Ongoing Royalties and Willful Infringement

Hat tip to Professor Michael Risch, for noting on Twitter a recent decision of the District of Delaware, Vectura Ltd. v. GlaxoSmithKline LLC, Civil Action No. 16-638-RGA.  To cut to the chase, a jury found that the defendant willfully infringed claim 3 of U.S. Patent No. 8,303,991, directed to certain "composite active particles for use in a pharmaceutical composition."  In a decision resolving various post-trial motions, however, the district court (applying the Read v. Portec factors) denied the plaintiff's motion for enhanced damages and attorneys' fees, noting among other matters the closeness of the case.  (Plaintiff was seeking a 33.3% enhancement.)  The court also denied the plaintiff's request for an ongoing (post-judgment) royalty of 4% (as compared with the 3% the jury awarded for the pre-judgment infringement).  The court denied this request as well, reasoning as follows:

. . . to determine the ongoing royalty rate, courts have used the Georgia-Pacific factors.  "Generally, the jury's damages award is a starting point for evaluating ongoing royalties." Apple, Inc., 2014 WL 6687122, at *14 (citing Bard Peripheral Vascular, Inc. v. WL. Gore & Assocs., Inc. , 670 F.3d 1171, 1193 (Fed. Cir. 2012), vacated on other grounds, 467 F. App'x 747. I must consider (1) the "change in the parties' bargaining positions, and the resulting change in economic circumstances, resulting from the determination of liability," Amado v. Microsoft Corp. , 517 F.3d 1353, 1360 (Fed. Cir. 2008), (2) "changed economic circumstances, such as changes related to the market for the patented products," XY, LLC v. Trans Ova Genetics, 890 F.3d 1282, 1297 (Fed. Cir. 2018), and (3) any other "post-verdict factor" that would impact "what a hypothetical negotiation would look like after the prior infringement verdict." Id.

Here, the parties have not addressed the Georgia-Pacific factors in full but focus on specific factors supporting a departure from the jury's verdict. Plaintiff asserts that the following factors support increasing the royalty rate above the 3% rate determined by the jury: (1) the jury has determined the '991 patent is valid and infringed; (2) the accused products are more commercially successful now than they were previously; and (3) the jury has determined that Defendants' infringement is willful. (D.I. 339 at 21).

First, I do not believe that the changed circumstances from the jury verdict and increased sales would significantly impact the post-verdict hypothetical negotiation as compared to the hypothetical negotiation considered by the jury. While Plaintiff would be in a stronger bargaining position based upon the jury verdict,^/10  I do not believe that this stronger position would allow Plaintiff to negotiate an "above-market" rate for the '991 patent. At trial, both parties' damages experts referred to a  comparable license between the parties which provided a license to Plaintiffs entire respiratory portfolio at a royalty rate of 3% with an annual worldwide cap. (D.I. 353 at 449:1-450:7, 451:34-454:8; D.I. 355 at 845:5-15, 858:11-859:8). Additionally, Defendants currently license other patents from Plaintiff at a lower 1% rate. (D.I. 353 at 500: 11-24). Plaintiff appears to tacitly agree with the assessment that without enhancement of the ongoing royalty rate for willfulness, the appropriate ongoing royalty rate is 3%. (D.I. 339 at 22).

^/10  I say this because the Federal Circuit has said this, but l am unsure of the logic behind this view. The Federal Circuit's ongoing royalty opinions suggest that it is the resolution of any uncertainty regarding validity and infringement that strengthens the patentee's bargaining position. See Amado, 517 F.3d at 1362 ("Prior to judgment, liability for infringement, as well as the validity of the patent, is uncertain, and damages are determined in the context of that uncertainty. Once a judgment of validity and infringement has been entered, however, the calculus is markedly different because different economic factors are involved."); see also Active Video Networks, In c. v. Verizon Comm 'ns, Inc., 694 F.3d 1312, 1342 (Fed. Cir. 2012) ("[W]hile it is likely true that Verizon would not have agreed to that amount prior to litigation, Verizon has been adjudicated to infringe and the patent has been held not invalid after a substantial challenge by Verizon.").

Yet, in the original hypothetical negotiation, the parties are assumed to be willing licensors and licensees and the patent is considered valid and infringed. After the trial, there is a jury verdict that the patent is valid and infringed. Thus, the jury verdict has not changed anything in that regard. In both hypothetical negotiations, the parties want to reach an agreement. The only difference is the fact of the jury verdict, but since the patent is valid and infringed in either scenario, and the parties know this, it seems to me the pressure on the infringer to take a license is the same in either hypothetical negotiation (assuming any changes in economic circumstances do not impact the negotiation). See Mark A. Lemley, The Ongoing Confusion over Ongoing Royalties, 76 Mo. L. Rev. 695, 704-05(2011); cf. Apple Inc., 2014 WL 6687122, at *2 (citing Lemley).

Defendants assert that the ongoing royalty rate should be capped consistent with the previous license agreement between the parties. (D.I. 344 at 21). Defendants note that every license agreement between the parties included a cap on royalties. (Id. at 22 ( citing testimony from both parties' experts)). I disagree. The jury considered this same evidence and concluded that an ongoing royalty of 3%^/11 would have resulted in the hypothetical negotiation of July 2016. (D.I. 321 ). I see no reason why a post-verdict hypothetical negotiation between the parties would reinstate a cap on the royalty payments, especially in light of the increased sales of the accused products and the strengthened bargaining position of Plaintiff.

^{/11While the jury was not asked to explicitly assess whether or not a cap would be appropriate on the royalty, the jury did have the option to assess a lump-sum payment for the life of the patent. Had the jury felt a capped amount was more appropriate, it could have assessed the lump-sum payment based upon the royalty cap for the remainder of the patent term. The jury did not do so. (See D.I. 321).}

Second, I do not believe that the jury's willfulness verdict would affect a post-verdict hypothetical negotiation. Nor will I enhance the ongoing royalty as the Read factors do not support such an enhancement. Additionally, as at least one court has explained: 

If it is improper to use willfulness as a basis for enhancing the ongoing royalty in a case in which an injunction has been granted and stayed, it would seem at least equally improper to use willfulness as a basis to enhance the ongoing royalty in a situation in which the equities would not even permit the issuance of an injunction  in the first place. And it may be even more improper in a case such as this one, which involves a non-competitor, in which the plaintiff chooses not to seek an injunction, and in which the defendant's continuing infringement in fact benefits the plaintiff by generating sales from which ongoing royalties can be awarded.

Erfindergemeinschaft UroPep GbR v. Eli Lilly & Co., 2017 WL 3034655, at *9 (E.D. Tex. 2017) (discussing Amado v. Microsoft Corp., 517 F.3d 1353, 1356 (Fed. Cir. 2008)). I agree with the Eastern District of Texas that willfulness is not an appropriate ground to enhance an ongoing royalty rate where the plaintiff is a non-competitor who benefits from the defendant's ongoing infringement.

Thus, I will grant Plaintiffs motion for an ongoing royalty rate of 3% to compensate Plaintiff for Defendants' post-judgment U.S. sales of the accused products through the expiration of the ' 991 patent (pp. 12-14).

Although I generally agree with the court's analysis, I do offer a few caveats.

First:  I agree with Mark Lemley (and with Judge Andrews) that when a court denies an injunction economic logic suggests that the ongoing royalty rate should be the same as the prejudgment royalty rate.  Indeed, I've said the same thing myself on numerous occasions.  As a general matter, however, the prejudgment royalty rate itself should be somewhat higher than the rate the parties would have negotiated prior to the infringement, in the real world given that (as footnote 10 indicates) the parties to a real-world license would have discounted the rate based on their views concerning the probability of noninfringement and invalidity.  The reason we calculate reasonable royalties based on the counterfactual assumption that the parties bargained knowing the patent to be valid and infringed is to avoid a double discounting problem.  In addition, real-world licensees sometimes bear risks that infringers avoid.  For these reasons, reasonable royalties generally should exceed negotiated royalties by some amount, all else being equal (though when real-world royalties are negotiated against a backdrop of potential holdup, things get more complicated still).  For discussion of the preceding topics, see Thomas F. Cotter et al., Reasonable Royalties, in Patent Remedies and Complex Products: Toward a Global Consensus 6, 23, 34-41 (Brad Biddle et al. eds. 2019)  

Second, the analysis also is a bit more complicated when there is a finding of willful infringement.  If willful infringement means that the defendant was aware of the patent before it started infringing (or otherwise committed to any particular technological path), then it's a bit hard for me to see why an injunction is improper.  Functionally, it makes sense to deny injunctions when there is a substantial risk of holdup, but when the defendant was aware of the patent before infringing it's not so clear to me that that risk is present--in which case, the social benefits of injunctions (reducing the risk of error and administrative costs in determining ongoing royalties) might outweigh the social costs.   Further, under the eBay factors, the balance of hardships arguably favors the plaintiff, if the defendant knowingly brought its own hardship on itself.  So I'm not sure why an injunction shouldn't issue in such a case.  (I haven't researched the facts of Vectura, or read the E.D. Tex. case the court cites. Maybe the plaintiff didn't request an injunction here?  Maybe it expected the court would apply the eBay factors in a less substantive, more wooden, manner?  I don't know.)

Third, though, if the court doesn't enter an injunction, I still adhere to the view that the ongoing royalty should be at the same rate as the prejudgment royalty, assuming that the latter was correctly calculated at a somewhat higher rate than a negotiated royalty would have been--which is arguably the case here, since (as footnote 11 indicates) the jury didn't impose the equivalent of the royalty cap found in some of the relevant negotiated licenses. 

Fourth,  however--and notwithstanding everything I've said so far--if the court had awarded enhanced damages for the prejudgment infringement, but still denied an injunction, would it make any sense to award a non-enhanced ongoing royalty?  I don't know; probably not.  (This appears to Professor Lemley's view as well, in the Ongoing Royalties article at pp. 702-03, if I'm reading it correctly.)  For reasons already stated, though, I would hope such cases are rare.  

Update:  Professor Erik Hovenkamp makes an interesting point in response to my fourth point above--namely that, contrary to point four, it may be better to concentrate the amount of the enhancement in the prejudgment royalties and not the postjudgment royalties.  He writes:  "In particular, a royalty will raise costs and, by extension, prices, which is bad for consumers. By contrast, damages for pre-judgment infringement are a fixed cost. In principle you could have as strong a punitive effect as you want by just increasing the latter to a sufficient extent. Then you could get your deterrence effect without having to harm consumers by increasing the post-judgment royalty rate."  I've been thinking it over, and I believe he's right.  An enhancement of the prejudgment royalty shouldn't affect the price at which the infringer will sell the infringing product, but an enhancement of the postjudgment royalty would raise the infringer's marginal cost--thus resulting in higher prices and lower output.  One possible caveat is that if the optimal penalty is more than three times the prejudgment rate, you can't concentrate all the deterrent effect in the prejudgment royalty under U.S. law--but then, it's unlikely anyone knows what the optimal penalty would be in any individual case.