even know how we prove what profits are if we don't even know what we're proving, is a really big problem. So I'm sorry that I have more questions than answer for you today, but I look forward to the discussion.
- John, let's turn to you. Your issue is more about how to prove a reasonable royalty. What kind of evidence can we recruit to prove a reasonable royalty for a patent infringement case? Can you talk a little bit about what the current state of the law is and where you see the possibility for some development?
- Yes, unfortunately in the utility patent space we don't get all the profits that we do in the design patent case, so I feel a little bit like mine is a second-class citizen, but the way it works in utility patents and a reasonable royalty, which is a tradition approach, it's a very simple equation. It's a royalty rate times a royalty base. So whatever the rate is, 1%, times this price of the product. So it sounds pretty simple, but for the last 10 years the Federal Circuit has been struggling with this, and changing the test and changing the direction. It started back in 2009 with a case that I was involved in, Lucent versus Gateway, Microsoft, and Dell, and went up to the Federal Circuit. In the Federal Circuit, the damages theory was a half a percent on the price of Microsoft Outlook, but the patent was only on the feature of Outlook. And so when the damages model went up to the Federal Circuit, they reversed because we applied the royalty to the full price of Outlook. But they said in the decision, and this is back in 2009, there's nothing inherently wrong with using the market value of the entire product in the reasonable royalty analysis, especially when there's no established market value for the infringing component or feature. So there's no market value for the piece that you've patented so long as the rate is adjusted to compensate for the smallness of the feature. So it's interesting language because they almost immediately went away from that in several decisions. Two years later in Uniloc versus Microsoft, they said that the court's precedence, quote, "do not allow for consideration of the entire "market value of the accused product." So almost immediately changing what they said in Lucent versus Gateway. Then a year later in LaserDynamics versus Quanta, they made it even more rigorous. They said that you have to apply the royalty rate to what they called the smallest patent-practicing saleable unit. So if you have multiple versions of the product, you have to find the thing that you sell that is the smallest that still contains the patented feature, and you've gotta apply the royalty to that. But then they drilled down on that even further just two years later in VirnetX versus Cisco. It was actually Apple was the defendant in the case, Cisco was a co-defendant, so the title of the case is sort of a misnomers. It's really VirnetX versus Apple. But there they want us to drill down even further than the smallest saleable unit. So the language the Federal Circuit gave us in 2014, they say, where the smallest saleable unit is in fact a multi-component product, so in this case it was an iPhone, that has several non-infringing features with no relation to the patented feature, the patentee must do more than to estimate what is the value of the patented piece of the product. And for anyone who does real world licensing, this actually makes no sense at all because licensees are not done that way. In fact, the court says later in the decision, in reaching this conclusion, we are cognizant of the difficulty that patentees may face in assigning value to a feature that may have never been sold individually. But I don't think they really understood how complicated that is. So what they did by these series of cases, they've made patent trials really complicated. So now instead of having a licensing expert as a witness, if you have a multi-component product, you have to get an economist, and the economist has to figure out what is the value of the feature that's patented in this broader product even if that feature's never sold separately. And the economist can't do that alone, they usually need a survey expert. So now we went from simple patent trials with a licensing expert where real world licenses were the best evidence to now an economist and a survey expert essentially making up values. And for anyone who tries cases, you know that if I have an economist and a survey expert, they're gonna come out one way and the other side's gonna have an economist and a survey expert that's gonna come out dramatically different, and this is all sort of amorphous and mushy, trying to get at a value, and I think it does a disservice to the sort of real world, what's happening in a licensing negotiation. When you have a licensing negotiation in the real world, the patent owner gets a royalty on a product and they may adjust the size of that royalty, make it really small if the patented feature's only a piece of the product, but they don't start in those negotiations assessing the value of individual components. I think at this point it's sort of the highlight of crazy in 2014 when the court came out with this decision, but then they almost immediately pulled away from it. And in the same year, in 2014, they came out with Ericsson versus D-Link where they said, sort of backtracking to 2009 with the Lucent decision, they said it's not that an appropriately apportioned royalty work can never be fashioned by using the entire market value. So just the opposite of what they said in the preceding three years for a multi-component product. But you can do so if you reduce the royalty rate applied in those cases. And they said they noticed the real world impact they were having 'cause they went on to say, otherwise, would you be making the real world licensing evidence irrelevant and not comparable, which is what was happening in the cases. And then they followed up Ericsson/D-Link with CSIRO versus Cisco in 2016, essentially saying the same thing that maybe we can do this apportionment by apportioning the rate as opposed to the base or the price of the product. Otherwise we would, and they say, "necessitate the exclusion "of all comparable license evaluations," which is essentially what was happening. And then they followed it up again, most recently 2018, with Exmark v. Briggs & Stratton where essentially they held exactly the same thing. So the onus in the law is there are a series of cases from 2009 to 2014 that tell us in definitive language we need to apportion the base and then there are series of cases from 2014 to 2018 saying, but it's okay to do it on the royalty as well. So the litigants are now in a quandary, but it's not like this second line of cases of apportioning the royalty is fail safe. One of the things that we noticed in actually trying the cases, plaintiffs can actually play games now under this new rubric. And in one of the same series of cases that I mentioned earlier, VirnetX versus Cisco and Apple, what VirnetX did in those cases is a symptom of what happens now in the real world because of this new line of cases. They had sued Apple and half a dozen other companies, and Apple was the big target. So prior to the trial against Apple, VirnetX settled with everybody else with sweetheart settlement deals, and these settlements were big royalty rates, but they were applied to the entirety of the switch products that each of the half a dozen defendants were selling, but then they only made the licenses apply to one model of the switch. So they didn't cover all the infringing sales, so ultimately those six licenses yielded like, one of them yielded like $50,000, one of them yielded $100,000. But when you apply that royalty rate to Apple sales of the iPhone, you're talking about $500 million in damages. And the court let those licenses in, and it shows you that even though I think that second line of cases is more real world, you have to be very careful about applying that logic and letting settlement agreements in or letting licenses that were litigation-inspired as opposed to real world. The point of that second line of cases is we're supposed to be capturing real world licenses and real market rates, not litigation-derived or litigation-inspired licenses. So that's sort of the state of the law. It's a little bit messy but litigations, it makes litigation fun.
- Great, so we're gonna have some time to pull these subjects together, I hope, in a few minutes, but, Kristelia, I wanted to give you a chance to talk a bit about both statutories and this new willing buyer/willing seller standard, which I think is, at least to me, very interesting how we're gonna do this.
- Sure, so, yeah. Apologies in advance for introducing two additional topics into the mix, but, yeah, I think there's some common threads. I'll start with talking about the current litigation landscape for statutory damages and copyright. Section 504 of the Copyright Act sets the rate for statutory damages, for copyright infringement, at somewhere between $750 to $30,000 per work, bumping that up to $150,000 if you can add a finding of willfulness, right? So the point just being there's this really dramatic range in possible statutory damages, and as you can imagine, this has resulted in strong feelings of pro and con statutory damages, perhaps depending on who your client is at the moment or what you're arguing. There are arguments that these are necessary in order to deter infringement, which is often very difficult to detect and certainly costly to detect, and then, of course, others who are concerned that these potentially very large statutory damage awards can induce rights holders to bring borderline claims, to bring frivolous claims, in hopes of just scaring these defendants into settlement, right? And the shift from analog to digital has only exacerbated these concerns. One of our copyright colleagues, Ben Depoorter, has an interesting new empirical piece that actually looks specifically at statutory damage claims and results in litigation in the copyright context. So I highly recommend that to you, but the primary takeaway, if I can ruin it, is that 80% of copyright infringement plaintiffs will claim willful infringement, but courts will only find willful infringement, right, the element that will bump up to $150,000 per work, in 2% of cases. Right, so 80% of plaintiffs bring willfulness, but the courts only find willfulness in 2% of cases. Why is it so low? I'm happy to talk more in the Q and A, but perhaps we might say, well, the meritorious cases settle, right, because people know that is what it is, perhaps many of the willfulness claims are made in areas of thin copyright, like fashion, right, where infringement is harder to detect ex-ante and so willfulness is gonna be harder to find, right? There could be some other explanations. Whatever the case is, yes, there does appear to be a overclaiming induced by the possibility of a really sizable statutory damage award, and this is undesirable however you feel about it because not only of the potential chilling effects but also a waste of court time and resources. So what do we do about it? Well, in his piece, Depoorter proposes, and I endorse a couple of possibilities that we could pursue for addressing this concern, we could aim to deter plaintiffs by making remedy overclaiming more risky and costly for them, perhaps by an amendment to Section 505. It would remove fee shifting in cases where you're found to have overclaimed. We might also reduce these super large outlier awards, the ones that intimidate defendants into settling, by adding some limitations on the availability or the size of statutory damage awards, right? So those are some possibilities, but that's where we're currently standing with statutory damages. I will segue into that something that's a little less damages and a little more costs, but it's really interesting and really timely, and that is the move that we're going to see with the new Music Modernization Act from what we call a 801b rate setting standard to a willing buyer/willing seller rate setting standard. As many of your know, the Copyright Act has several statutory licenses under which copyrighted works can be used, and these licenses traditionally utilize one of two standards for setting those statutory rates. There's one standard comes from section 801b1 and the other is something that we call the willing buyer/willing seller standard. The 801b standard is the more, well, I guess they're both quite subjective, but the 801b standard considers a series of four factors when it's setting a rate. It thinks about the maximizing the availability of creative works to the public, fair return to copyright owner and copyright user, relative technological contributions, capital investment to risks, and the minimization of destructive impact on the industries involved. So the courts go through these four factors, come up with rates, the CRB. By contrast, the willing buyer/willing seller standard is intended to emulate fair market value by saying that what it's gonna do is consider what a willing buyer and a willing seller would agree to pay in some hypothetical marketplace. Now the mechanical license that Chris mentioned that the MMA is going to be modifying, it has historically been set under the 801b standard, this four-factor standard. The Music Modernization Act that was passed this past October will change this in January 2021 to a willing buyer/willing seller standard. The import of this, potentially, or has some people concerned, is that traditionally royalties rate set under willing buyer/willing seller standard are substantially higher than the rates that are set under an 801b standard, so that a shift from an 801b standard to a willing buyer/willing selling standard will almost inevitably lead to a higher rate. And when I say higher, I mean a lot higher, right? So for example, digital performance royalties for internet radio providers like Pandora are set under willing buyer/willing standard, and there's paying around 50% of revenues compared to those same royalties but for satellite cable audio providers like SiriusXM who under an 801b standard pay out at 8%, right? So this is almost sure to make a difference, and some worry that the MMA's shift to willing buyer/willing seller will just be a coup for licensors. This assumes, of course, that the current 801b1 rate is the right one, that it's not set too high or set too low, which I think assumes a lot, and, in fact, my concern with the MMA's move from 801b to willing buyer/willing seller has little to do with the potential increase in rate, which for me is neither here nor there, and everything to do with the willing buyer/willing seller standard's willy nilly use of the term market value, particularly in the copyright context, where in many cases the establishment of the market and the establishment of the statutory license, such as in the instance of digital streams, were simultaneous, so that we've never had a market and we're not going to unless we deactivate these statutory licenses or remove things out from under them, meaning that, by definition, no market exists and we don't know the market rate would be. We don't know what a willing buyer and a willing seller would do because they've never had a marketing which to do. In addition, given the way that copyright has trended, we have so few players on either side. It's very difficult to reach any simulation of a negotiation in these bilateral markets, and for me this makes the willing buyer/willing seller standard not potentially unfair but highly subjective. And, again, in the Q and A we can talk about a couple of examples where people have, parties have attempted to manipulate what they'll present as these values. And so here I worry about the same thing that I tend to worry about in all of these cases of one-size-fits-all compulsory licenses, that they're untailored, meaning that they're almost necessarily going to undercompensate and overcompensate in different cases, and as such, they effectively will service penalty defaults, which is fine perhaps for those that can contract around them, but it can really be punishing for those who are not in a position to do so. So I'll stop there and let our discussion kick off.
- So first question, and I wonder what you guys think of this collectively as a panel, we just heard a lot of diversity and how we treat damages among these three different areas of law, which, of course, are closely related. They're all orbiting around the same question of innovation and protecting innovation, but the damages measures are really different. Is that a problem? Is there something we can point to and explain, well, this is why copyright is so different from utility patent, and this is why the question about what portion of the product you get in utility patent is so different from design patent, such that we can all kind of go like this and be satisfied, or is there a conceptual problem?
- I think the biggest problem for utility patents is we have gotten away from granting patents to substantial inventions and now we give patents for every little doohickey on a product, and it necessarily devalues what the patents are worth, and I think that's why the Federal Circuit is struggling so much with, what do you do with all these utility patents if you have an iPhone or a big switch and there's 10,000 patents on there, or 30,000 patents on there? What are you gonna do? And the inherent unfairness of the way try cases to jury is the judges often won't let you tell the jury that there are 30,000 patents and prove to the jury that the patent is just one little thing, so the jury get this skewed view, so I think, but I think that trade dress, or design patents, and copyrights are much different, in my view.
- What's the difference?
- Copyright I think generally is really what the product is about. It's more all consuming, and I think the same if it's a real design patent where it really is the design of the product that the consumer's after. It seems more all encompassing of the product, so you can understand why you'd have a damages rubric here where you're gonna give over all the profits because you stole the look and feel that the customer wants.
- So let me as Kristelia and Sarah whether they think the doohickies problem, as I'll frame it, afflicts--
- I'm the only non-professor up here, so we talk about doohickies.
- So, first of all, doohickey is a great word, we should use it more often, and I do think the doohickey problem is something I wanna focus on across these fields, and I wonder to what extent does the doohickey problem crop up?
- So I think it's extensive in product design. In trade dress we've definitely seen a movement. If you read the old cases, they say the trade dress is the total look and feel of the whole product, the colors, and now people claim, Mark and he wrote a great paper about this--
- Thank you.
- You can claim doohickey, or doohickeyization of trade dress law, right? Like we've seen this. It's certainly happened in design patents, right? This total profit remedy made a lot of, well, I'm not gonna say a lot of sense, more sense in 1887 where you had to claim the whole shape or the whole pattern of an article, but since 1980, you can do the doohickey thing in design patents too. You know, Apple's were not the worst I've ever seen, at least the ones that started in this litigation, but you can claim ridiculous little, tiny bits of things, and I think that was one thing that just, like you said, the inherent unfairness just doesn't pass the sort of what my tort professor used to call the brown paper bag test, right? Can I go into court and make this argument without a paper bag over my head? You know, just dying of embarrassment. So I think we see it there too.
- Yeah, so I'm thinking in theory copyright is intended to have some protection against sort of doohickeyization, but in the sense that like we don't allow you to copyright short phrases, right? If you have short, popular phrase, you go to trademark for that kind of thing. We don't let you copyright band names and song titles, things of that nature, right? So that said, there certainly are, if you write a poem and it's one sentence long, you can still copyright it assuming it meets the threshold for it's been fixed and it some modicum of creativity and so forth. So, yes and no. Where it comes up? I mean, potentially I see that coming up a little bit in arguments around compilations and sort of what value that is of putting a bunch of pre-existing copyrighted things together sort of. I think the doohickey as it happens is more contribution, sort of what you added when you put a bunch of things together as sort of the, more the place where I see it.
- Yeah, and I wouldn't necessarily put this in a basket of doohickeyization, but I do think there is an element in music of this. So think about what goes into a stream on, say, Apple Music, or Spotify, or Amazon Prime. So one thing is the musical composition, which might be owned by one or more copyright owners. Another thing is the sound recording of that musical composition, so the actual recorded performance of that musical composition, which may be owned, and often is, by a completely different copyright owner or set of copyright owners. A third thing is the service, right, that is providing the stream. So they provide playlists, they provide technological infrastructure, they provide marketing, they provide the licensing infrastructure, right, that pulls all the stuff together. So in the 115 proceedings before the Copyright Royalty Board which sets the rate, the last time they did it before the MMA they ran this incredibly complex game theoretic model called Shapley Values where they had these economists trying to measure the relative contribution to the success of the song, the stream, of the composition, the sound recording, and the contributions of the streaming services. I think the copyright in the composition came out to below 10% of the value.
- It was low.
- It's close to a doohickey. Not quite, but it's flirting with the edge of doohickeyization, so--
- Did you see that in the utility patent context when you're talking about patent pools or industry standards? When you set up a patent pool where everybody puts their patents in and there is ordinarily an algorithm that assesses the value of each, and there's a formula for who gets what, that is a much more realistic valueization of what these individual utility patents are worth, but we don't do that in District Court litigation, and very often in a District Court litigation you're not allowed to prove all the patents that are surrounding, and so the jury gets a very skewed view, but I think the patent pool context, it works much like what you were just describing.
- So can I ask a question about something, Sarah, that you said that really stuck me, which is the Supreme Court says the Federal Circuit has it all wrong when they say you can get the entire profits from the product, there must be some partitioning, and then basically says, you deal with it, right? So we're not gonna deal with this. So, okay, you deal with it. Could be us dealing with it. So what should the test be?
- Yeah, so, I mean, they didn't exactly, they said the rule that the relevant article manufacture always has to be the end product was wrong. That's literally all they said. Some people have overread that decision to say that in cases like Apple with the screen versus phone that the plaintiff has a right to always put that in front of the jury and I gotta be able to argue for the phone, I don't think that's right. That's not what the Supreme Court said. They just said it's not automatically the phone. So what do we do? My argument that I have made in the past, and I still continue to believe is not insane despite criticism to the contrary, is that we need to figure out, first of all, what's an article manufacture, right? The Supreme Court says it's anything made by people. Okay, that's fine. I would go back and adopt a slightly more specific historic definition. When this remedy got passed, the practice, the law was that we were talking about tangible items that are made by people, that have a unitary structure, and are complete in themselves for use or for sale. Okay, what does that mean, right? Something that's made in a whole piece, it exists in a whole piece at some moment in time. So like the phone screen, right? That exists as a physical piece. It is later combined with other articles of manufacture to form a larger composite article. So what do we do? I don't think this is rocket science, right? The SC got up and they kind of wanted to have this argument about what was important and like the qualitative value cut at the heart of the design or something. I think that's just making it too hard, right? Like let's look at what you claimed. If you claimed a design that sets the shape of a screen, and a screen is an article of manufacture, we're done. I mean, I just don't think it has to be complicated than that.
- So the more difficult question would be if you claim an article that is the arrangement of buttons on a screen, or the arrangement of microphone ports on a screen, which Apple actually did, and that's not an article of manufacture but the screen is an article of manufacture, right now we have turtles all the way down. So what do we do we that?
- Well, first of all we need to get rid of GUI design patents which is what I think you're alluding to.
- I am.
- So Apple also had this patent for one screen shot from a graphical user interface, or GUI. The patent office has been granting those for since the '90s, so quite a long time, but no court has ever said, yes, this actually statutory subject matter. I think I we're gonna protect them, what we have have to do, courts dealing with this now, the patent office says the article of manufacture is the screen, fine, take the profits for the screen. I don't think that's right, I think we should get rid of 'em, but in the interim, I think that's the most reasonable thing to do.
- Okay, so, John, just wanted to get back to this question of the royalty rate versus the royalty base. Okay, so right now we have deals out there that are really focused on the royalty rate when there's some feature or product that is patented and there is some question about the value of an infringement claim on that feature. The Federal Circuit, as you put it, has been kind of all over the map on whether you have to focus on the royalty base or whether you focus on the royalty rate. One thing that contracts people would say, people who do a lot of contracting work, is that just give us a clear rule and contracts will bend around the rule, right? So if you tell us we have to focus on the rate, we'll focus on the rate. If you tell us we have to focus on the base, we'll focus on the base. But tell us something clear. If you were to have the court articulate a clear rule, I get the sense that you want to focus on the rate. And why do you think that's better now?
- I think ultimately it would be better now, and it would have been better all along, if we focused on that because I think that, especially in jury trials, I think the best evidence is what happens in the real world. And if we focus on apportioning the rate, the way actual license negotiations work is the patent owner is sitting across the table from the alleged infringer and they negotiate a license agreement. And every license agreement I've ever seen, and I've done a lot of licensing, is if it's a running royalty license agreement, it's a rate applied to a product that you can account for. So it's a product that's reported in the annual or quarter, in a 10-K, or in some sort of filing, so that the patent owner knows how many there are and can easily account for how many there are, and there doesn't have to be some sort of detailed accounting. So that's what people do in the real world, and so that rubric sets great real world evidence on what the value of the patents are. In a jury trial, that then becomes the ceiling on what you're gonna get because that's what happened in the real world, and it's great evidence of the value of the technology at issue. Whereas in the apportionment of the base world that the Federal Circuit brought us to, which culminated in the VirnetX-Cisco case, what we prove at the trial has no relationship to the real world. We put on an economist who does a conjoint analysis, or hedonic study, or relies on survey evidence about whether a consumer finds a feature valuable, I'm not even sure a survey about whether a consumer finds a feature valuable is really valuing the feature in the way that we're supposed to be getting after. So you're gonna put on a economist, you're gonna put on a survey. I think it's suspect valuation methodology, and the other side has opposite economists and survey experts. So it seems to me a world of make pretend, not a world getting at the actual real world.
- Let me ask you to assess the size of one problem that you mentioned, which is, okay, so you get the example of the switches. So there's a bit of chicanery going on where the rate looks really high, but if you look at what's actually happened, they've restricted it to a certain less popular switch.
- Yes, exactly.
- Okay, so that's an abuse, essentially. That's an attempt to introduce this rate and use it across the board for something that really isn't across the board.
- Right.
- So here's another trick that they could use. There's a flow of payments between Company A and Company B, and the rate on it for certain patent license is artificially high because there's a side deal. There's a flow of payment regarding something else, and that flow of payments is made slightly smaller to compensate for the slightly larger royalty rate, and why? Because there's a side payment to get the rate higher so that we can now introduce some litigation and say, here's the rate across the base. So that's also a possible abuse. How confident are you that those abuses can be ferreted out, exposed for what they are and avoided?
- Well, it will be a jury trial, so I think if you have an experienced licensing expert, they will do an analysis of the license agreements, all of them with all of their, they all have different features. So some of them are portfolio license, some of them are cross licenses, some of them have extra value. They might be like a manufacturing or agreement of something else that goes along with it. And I think the licensing experts would value each of those features and would give an opinion then on what the value of the patents in the litigation are, but at least it's anchored in something that's real and is real world, and I think it has a better chance of getting to something reasonable, and at least the licenses then fix the universe of the most it's gonna be or the least it's gonna be. Whereas when you're operating in the VirnetX versus Cisco world where the economists essentially have free reign to opine on things, like, for instance, in this one case, Summit Six, which was a fascinating departure from real world, Summit Six put forward an analysis where they had an economist and a survey expert value the infringing feature in the defendants product, and then value through survey evidence and annual reports, and all this sort of suspect evidence, what it costs to manufacture that, and then did an analysis of, okay, with that cost to manufacture, this amount of profit from the sale of the product goes to this feature and a willing licensor and a willing licensee would look at that profit and say, that's a trivial patent, let's split it. And, you know, that's not how the real world works, and that number that they came up with was a large number and wasn't tethered or anchored to anything real world, which is why I get troubled with that sort of analysis.
- I think I agree that the rate's the way to go, but this whole discussion suggest to me that lawyers better ask the right questions in discovery to get all the other sources of value that may affect what that royalty rate is set at, and that would, I think, make the analysis much more robust.
- But even more so, we have to get the judiciary to let the litigants actually prove the other factors because what happens in most of the cases, the judges will say, this case is about these patents, so I don't wanna hear about any other patents, and they corral the evidence you can prove and what that does is it doesn't allow you to tell the jury, this product has 10,000 patents on it. Hear what they are, here's how this one, the patent in suit, compares to these others. And that real world evidence is often excluded, and so the jury gets a skewed view of the real value.
- So I wanna open this up to questions, but one last question, Kristelia, for you, which is about this enormous range of statutory damages in a copyright law. So the copyright section's pretty bare on this. So basically it just says here's a range. The range goes from 750 to 30,000 bucks per work infringe, for non-willful infringement. The top of the range gets bumped up to 150,000 bucks max per work infringe, for willful infringement. Willful is defined a little bit differently by courts, but it all focuses on the same thing, either knowledge that what you're doing is copyright infringement or reckless indifference with regard to that fact. So it's a pretty high level of scienter required to get you into the willfulness category, which may also help explain why courts so rarely push defendants into the willfulness category. Okay, so the statute says a little bit and then stops, and that's often what the Copyright Act does. A Copyright Act says a little bit and then stops, okay. And in the open spaces, courts tend to fill in the blanks, right? So the Copyright Act says nothing, for example, about what the infringement standard is, and the courts have filled in the blanks. Very helpfully they've provided us with all kinds of rules that were spoken about earlier about how, like in the Rentmeester case, how infringement analysis is supposed to work. Okay, there has not, to my observation, in the statutory damages area been a development of common law rules about how this range is supposed to be approached. Do you have any views as to why not, and if we were to start thinking about that, if judges were to start making these blank spaces a little bit more filled with stuff, what would you wanna put in there?
- Well, one thing that I've always found frustrating from both sides in litigation, and I think other people would agree, or one of the things that people often find most surprising, is that you can go for statutory damages and because we don't have these well developed sort of common law suggestions that these are the things that you should do to tell us where on the range of 750 to $30,000 you fall, you don't have to, for example, you're suing for someone using your photograph, present any evidence to show what you would have licensed your photograph for, right? Like that's not, that would be a basic starting point. I would like to see, like, okay, I used your photograph and I wasn't supposed to, well, if I'd approached you properly for a license, what would that range have looked like, right? But we don't even bring those in and there are a variety of reasons for that, right? Well, maybe I would flex my licensing right depending on who you were, and that sort of, what the use was gonna be and these sorts of things. But maybe something that just allowed us to show what the market looked like, what the market for your particular work looks like, because what one photographer might license their photos for can vary dramatically to what another photographer might license their rights for, and in within different context. And so I think requiring some sort of, you know, and this is why we say, well, statutory damages also doesn't, you don't require any showing of harm, right? Okay, maybe fine, but showing of potential loss would even be, your loss profits of some sort, I think would be a good start in that direction.
- Yeah, so imagine this as a methodology. So the photographer brings a claim and there's some evidence that suggest that the licensing rate would have been about 1,000 bucks, right? And the judges says something like, well, it's unlikely that, pretty unlikely, that this infringement would have been detected. So it was kind of a bitter miracle, so maybe there's a 10% chance that the individual would have been detected. Okay, so I'm thinking if the cost of license was 1,000, I wanna give the photographer at least 10 grand, plus I wanna give the photographer the cost of bringing the lawsuit. So I wanna give attorney's fees and costs. There I think I have some, at least a math that I can point to that says now I've made it not pay for this defendant to have infringed in the first place, right? I haven't--
- Sure, 'cause you don't want the defendant making the calculation of like, do I spend the grand, or do I take the 90% chance I won't get caught? Well, take the 90% chance.
- Right, so now it looks like, it looks like I've deterred them, right? I haven't punished them, but I've at least deterred them. And there some law that says copyright damages aren't suppose to be about punishment, they're supposed to be about compensation and deterrence, right? So I've achieved that. Is there a way to get that into the law?
- [Woman in Audience] Yes.
- Yes, you can write it in, right? Like you could get 504, we could have some guidelines there, or we could have courts, and I've argued this in different context, take a more purposive, substantive approach, and say like, here's what we are intending to do with statutory damages, and have a line of cases that we can look to that start to look at those kinds of things.
- I'll tell you, though, in this country, it's not gonna happen because the District Courts in this country do not like assessing fees and shifting fees. So I don't think, you know, I do think it's unfortunate because it creates this what we call in the utility patent world the efficient infringer who makes an assessment of if at the end of the day unjustly being ordered to pay a royalty, why would I do that up front? I'll just wait till the end of the day, and maybe the end of the day will never come. And so even in those scenarios, District Courts are loathe to shift fees, which I find surprising, but they won't do it. They have this American rule stuck in their head and they won't come off of it.
- Yeah, so on that observation, I'd like to open it up to questions.
- [Woman in Audience] So the statute actually says, the copyright statute says that the award of statutory damages is supposed to be just, and I think that's actually a place where you could start building a jurisprudence. And forgive me for the self-promotion, but my article in the William & Mary Law Review basically set out a set of ways in which you could develop and use cases that some courts have actually done to say in this kind of situation, this is the kind of award, an approximation of actual damages. In this kind of situation where somebody knew that they should have gotten it, maybe you sort of do a multiplier over the license fee, and it's not hard to do. It's just that you actually have to try and once some courts actually got started. What worries me about this new small claims bill that's before Congress is creating an entirely new statutory damage regime without any kind of limits so that somebody who has a $500 license fee can go in and ask for $15,000 and the Copyright Office is not contributing to the development or the confidence that they would actually use some just principles to try to do this. So I'm against the creation of any new statutory damage regime, given that we basically don't have a good one in copyright right now from my standpoint. I will also say Sarah Burstein's articles on the article manufacture issue in the design patent context are brilliant. You really need to read them, they're really super special, and I will also announce I have a project to try to look at the disgorgement award across intellectual property regimes. There's not the same rule in any of them, which is like really, really weird. So we're gonna try to come up with ideas about if you were going to try to justify having disgorgement here but not here, what would that look like? So anybody who has interesting ideas about that, please contact me.
- Yes.
- [Man in Audience] So one of the things that's always interesting litigating cases is when you first meet the judge. The judge always wants to know what's the case about, but they never ask what's it worth. And I think part of the distortion, particularly in the patent context, is because damages gets pushed off way towards the end of discovery, experts get involved, and by the time the judge is really think about damages, he or she have already decided that it's gonna go to trial, and they're thinking about in the context of what evidence should I let the jury hear, and what can I do to make the jury's job easier. Now I remember a few year ago, I think Magistrate Judge Laporte tried to introduce the idea of early damages contentions kind of in line with infringement contentions and patent cases, and I just wondered what the panel's thoughts were whether, it doesn't really matter if it's utility patents or design patents, about that idea and why hasn't it taken off because it seems to make a lot of good sense and really be a filter for addressing complicated damages questions when there's time to do it as opposed to right before trial.
- Yeah, I mean, I'm happy to start. I think that that would be a good development in civil litigation. I mean, one of the problems with utility patent litigation is that it's outrageously expensive and you could easily spend $10 million on a medium-sized case or 15 or $20 million on a big case. And often times because of what you suggested, the District Courts are treating cases all the same, and if we did have at the first conference a what-is-this-case-worth conversation, you could put small cases on one track, and medium cases on a different track, and the big cases that deserve a $15 million fee in their own track, and you could streamline discovery, you could make it affordable for the smaller disputes, but it doesn't happen. And I've heard those proposals too, but it never takes off, and I think one of the reasons it might not take off is because plaintiffs are always loathe to put their cards on the table early. So maybe that's why, and it takes a District Court judge willing to do that too, but it would be a sensible reform, I think.
- I've never heard of that in a copyright case either. Have you?
- Mm-mm.
- Everyone just says it's $150,000, right?
- Yeah.
- Yes.
- Strom?
- Sorry to interrupt.
- [Strom] Yeah, so I'd like to invite your reactions to a couple of related points that I heard during the talk. One is the doohickey problem, right? If I understand you correctly, John--
- I got a copyright on that. Okay, consider it done.
- Trademark sort of thing.
- Trademark.
- Trademark, damn it.
- [Strom] So one way to look at it is that--
- Still good.
- The compartmentalization and incrementalization of IP rights, whether patents or otherwise, is bad in and of itself because it devalues the whole enterprise. We should award IP right only for things that are really sort of substantial innovations. The other, which I thought I, it was at least implicit in Sarah's comments, but maybe she didn't mean it that way, is that it's only a problem if and to the extent that remedies are disproportionate. So if you're claiming remedies on the whole thing, then the design better damn well cover the whole thing. And if not, then that's the problem, not the inherent doohickeyization. And it may be that this problem is aggravated in litigation because we tend to focus on individual patents rather than portfolios of patents. Even if a portfolio is asserted, certainly as it would be licensed, cross license might be portfolio-wide, courts would say, well, we'll take five representative patents for judicial efficiency and economy, and take it that way. The related point is, John, your view on rates rather than the base. It seems to me that if the reservation price of the seller and the reservation price that the buyer's willing to pay overlap substantially, then any amount in between those two points, if there's overlap, is an economically defensible answer. Now the Nash bargaining equilibrium might be a 50/50 split if there are two counterparties, and that's a theoretical answer, that's fine, but as we're doing those two, any answer would be economically sort of reachable in the marketplace. So since all bargaining takes place in the shadow of litigation, why isn't the base the more important thing to focus on because the rate could literally be anything in that range, and we don't really have a principle way of getting to the right answer in all cases.
- So to pick up on the last point first, when I said that we should be apportioning the rate in the analysis, I didn't mean without reflection on the base because you need to look at the patent and understand what it is claiming, and how much of a portion of the base that affects, but then the math should be done on the rate. And the reason I think that is simply because the real world licenses that are gonna come into evidence do that, right? Because they apply a rate to the whole base. So I didn't mean that we would apply the rate regardless of the claimed invention, you have to look at the base and you have to see how big it is, but then you look at the real world licenses and you see how the people in the real world adjusted the rate 'cause the way the licensing works in my experience is you might have a .35 rate in a real world license, meanwhile in litigation, that patent's being, somebody's asking for 2%. But if you look at the real world license, it's .3 or something like that. But on the doohickey problem, if I were czar of the Patent Office, and I'm not and I never will be, but if I was, I would not be granting the patents on the things they're granting patents on. The number of individual tiny little things that we are giving people exclusive rights to seems like the wrong approach to me, and that's why we're coming up to 10 million patents. I mean, it does, that doesn't make sense to me and I don't think it's necessary, too, for innovation. I don't think a chip company is making a new transistor with a gate made out of a particular compound because they're gonna get a patent on it. They're doing it 'cause it's gonna make a better process. So I think that a lot of patents we're rewarding are creating a problem. But if we're gonna be in that world, I think the District Court judges need to let the parties prove that we're in that world, and we need to be showing the jury there are 10,000 patents on this device, let me explain them to you. This one that we're litigating is this teeny little thing, so don't get crazy with the damages. And we're not allowed to do that, at least in any litigation that I've been involved in. We're not allowed to do that.
- Ron?
- [Ron] I want to pursue doohickies and smallness. Why isn't it easy simply to see how much the doohickey costs in relation to the total cost of the item? And, say, well, if the doohickey's 2%, then you get 2% of the profits, or you get 2% of the cost. And if the plaintiff wants to prove that, no, my doohickey has 50% more value, then he's got a clear and convincing burden to show that because of his doohickey, the price of the machine is 50% more than it would be otherwise. And otherwise it's too speculative and he doesn't get damage. He has to prove that his doohickey has greater value. Why isn't that viable?
- That's what they did in Summit Six. What Summit Six was they did a manufacturing cost assessment of what it cost to make the part that was patented, and then they did the math like you're talking about, about, well, how much this thing sell for and what the incremental margin, and so we'll give this, this is the profit that's associated with that. I think that's, in my view, there's a flaw in that logic because I think just because a part costs 2% of the full cost is not telling you what that part is really worth in the marketplace, right?
- Well, then you give--
- What would somebody pay for that and what is the real value of the contribution
- You'd get the opportunity
- of that particular part.
- to prove that override.
- So I wanna get Andrew and then Dale in, and then I think, we have a reception. I don't want to keep everyone from that reception for too much longer. Andrew.
- Thanks, I wanted to follow up with a comment about statutory damages that Chris made, but then also ask Kristelia a question about statutory damages. My experience is that copyright holders who are plaintiffs in cases do and don't wanna say how much the case is worth. They actually will say, we're suing for 600,000 works at $150,000 each, but you can never get 'em to say, and that equals $19 billion dollars. They never give the result of the multiplication, but part of the reason they wanna be clear is that many of these cases are against private companies and they're gonna last for two to three years. Private company has a six billion, 10 billion, $50 billion contingent liability on the books, and someday wants to have a liquidity event, in which case they say to their lawyers, do whatever it takes to settle the case within a week, and that's how those cases operate. But a question for Kristelia, I run everything, all these principles we're debating, and appellate decision and the like. Through the lens jury instructions, has anybody ever seen a set of jury instructions for the Safe Harbor defense, under five, one of the 512? I mean, it would go like 15 minutes in court. It's complicated. But I'm concerned about statutory damages for contributory infringement because after Grokster, maybe all contributory infringement is willful, because it requires intentional, culpable misconduct. So does a finding, if you're trying to instruct a jury on the standard for contributory infringement, and you're trying to instruct a jury on the standard for finding willfulness for a higher range of statutory damages, are you supposed to say, well, if there's contributory infringement, ordinary contributory infringement, that is willful and intentional inducement, you're 750 to 30,000, and for like super, super willful infringement, you can bump it up, there's no principled distinction, I think, in the law right now that allows a court to explain coherently to a jury which level of statutory damages corresponds to what sort of degree of culpability and contributory infringement. Have you addressed that at all?
- Yeah, I think that right and that's part of what Chris was saying, is we don't have a well-endorsed and accepted definition of what willful is going to be. It's a little bit of like, I think if we're teaching it, there's this knowledge factor that you could potentially contribute to really infringe without intending to do this without the ish, right? On the books. So I would think, like, well, if I were teaching, I'd say, well, then the difference between that and willful would be this knowledge requirement. In practice the court will say I'll know it when I see it. You lay the facts out, or the jury will, you know, we'll give the jury the instructions, but I have not seen anything consistent to describe like when we know we're going from one to the other. The numbers that are in the new Depoorter piece suggests that maybe the juries don't see much of that either, or whatever that is that the courts are giving them, maybe they're saying some super duper thing. The juries are not finding any thing in the facts that kick it up to this super duper level, and that's also where we're getting really low percentages. I haven't seen anything.
- [Ron] Thank you.
- Thanks.
- [Dale] Hey, I just wanted to follow up on John's comment about not being able to present, hi, John--
- Hi, Dale, nice to see you.
- [Dale] To the jury, the thousand other patents that issue in making your damages presentation. I have two interrelated questions. One, have you or someone else ever tried to preserve that issue so that an appellate court has ruled on it? And the second part of the question is in copyright cases when we're litigating apportionment of damages, frequently just like in the patent field, we would use surveys to try to show, well, why did you buy this? And what's so important about this? And isn't that a way of, it doesn't mention the thousand patents, but doesn't that get it the same thing of assessing the relative value of the doohickey to the whole thing?
- So let me answer the first question first. So we have preserved the issue. We preserved it in two cases for Cisco, and then we won both of the trials, so it never got out. We never got it. We got non-infringement verdict, so I'm looking for one, but hopefully we'll stay on the winning streak and I'll never get there. But on the second point the, so in any event, I have not seen that issue appealed. I think it's a great issue and hopefully it'll make it up to the Federal Circuit one of these days. On the second issue, the trouble I have with surveys in the patent context, and maybe you guys in your context will be different, but, and this goes back to the question we had earlier, when you ask a consumer after the fact about a product and how important was this feature to you, I have two fundamental logic problems with that. One is they're making up that answer. They don't really know how important that feature was. But secondly, very rarely can you ask a question in the utility patent context that actually gets at the invention, right? So the invention is not often the screen, it's that you used a particular silicon in the screen. And so you're like, well, consumer, did you like that silicon? So you can't say, of course you're gonna say I like the screen, but it doesn't get at the invention. In the utility patent context, we're litigating patents that are super technical, and if you were to ask the consumer really about the invention, they wouldn't have an answer. So that's the trouble I have. I think it's different in trademarks and copyrights, but that's my problem with surveys, so but we're doing it anyway. And this is why these cases that apportion the base I think are all wrong because they're relying on consumer surveys that ask consumers after the fact, what'd you think of that feature? And we're using that as a proxy for the invention, and I think it's a disconnect.
- [Dale] Thank you.
- All right, thanks everybody, I really appreciate--