Presumption That Marginal Rate of Return Equals Average Rate of Return

Apotex Inc v Eli Lilly and Company 2021 FCA 149 Boivin JA: Webb, Near JJA affg 2019 FC 1463 Zinn J [Cefaclor Interest]

            1,133,007 – 1,146,536 – 1,133,468 – 1,150,725 [the Lilly Patents]

            1,095,026 – 1,132,547 – 1,136,132 – 1,144,924 [the Shionogi Patents]

Is the Cefaclor litigation finally winding to a close, 25 years after the action was commenced? In Cefaclor liability FC 2009 FC 991 affd Cefaclor liability FCA 2010 FCA 240, Gauthier J found that at least one valid claim of each of Lilly’s patents had been infringed by Apotex. After Gauthier J was appointed to the FCA, Zinn J was assigned to hear the damages reference. His lost profits damages award of $31m was affirmed by the FCA, except in respect of the interest calculation: Cefaclor Damages FC 2014 FC 1254 affd Cefaclor Damages FCA 2018 FCA 217. Zinn J had awarded compound interest on the basis of a presumption that a plaintiff would have generated compound interest [FC 118] (see here), and the FCA remitted on the basis that there is no such presumption [FCA 156] and “a loss of interest must be proved in the same way as any other form of loss or damage” [FCA 158] (as discussed here).

The decision under appeal in this case is the interest decision on remand, which I’ll call Cefaclor Interest FC. In Cefaclor Interest FC, Zinn J set the compound interest rate by assuming that Lilly would use the money it would have had in such a way as to maximize its rate of return [30], [FC 27]. That’s fine, but he also set the rate at a rate equal to Lilly’s average rate of return [72]. The problem, as discussed here, is that as a matter of financial logic, the marginal rate of return—what Lilly would have done with the extra money—is necessarily lower than the average rate of return. Now of course, this is a matter for financial evidence, not just financial logic, but Zinn J considered the evidence of the financial experts to be “unhelpful”, because they carried out their analysis on the assumption that the lost profits were “a sum separate and apart from the other Lilly profits” [12], [FC 53]. That is, he disregarded the expert evidence as to the rate of return precisely because they carried out a marginal analysis. He preferred the evidence of Lilly’s fact witness, which was to the effect that Lilly would have spread the extra money among the same investments that it had made in the real world [13], [FC 55–56]. But there is actually no conflict between the financial experts and the fact witness. That Lilly would have spread the extra money among the same projects does not imply that the rate of return on extra money spent on those projects would be the same as the average rate of return on those projects. On the contrary, as a matter of financial logic, even if it is true that in fact Lilly would have spread the money among the same uses, the marginal return from an extra investment in those uses would have been less than the average return from those uses, again as discussed here. So far as I can tell, there was no evidence on the issue of the marginal rate of return on the same projects, presumably because the experts had focused their analysis on treating the extra money as a separate sum. In effect, Zinn J applied a presumption that the marginal rate of return would be the same as the average rate of return: “Where that proposed use of the slightly larger pool of profits parallels the use Lilly made in the real-world, there must be a heavy burden on Apotex to show that there was something making it impossible for Lilly to do so again” [FC 57].

Where does this leave the law? The FCA decision establishes that a trial judge is entitled to find that the marginal rate of return is the same as the average rate of return. But the issue was treated as a matter of fact at both levels of court, so this doesn’t establish a legal rule that the marginal rate of return is equal to the average rate of return. There is a “heavy burden” on a party seeking to show that average and marginal rates of return are different, but presumably that burden can be discharged by evidence directly on point, which we now know is necessary.

Market Share Expert Not Required for Market Share Damages Analysis

DNOW Canada ULC v Estate Grenke 2020 FCA 61 Dawson JA; Rennie, Rivoalen JJA var’g 2018 FC 564 Phelan J
            2,095,937 / stuffing box

It has been a decade now since Phelan J held Grenke’s 937 patent to be valid and infringed in Weatherford v Corlac 2010 FC 602 — so long that the parties or their names have almost all changed for one reason or another — and this is the fourth (dare I say final?) trip to the FCA. The decision now under appeal, 2018 FC 564 (here), is Phelan J’s award of $8 million in damages [9]-[10]. The defendants alleged a variety of errors by Phelan J in his assessment of the facts. While the FCA noted in couple of places that Phelan J’s reasons might have been more detailed [61], [113], the Court rejected most of these challenges on a deferential standard of review [19]-[20]. No new law was applied (the decision starts with a handy review of damages principles [18]), but there are a couple of points worth mentioning.

The 937 patent relates to an environmentally friendly stuffing box that prevents oil from spilling out of the top of a production oil well. During the relevant period, there were three main manufacturers in this market, including the plaintiffs and the defendants, along with some smaller players [35]. Total sales volume was on the order of a couple of thousand units annually [41]. Phelan J used historical market share as the basis (with adjustments) for assessing lost sales. The FCA affirmed that a market share analysis was appropriate on the facts; this is not a case, such as Alliedsignal 1998 CanLII 7464, with only nine customers in the market, in which a review of the evidence on a customer-by-customer basis would be necessary [68].

The plaintiffs relied on only one witness for the market share analysis [46], and the defendants objected to the witness’ qualifications on the basis that he not qualified as a “market share expert” [48]. (Presumably that means an accountant or economist with expertise in market assessment and modeling.) However, the witness in question was an expert on the industry and its purchasing preferences; he had personal knowledge of product offerings and the merits of available products, and of the market generally, from his long experience as a purchaser in the industry. The FCA held that it was not an error for Phelan J to have relied on this evidence in assessing the market share lost as a result of the infringement [76]-[78].

A final point of interest concerns loss of “convoyed” sales. “Convoyed” products are products that are not themselves protected by the patent, but that are typically sold with, or as a result of, the sale of a patented product [142]. The FCA reaffirmed that “An entity claiming under a patent is ‘entitled to damages assessed upon the sale of non-infringing components when there is a finding of fact that such sale arose from infringing the patented component’ [143] (quoting with approval Beloit v Valmet-Dominion Inc [1997] 3 FC 497 (FCA). However, the FCA stressed that the mere fact that the convoyed goods in question are commonly sold with the infringing goods is not sufficient to establish causation: [153-55]. What is required is “a specific finding based on evidence” that the loss of sales of the convoyed parts was caused by the loss of sales of the infringing products [154]: “‘[s]imply because a non-infringing product appears on the same invoice as a drive is not sufficient to establish causation’” [151] (quoting and agreeing with the appellants’ submission). The FCA was of the view that Phelan J had erred on this point [156], and, after assessing the evidence itself, held that no damages for lost convoyed sales should be awarded [169]. This was the sole point on which the FCA reversed Phelan J.

Determining the Rate of Return for Compound Interest Damages

Eli Lilly and Co v Apotex Inc 2019 FC 1463 Zinn J [Cefaclor Interest] remitted from 2018 FCA 217 Gauthier JA: Gleason, Laskin JJA [Cefaclor Damages] aff’g 2014 FC 1254 Zinn J [Cefaclor Damages FC]

1,133,007 – 1,146,536 – 1,133,468 – 1,150,725 [the Lilly Patents]

1,095,026 – 1,132,547 – 1,136,132 – 1,144,924 [the Shionogi Patents]

Cefaclor Interest is devoted to assessing the proper rate of return for compound interest damages. While the determination was ultimately a matter of fact, it raises issues that will arise in any similar determination. I also suggest that an overarching lesson from the decision is that whatever particular approach is used, the determination of the rate of return as a matter of fact is likely to be a burdensome undertaking. As a matter of policy is might be desirable to amend the relevant statutory provisions to permit a successful party to claim compound interest at a modest rate as a matter of law, without the need for a fact intensive inquiry as to what exactly it would have done with the money.

As discussed here and here, the common law traditionally prohibited recovery of interest on a damages award. While that prohibition was reversed by statutory provisions permitting the award of pre-judgment interest, those statutes generally did not allow for compound interest. In Bank of America 2002 SCC 43 the SCC recognized that simple interest is not fully compensatory, and consequently, the SCC held that even if compound interest was not available under the relevant statute, it was available under the common law of contract as compensation, so long as it was claimed as such and proven.

In Cefaclor Damages FC, the damages portion of a bifurcated trial, Zinn J had awarded lost profits of just over $31m (the Lost Profits), with compound interest on the Lost Profits as damages: [4]. On appeal, in Cefaclor Damages the FCA confirmed that compound interest is indeed available in the patent context when interest is claimed as a head of damages: see here. However, Zinn J had held that the loss could be presumed, and the FCA reversed on this point alone, saying “a loss of interest must be proved in the same way as any other form of loss or damage” [158]. The FCA therefore remitted the matter to Zinn J for reconsideration of this point alone [164].

When compound interest is claimed as damages, the loss is the value of the lost opportunity to use the funds that the plaintiff would otherwise have had — in this case, the Lost Profits. The question in Cefaclor Interest therefore, was whether Lilly had proven a lost opportunity to use the Lost Profits, and how prove the applicable rate of interest: [18].

Use of ex Post Information in Assessing a Reasonable Royalty

Seedlings Life Science Ventures, LLC v Pfizer Canada ULC 2020 FC 1 Grammond J
             2,486,935 / Auto-Injection of Medication

The facts in Seedlings are summarized here along with a discussion of claim construction. Novelty and utility are discussed in yesterday’s post. Despite his holding that all the asserted claims were either invalid or not infringed, Grammond J went on to consider remedies, which had been fully argued. This post discusses a timing issue that arises in assessing a reasonable royalty. In particular, I will argue that the timing of the hypothetical negotiation construct used in assessing a reasonable royalty must reflect its purpose, which is to ensure proper compensation. This implies that there is no strict rule that the hypothetical negotiation takes place at the time of first infringement, nor is there a rule that ex post information cannot be used.

Recovery of Loss from Displaced Sales

Arysta Lifescience North America, LLC v AgraCity Crop & Nutrition Ltd 2019 FC 530 Pentney J
            2,346,021 / flucarbazone sodium herbicide / EVEREST

In Arysta v AgraCity the patentee, Arysta, sought, and was granted, an interim injunction, on the basis that AgraCity’s sale of a generic herbicide would infringe Arysta’s 021 patent. Yesterday’s post discussed the ability-to-pay aspect of the irreparable harm issue. On a separate point, AgraCity argued Arysta would not suffer substantial damages because the 021 patent only covers the granular or powder form of the herbicide.

AgraCity’s product is a generic of the EVEREST 1.0 granular product formerly sold by Arysta [9], but Arysta no now only sells a new and improved liquid product, EVEREST 3.0 [54]. AgraCity argued that because Arysta no longer sold a product embodying the patented invention, “any losses that may flow to Arysta from the sale of its generic product must be limited to a reasonable royalty” [54]. Pentney J rejected this argument, essentially on the basis that the patentee is entitled to any losses caused by the infringement; if the sale of the generic granular product displaced sales of Arysta’s liquid product, those losses would be recoverable [68]. I won’t go through Pentney J’s analysis in detail, except to say that it strikes me as entirely correct.

Airbus v Bell Punitive Damages Award Upheld

Airbus Helicopters S.A.S. v. Bell Helicopter Textron Canada Limited 2019 FCA 29 de Montigny JA: Pelletier, Boivin JJA aff’g 2017 FC 170 Martineau J
            2,207,787 / helicopter landing gear

In Eurocopter v. Bell Helicopter Textron Canada Limitée 2012 FC 113 aff’d 2013 FCA 219 Martineau J held that one of the claims of the 787 patent, owned by Eurocopter (now the helicopter division of Airbus), to be valid and infringed. The patent covers sleigh type skid landing gear for helicopters [8]. In the subsequent Airbus Damages decision, Martineau J awarded $500,000 in compensatory damages (see here and here) and $1,000,000 in punitive damages, plus pre-judgment and post-judgment interest. (I wrote far too much on the punitive damages aspect of the trial decision, trying to analyze the relevant principles. I won’t repeat that here, but for those who are interested the posts are here, here, and here.) Airbus appealed, arguing that the award of punitive damages should have been higher, alleging a variety of errors [29]. Bell cross-appealed on the basis that the award of punitive damages was too high [71]. A few general principles arise from the FCA decision.

The standard of review is deferential

The Court noted that:

[30] It is now well established that appellate intervention with respect to an award of punitive damages will only be warranted where the trial court made an error of law or a “wholly erroneous assessment” of the quantum of damages (Richard v. Time Inc., 2012 SCC 8, [2012] 1 S.C.R. 265 at para. 190 (Time)). In Cinar Corporation v. Robinson, 2013 SCC 73, [2013] 3 S.C.R. 1168 (Cinar), the Supreme Court summarized the applicable standard of review in the following way:

In [Time], this Court held that an appellate court may only interfere with a trial judge’s assessment of punitive damages (1) if there is an error of law; or (2) if the amount is not rationally connected to the purposes for which the damages are awarded, namely prevention, deterrence (both specific and general), and denunciation …

                                    Cinar at para. 134.

Airbus argued that Martineau J had not attached sufficient importance to various factors, such as the blameworthiness of the respondent’s conduct, and the financial means of the defendant [39]. However, Martineau J “correctly identified the factors going to the proportionality of a permissible quantum of punitive damages, and properly applied them to the facts of this case” [50]. So long as that is done, the FCA will not be inclined to reweigh the relevant factors itself [39], [45]-[52], [57].

Prior awards may be used as guideposts

Airbus argued that “the judge’s determination of the quantum of punitive damages was based on his erroneous belief in the existence of a $2,000,000 ceiling” on punitive damages awards [34]. The FCA concluded that Martineau J had not in fact felt himself limited by any such ceiling [36], [38]. Rather, “he saw this scale simply as a helpful indication of the range of punitive damages previously awarded” [36]. Moreover (emphasis added):

[37] In my view, there is nothing untoward or inappropriate in using a range of previous awards, as guideposts, in assessing the quantum of punitive damages. Even the Supreme Court, in Whiten, took such a range into account, when it concluded that the award was “certainly at the upper end of a sustainable award on these facts but not beyond it” (at para. 4). It is, in fact, a very common practice, and a sound one for that matter, to consider previous awards in assessing the quantum of punitive damages in a particular case.

Lubrizol is of limited precedential value

In Lubrizol 58 CPR(3d) 167 (FCTD) rev’d 67 CPR(3d) 1 (FCA) the FC had initially awarded $15,000,000 in punitive damages. This was overturned on appeal because the quantum of the compensatory damages had not been considered in the analysis [40] and they had not yet even been assessed. As the FCA stated in Lubrizol, “the Court cannot decide whether exemplary damages are required until after it decides whether the general damages were insufficient for punishment and deterrent purposes.” Airbus argued that nonetheless the FCA in Lubrizol had “implicitly agreed” that $15,000,000 was appropriate [41]. The FCA in this case rejected that reading of Lubrizol [41], and went on to state that

[43] Lubrizol was decided almost 25 years ago and does not seem to have been given much precedential value. I have been unable to find any other case where such a large award of punitive damages has been made, and counsel has not drawn our attention to any such case. [On review of other awards] Lubrizol would therefore appear to be the outlier in terms of the significance of the award in punitive damages, and the judge was certainly entitled to distinguish that decision from the present case.

Post infringement mitigating conduct should be considered

Bell had originally used the infringing gear, known as the “Legacy” gear, but after an infringement action was brought in Canada and other jurisdictions (including the US and France), had switched to non-infringing Production gear [10]. Martineau J had taken this into account as a mitigating factor in assessing punitive damages. Airbus argued that doing so was an error of law. It argued, based on the discussion of “potential harm” in Whiten 2002 SCC 18 [117], that “the seriousness of the prejudice must be assessed at the time of the wrongful behaviour, whether or not actual prejudice ensued” [60]. This FCA rejected this, pointing out that it was clear from the SCC discussion in Whiten that “what the Court really wanted to prevent was for bare luck to be considered as a mitigating factor” [61].

Further, the FCA held that “it was open to the judge to regard the steps taken by the respondent after the infringement as a mitigating factor” [62], and indeed,“[n]ot taking into account the conduct of the respondent after being notified of the violation would be antithetical to [the holistic and balancing approach endorsed in Whiten]” [63, my emphasis]: and see [64].

Sanctions for the same infringement in other jurisdictions may be taken into account

Airbus argued that Martineau J had erred in law in considering other penalties, in particular the likelihood of damages for infringement in parallel US and French proceedings [65],[66]. (And see [FC 435], [FC 440]. The FCA held that this may be taken into account, at least so long as the proceedings related to the same infringement [66] and are actually likely to result in a sanction [67]. (Though in any event, on the facts Martineau J appears to have considered this factor to be neutral [69].)

Finally, Bell cross-appealed on the basis that the award of $1,000,000 was more than the minimum necessary to meet the purposes of punitive damages [71]. The FCA rejected this argument, holding that on the facts, it was open to Martineau J to conclude that substantial punitive damages were warranted [75].

Miscellaneous Issues in Cefaclor Damages

Apotex Inc v Eli Lilly and Co 2018 FCA 217 Gauthier JA: Gleason, Laskin JJA aff’g 2014 FC 1254 Zinn J
            1,133,007 – 1,146,536 – 1,133,468 – 1,150,725 [Lilly Patents]
            1,095,026 – 1,132,547 – 1,136,132 – 1,144,924 [Shionogi Patents]

There are a couple of miscellaneous points arising out of Cefaclor Damages FCA that I’d like to address.

Grain Processing
In her remarks on general principles, Gauthier JA noted that

[48] [I]t is important to understand that our Court did not simply import an American law concept in a wholesale fashion. The Court in Lovastatin may indeed have referred to American authorities in order to better ground the concept. But one must be careful not to construe references to American jurisprudence lending support for the NIA defence as a blind incorporation of, or strict adherence to, the reasoning adopted by American courts.

This is all fair enough in the abstract, but a bit obscure; the FCA seems to be implying that there is some particular aspect of American law which is not good law in Canada, but without specifying exactly what.

I wonder if the remark might have been aimed at Grain Processing 85 F3d 1341 (Fed Cir 1999), in which the Fed Cir allowed the infringer to rely on an NIA which was not in existence at the relevant time. Grain Processing was relied on by Apotex both in Lovastatin FCA and in this case. Gauthier JA pointed out that Grain Processing, can be distinguished on the facts, because in that case the NIA was a process that increased the cost by only 2.3%, while in this case the increased costs were raised by at least 40%. I agree; and there is another related distinction. In Grain Processing, the only reason the infringer did not develop the infringing process earlier was that it did not know it was infringing. The infringer knew of the patent, and was trying to design around it, and thought it had succeeded, but there was a technical dispute over exactly how the “dextrose equivalent value” specified by the claim was to be measured (using the “Schoorl test” or the “Lane-Eynon test”). The infringer guessed wrong. But the point remains that once the true construction of the claim had been determined, the infringer had no difficulty designing around the claim and developing a substantially equivalent non-infringing process. That is why the Fed Cir in Grain Processing ultimately held that the infringer would have used the non-infringing process from the outset; it was very clear on the facts that it could have developed the NIA much earlier and would have done so had it known that the process it was using was infringing. This is very far from facts in this case, in which it was difficult to design around the patent and the resulting process was not economically competitive [44e, n]. Again, I’m not sure that Gauthier JA’s general remark was actually aimed at Grain Processing, and I do agree with Gauthier JA that we cannot assume that US law on this issue is correct in every respect. But Grain Processing itself is entirely sound, given its unusual facts, and because of those unusual facts it provides a particularly striking example of the NIA defence; but it is only equally unusual facts that will generate the same kind of result.

Convoyed Sales and Remoteness
On a different issue, Apotex also argued that it is an error of law “to award damages for sales displaced by non-infringing products because such sales are beyond the scope of the Patent Act [and] are too remote” [112]. Gauthier JA re-affirmed that the well-established law that damages for lost sales of so-called convoyed goods are indeed recoverable [114], [122], and on the facts the sales at issue were not too remote [127]. But Gauthier JA did appear to acknowledge that in some circumstances convoyed sales might be too remote for damages to be recoverable, even though the lost sales were caused by infringement [123]-24]. That is, causation is not necessarily the only limitation on recovery in patent damages. Gauthier JA did warn that remoteness is not usually an issue, and it should be raised as soon as possible or it may be considered to have been waived [123].

Compound Interest is Here to Stay

Grenke v. DNOW Canada ULC 2018 FC 564 Phelan J
            2,095,937

This decision is the damages portion of a bifurcated action, following Weatherford Canada Ltd v Corlac Inc 2010 FC 602,, in which Phelan J held the plaintiff Grenke’s 937 patent to be valid and infringed by the defendants. (Weatherford, a licensee of Grenke, has settled, and the names of various parties, including Corlac, have changed through acquisition and otherwise, which is why the case names are so different: [11]-[17]). The patentee elected damages rather than an accounting [2].

In oil production, a “stuffing box” is used to seal the top of an oil well and prevent leakage of the oil as it is being pumped out of the ground. The 937 patent claims a new type of environmentally-friendly seal assembly, which minimizes leakage [18]-[19]. There was substantial regulatory and environmental pressure to minimize leakage, and the patented technology was a game-changer [21], [45]. The decision is quite straightforward, and it is a good example of damages assessment in the context of a simple product, in which the patented technology contributed a major part of the value of the end-product. (This is in contrast to complex products, such as a smartphone, where any single technology contributes only a small part of the overall value.) Perhaps the most legally interesting aspects are that Phelan J allowed full recovery for convoyed sales, refused to grant punitive damages, and granted compound interest.

To begin, Phelan J emphasized the need to use a “broad axe” in assessing damages: while perfect compensation is the goal, the reality of litigation is that perfection can rarely be achieved: [71]-[73]. This point was repeated at a number of instances, though in most cases the inaccuracy in question would be small in any event: [84], [99], [101], [119], [130], [168]

The basic framework is to assess how many sales the patentee lost as a result of the infringing competition, and to calculate lost profits on those sales, and then to assess a reasonable royalty on all the remaining infringing sales. The parties did not really disagree on this framework, or indeed any major conceptual issues, but only on the application to the facts.

Phelan J accepted that market share was appropriately used by the plaintiffs in this case as a proxy for determining the plaintiff’s lost sales [91], [96], pointing out that “This is not a case with a small number of customers. . . which might require “customer-specific” evidence” [97]. Phelan J held on the facts that the appropriate market was “the ‘environmentally friendly stuffing box market’” [105]. There is something of a debate as to whether and when fixed costs can be deducted in calculating profits (see here, here, and here), but in this case the parties apparently agreed that only variable costs should be deducted [115].

Reasonable Royalty
With respect to the reasonable royalty, Phelan J remarked:

[136] In constructing the hypothetical negotiation of the royalty rate, it is important to take into account the realities on the ground. It is particularly relevant that GrenCo had market strength because of its technology and also important is that all the evidence suggests Grenke would have been an unwilling seller, difficult to deal with, and prepared to push the limits of “reasonableness” to the edge (and perhaps over). This all suggests a rate at the upper end of reasonableness. It is somewhat ironic that Grenke would be deemed, under this hypothetical negotiation, to accept any amount of royalty rate given his general intransigence to the point of “going down with his ship” rather than to settle. However, the Court must accept some of the realities in constructing the hypothetical negotiation and resolution.

There is indeed considerable artificiality in constructing a hypothetical negotiation in a case in which it is clear on the facts that the patentee would never have licensed in reality. But recall that a reasonable royalty is only assessed in respect of sales which the patentee would not have made in the real world; if the patentee can prove that it would have made the sale, then it will be entitled to lost profits. Sometimes, this will be because the infringer was selling into a market that the patentee had not entered. In this case, the parties were competing in the same market, and patentee is seeking lost profits on infringing sales that would not have been captured by the patentee, but which would instead have gone to a different competitor with a non-infringing product. That is, for reasons of pre-existing relationships, distribution, etc, if the particular customer had not bought from the defendant, it would have bought from a non-infringing third party. In that case it makes sense for the patentee to accept that since it would not have made the sale in any event, it would be better to allow the infringer to make the sale as a licensee, so that the patentee gets a return in the form of a royalty, rather than losing out altogether. The artificiality arises because it is only with hindsight that we know that the patentee would lose the sale in any event. At the time, it would have made sense for the patentee to fight for the sale, rather than license the defendant, because (unlike the case in which the defendant is selling into a different market), the patentee would rather have the profits on the sale than a royalty, and it can’t license only for those sales it won’t get, because it doesn’t know at the time that it won’t get the sale.

The larger point, perhaps, is that the objective is to assess a royalty which is reasonable in the circumstances, and a hypothetical negotiation is no more than a useful heuristic, which should not be taken too literally if problems arise in applying it. With that all said, I am not suggesting that it is wrong to assess a reasonable royalty “at the upper end of reasonableness” when the patentee was competing in the same market. I don’t have an opinion on that issue. My point is that the hypothetical negotiation framework is not particularly helpful in resolving the issue.

Turning to the calculation itself, the parties both used the “anticipated profits approach,” supplementing this analysis with the “minimum willingness to accept (MWA) versus maximum willingness to pay (MWP)” [135].

The anticipated profits approach supposes that the parties to a hypothetical negotiation, which takes place at the time of first infringement, split the profits they anticipate the infringer will make from the use of the patented technology. As noted by Phelan J, this approach has previously been used in Jay-Lor 2007 FC 358, and AlliedSignal 1998 CanLII 7464, 78 CPR (3d) 129 aff`d (1999) 86 CPR (3d) (FCA). In “A New Framework for Determining Reasonable Royalties in Patent Litigation,” (2016) 68 Florida Law Review 929, Tom Cotter and I argue that instead of anticipated profits, a split of the actual profits should be used. The difference is important only when actual profits are different from the profits that would have been anticipated at the time of first infringement. Despite the name “anticipated profits” approach, I believe that using actual profits would be consistent with Canadian cases. Note that the anticipated profits already relies on actual profits to a large extent. Profits are sales multiplied by profit margin per unit, and the “anticipated” profits approach normally (always, in Canadian cases) uses the actual sales, not anticipated sales, in calculating the “anticipated” profits. We argue that there is no good reason to use actual sales combined with anticipated profit margin. Further, it appears that in AlliedSignal, the calculation did use the actual profits [213]-[214],though it is a bit difficult to tell as the detailed calculations are in a confidential Appendix. (Note that AlliedSignal never used the “anticipated profits” terminology.) Jay-Lor did refer to anticipated profits, but in calculating anticipated profit margin, it appears that actual post-infringement costs (“over the past four years”) were used to assess the profits that would have been anticipated [142]. If that is right, Jay-Lor used an actual profits approach, with an anticipated profits terminology, or, at most, there was no difference between actual and anticipated profits. In this case, the details of the calculation are not discussed, as the parties were largely in agreement, so it is not clear whether actual or anticipated per unit profits were used. In summary then, if there is a case in which anticipated and actual profits are different, it would be open to Canadian courts to use an actual profits methodology, and in my view this would be preferable in principle, for reasons explained in detail in my article with Professor Cotter.

With respect to the MWP versus MWA analysis, Phelan J noted that in Merck v Apotex 2013 FC 751, (blogged here) “the Federal Court indicated that if there were no overlap between the MWA and the MWP, then the reasonable royalty is the MWA. The “minimum” of the MWA must, logically, be the lower end of the range – therefore, the application of Merck FC to this case leads to the conclusion that 8% is a reasonable royalty rate” [143]. Here again, I would suggest that we keep in mind that a reasonable royalty is being assessed only in respect of sales which the patentee would not have captured. In such circumstances, the patentee’s MWA will be its marginal cost, normally zero, because if the negotiation breaks down, the patentee will not make the sale itself, and so will not get any return at all. If the MWA is higher than the MWP, it must be that there was a conceptual error in assessing the MWA (assuming the patented technology is actually valuable, not detrimental, as was clearly true in this case). The details of the assessment of the MWA was not described. However, the MWP / MWA analysis gave much the same result as the anticipated profits analysis.

In US law, the most commonly used approach to assessing a reasonable royalty is to look to comparable licences. In this case, the defendants suggested that lower rates that had been negotiated in respect of the same technology by Weatherford should be used. Phelan J rejected this, correctly in my view, noting that the deal included cross-licensing which would have depressed the rates.

Phelan J also allowed recovery of lost profits for convoyed sales, on the basis of but for causation, without any further restriction:

[163] If compensation is not provided for such add-on items, then the patentee is not being put in the position that they would have been in but-for the infringement (resulting in less than perfect compensation). If the Plaintiffs can show, on a balance of probabilities, that such sales would have been made by the Plaintiffs in the but-for world, then in my view this is a loss for which they should be compensated. This approach is consistent with the limited Canadian case law on the topic, and I would reject the requirement suggested by the Defendants that such convoyed sales must have no function independent of the patented object.

In my view, Phelan J’s conclusion is sound for the reasons he gives. If the convoyed sales are reasonably foreseeable from the sale of the patented technology, as was clearly the case here, then the patentee would face a choice between charging a higher price for the patented goods, and losing some sales and convoyed sales, or charging a lower price for the patented good, in anticipation of profits from convoyed sales. These are just different strategies for realizing the value of the patented technology, and the damages award should not turn on which marketing strategy was chosen.

Punitive Damages
The plaintiff sought punitive damages. Phelan J noted that:

[186] At the liability phase of this action, the Court found that the Defendants:

a) intentionally set out to create a product which they knew or ought to have known would infringe the Patent;

b) have used a stuffing box design which is the same as the Grenke design, and profited from the sale of the infringing articles over a 10 year span (until enjoined by this Court from further infringement);

c) were considered by this Court to have been unjustified and egregious; and

d) had denied infringement of the Patent, yet at the same time claimed ownership in the Patent and argued that the Plaintiffs had in fact infringed their patent rights.

Phelan J refused to award punitive damages, emphasizing that “patent infringement alone, even knowing infringement, is not sufficient to ground an award of punitive damages,” [185] (citing Bauer Hockey 2014 FCA 158 [29]), and similarly, “there must be something more than knowing infringement to support an award of punitive damages”[189]. The various allegations regarding infringement and validity were considered part of the “brawl” over rights in the technology. No doubt this can be distinguished on the facts from Martineau J’s Eurocopter decision (blogged here), but it does evince the traditional reluctance of the Canadian courts to award punitive damages.

Interest
The Federal Court has historically been reluctant to grant compound pre-judgment interest, in part for the very good reason that the relevant statutes relating specifically to interest did not allow compound interest: see Federal Courts Act, s 36(4)(b), as noted by Phelan J at [205] (and the discussion here). There has been a recent and welcome trend to granting compound interest nonetheless, based not on the interest provisions, but on the basis that the loss is recoverable as an element of compensation, under s 55(1). The rationale is that in reality, a party with more money in its pocket will put that money to work, and reinvest the profits, which will also go to work, and compound interest is therefore required to provide full compensation. That is the approach taken by Phelan J at [193], [209] following Zinn J in Eli Lilly / cefaclor, 2014 FC 1254 (discussed here). Phelan J stated that “I concur with Justice Zinn’s comments that in today’s world, compound interest is an accepted form of redress” [194].

An interesting point arises from the approach of awarding compound interest on the basis of s 55(1). Phelan J began his discussion of interest by noting that the award of interest under s 36 of the Federal Courts Act is subject to judicial discretion to disallow interest under s 36(5), which discretion is “to assist the court in controlling the litigation process and to avoid inappropriate compensation” [191], quoting Apotex v Merck 2006 FCA 323 [140]. Phelan J also stated that compound interest was also awarded “[a]s a matter of discretion taking into account the equities and the conduct of the Defendants” [211]. But s 55(1) is normally understood as entitling the patentee to its legal damages, with only equitable remedies, such as an accounting, being subject to discretion. If compound interest is awarded as compensation under s 55(1), can it be discretionary? I’m not sure of the answer to this question, particularly in light of the argument for allowing the court to use its discretion in order to control the litigation process. The question did not arise in this case, as Phelan J did not exercise his discretion to disallow compound interest for the period for which it was claimed, but it presents something of a conundrum. No doubt the best solution would be updated interest provisions in the Federal Court Act, but there is little prospect of that happening, and I expect the issue will have to be resolved judicially.

Legal Facts in the “But For” World: Does it Matter Who Might Have Been on the SCC in Calculating Damages?

Apotex Inc v AstraZeneca Canada Inc 2018 FC 181 Locke J
            2,139,653 / esomeprazole / NEXIUM / S 8 NOC

In AstraZeneca 2017 SCC 36, an action involving the 653 patent, the SCC abolished the promise doctrine (here). In so doing, it held the 653 patent to be valid,¹ reversing the decisions in the courts below which had applied the promise doctrine to invalidate it. But before that action, in 2010 FC 714, Apotex had prevailed in the NOC proceeding. This decision, by Locke J, is the s 8 damages decision following from that NOC decision. Section 8 proceedings entitle the generic to recover damages for having wrongly been kept out of the market as a result of the statutory stay triggered by the NOC proceedings in which it prevailed. But what if ultimately, in the infringement action, the patentee prevails? The key question is whether Apotex should be entitled to recover s 8 damages for having been wrongly kept out of the market, when it turns out that it never had the right to enter the market. In AstraZeneca v Apotex 2017 FC 726 Barnes J held that the answer was no: see here. Locke J has now come to the same conclusion, and he therefore dismissed Apotex’s s 8 action [112]. While Locke J’s reasoning is somewhat different from that of Barnes J, the decisions are consistent. 

The relationship between s 8 damages and damages in a subsequent infringement action will soon be of historical interest only, as the new NOC proceeding will take the form of an action. This decision does, nonetheless raise a point of more general importance regarding the nature of hypothetical legal facts in the “but for” world. For example, does it matter who is on the Supreme Court of Canada in the but for world?

The technical point is that s 8 damages, like all damages, are assessed as the difference between the plaintiff’s actual position and the position it would have been in “but for” the wrong (ie but for the statutory stay in the case of s 8 damages). The patentee’s general argument in a case in which the patent is subsequently held to be valid is that in the but for world, the generic would have infringed, and would have liable to the patentee for infringement, and that liability in the “but for” world would offset any profits the generic would have made had the statutory stay not prevented it from entering the market. In effect, the generic did not suffer any lost profits as a result of the statutory stay, because the patent, which we now know to be valid, would have prohibited it from entering the market anyway.

It is clear from prior FCA decisions, in particular Lovastatin 2011 FCA 364 (here) and Omeprazole 2013 FCA 77 (here), that liability for damages for infringement should be taken into account in assessing s 8 NOC damages. There was a preliminary dispute in this case as to whether it should be taken into account under s 8(1), which establishes liability (AstraZeneca’s position), or under s 8(5), which allows the court to consider all relevant matters in assessing quantum (Apotex’s position). Locke J, relying on Lovastatin, held that Apotex’s “but for” infringement should be taken into account using s 8(5) [62]. I think that’s right as a matter of the interpretation of Lovastatin, but I have to admit that it is not entirely clear to me what difference it makes: as Locke J pointed out, it wouldn’t have made any difference on the facts of this case: [64].

The more interesting point was how to construct the “but for” world:

[76] . . .Apotex argues that there is no evidence that any of the following would have happened in the but-for world: (i) AstraZeneca would have sued Apotex for patent infringement, (ii) the parties would not have settled the matter before a trial, (iii) the SCC would have granted leave to appeal, or (iv) the SCC would have decided such an appeal the same way. Accordingly, Apotex argues that, for the purposes of constructing the but-for world, the 653 Patent should be treated as invalid just as it was in the real world during the Delay Period.

On the details, Apotex argued, for example, that the SCC would not necessarily have decided the same way because in the “but for” world, the matter would have come to trial at a time when Rowe J, who wrote the AstraZeneca decision, was not on the Court [88].

AstraZeneca’s position was that this should all be irrelevant:

[77] It argues that we now know that the 653 Patent was valid during the Delay Period, and that it should be treated as such for the purposes of the assessment of compensation under s. 8 of the Regulations, regardless of whether it would have been found valid in the but-for world.

Locke J found it unnecessary to decide the question. Relying in large part on the principle from Ramipril s 8 2014 FCA 67 [145] that “All steps that were taken in the real world should be assumed to have been taken in the but-for world unless there is evidence upon which the trier of fact may reasonably conclude that different steps would have been taken,” he concluded that all of Apotex’s hypotheticals should be resolved against Apotex. Since AstraZeneca prevailed on the facts, it was not necessary for Locke J address the legal argument.

The point is nonetheless an important, to the extent that different facts might warrant a different conclusion. In my view, AstraZeneca was right on this point. The reason for this is emphasized in an article by David Taylor: “reasonable royalties should reflect the value of patented technology rather than patent rights” (49 Ga Law Rev 79. 89, original emphasis). Taylor made that observation in the context of reasonable royalties, by way of explaining the established rule that in determining a reasonable royalty using the hypothetical negotiation approach, the patent should be assumed to be valid, even though in the hypothetical “but for” negotiations, the parties would not have known it to be valid. The same principle applies in the context of lost profit damages. To elaborate, the purpose of assessing damages for patent infringement is to preserve the incentive to invent. In order to promote socially beneficial inventions, the incentive to invent should be commensurate with the social value of the invention. That social value is equal to the incremental value of the invention over the best non-infringing alternative: see eg The Sedona Conference, Commentary on Patent Damages and Remedies, 23-24; David O. Taylor, David, Using Reasonable Royalties to Value Patented Technology,” 49 Ga Law Rev 79. 91-97; Lee & Melamed, “Breaking the Vicious Cycle of Patent Damages,” (2016) 101 Cornell Law Rev 385, 411-12. As I said in my article “A Remedial Benefit-Based Approach to the Innocent User Problem” 920014) 20 CIPR 79, “A patent is ‘nothing very special’ when it provides only a small advantage over the next best publically available alternative. The differential profit approach, by comparing the actual profits with those which would have been made with the next best alternative, provides a precise means of identifying a patent which is ‘nothing very special.’”

The “but for” world is a construct for determining the true social value of the patent invention. The “but for” world should therefore be sensitive to factors which affect the value of the patented technology. If a drug treats a disease that has become an epidemic, it should be more valuable than one that treats a disease that was anticipated but which never emerges. This gives the inventor a strong incentive to focus its efforts on inventions that it believes will be socially valuable. The “but for” world should not be sensitive to factors which solely affect the value of the patent rights, such as whether an injunction would have been granted, or whether the patent would have been held invalid. The grant of an injunction by a court may increase the value of the patent right to the patentee, but it does not increase the social value of the patented technology.

This suggests that legal facts about the “but for” world should be treated like past events: “past events must be proven, and once proven they are treated as certainties” (Athey v Leonati [1996] 3 SCR 458, [28]). This is consistent with Locke J’s observation in response to Apotex’s argument that the SCC’s decision to change the law and reject the promise doctrine was not foreseeable (my emphasis):

[96] Even accepting all of the facts asserted in the preceding two paragraphs, the fact remains that the 653 Patent is, and always was, valid. This is the case in the real world even if I accept Apotex’s allegations that the law on the Promise Doctrine would not have evolved in the but-for world as it did in the real world. Whether or not such allegations are justified, the fact is that Apotex claims compensation for loss as a result of being prevented from infringing AstraZeneca’s valid patent.

¹There was a preliminary dispute as to whether the SCC in AstraZeneca had held the 653 patent to be valid, or merely that it is not invalid for lack of utility. Locke J held that the SCC had held the 653 patent to be valid [30].

Standing of Plaintiffs Carrying on Activity Outside of Canada

Teva Canada Ltd v Janssen Inc 2018 FCA 33 Dawson JA: Webb, Gleason JJA aff’g 2016 FC 593 Hughes J and 2016 FC 727
            1,304,080 / levofloxacin / LEVAQUIN

As discussed in Tuesday’s post, the FCA dismissed Teva’s appeal on a number of factual points that do not raise issues of general importance. But Teva v Janssen is a significant decision on standing. It raised a novel question as to whether a foreign party which has a licence from the Canadian patent owner, but which does not need the licence because it does not carry out any infringing activity in Canada, has standing as a person claiming under the patentee pursuant to s 55(1). The FCA, affirming Hughes J, held that the claimant does indeed have standing in this situation:

[126] I reject Teva’s submission that Janssen US was required to demonstrate that it engaged in conduct in Canada that would otherwise amount to infringement.

[127] A party need only establish that they enjoy rights under a patent in order to be a person claiming under the patentee.

In so doing, the FCA also affirmed and distinguished Servier 2008 FC 825 (aff’d without discussion on this point, 2009 FCA 222), in which Snider J held that the foreign parties did not have standing in similar circumstances. Between the two cases, this should provide good guidance as to when a licensee which does not carry on activity in Canada does or does not have standing. However, I must say that I do not find the FCAs explanation of Servier entirely persuasive, and I am inclined to think it is better to regard Servier as having been effectively overruled.

Quantum, Proportionality, and Uncertainty in Punitive Damages

Airbus Helicopters SAS v Bell Helicopter Textron Canada Limitée 2017 FC 170 Martineau J [FC Damages]
            2,207,787 / helicopter landing gear

As I noted in my post on Airbus’ entitlement to punitive damages, most jurisdictions do not award enhanced or punitive damages for patent infringement, or for civil actions more generally, because, as explained by Lord Reid in Cassell & Co Ltd v Broome [1972] UKHL 3 [1972] 1 All ER 801, it is undesirable to expose a party to a quasi-criminal sanction when both entitlement and quantum are so vaguely defined. With respect to quantum, he noted “[t]here is no limit to the punishment except that it must not be unreasonable.”

In US patent law uncertainty regarding quantum is addressed by limiting the total award to a maximum of three times compensatory damages assessed: 35 USC § 284. This means that the punitive damages are limited to twice the compensatory damages. So, in this case, the award of $1,500,000, comprised of $500,000 in compensatory damages and $1,000,000 in punitive damages, would be the maximum allowable under US law.

In Canadian law there is no such hard limit, either with respect to patents or more generally, though this may be because punitive awards have tended to be more modest and less of often granted. In Whiten v. Pilot Insurance Co 2002 SCC 18, [70], the SCC recognized the problem of uncertainty relating to quantum:

[70 [T]he incantation of the time-honoured pejoratives (“high-handed”, “oppressive”, “vindictive”, etc.) provides insufficient guidance (or discipline) to the judge or jury setting the amount. . . . A more principled and less exhortatory approach is desirable.

See similarly [39]. Thus the purpose of the principled approach is to provide guidance in assessing quantum. Does it?

Record Quantum

The Dow Chemical Co v NOVA Chemicals Corp 2017 FC 637 Fothergill J
            2,160,705 / film-grade polymers / ELITE, SURPASS

In the liability phase of this action, Dow v NOVA (No 1) 2014 FC 844 aff’d 2016 FCA 216, O'Keefe J held Dow’s 705 patent related to advanced film-grade “mLLDPE” polymers, to be valid and infringed by Nova. He also held that Dow was entitled to damages under s 55(2) of the Patent Act for pre-grant ‘infringement’, and that Dow was entitled to elect between damages and an accounting for post-grant infringement. Dow subsequently elected an accounting [107]. Fothergill J’s subsequent decision in the remedies phase, Dow v Nova (No 2) 2017 FC 350, addressed various issues to allow the parties’ accountants to calculate the actual sums owed by Nova to Dow [6]. (See my posts on reasonable compensation, springboard profits, fixed costs, and interest.) The parties were then able to largely agree on the quantum, but they returned to the court for further direction on three issues, which were addressed in this decision.

The first issue was whether Nova was entitled to deduct capital depreciation expenses for the division used to make the infringing product [5]. Fothergill J held it was not so entitled, as a matter of interpreting his own prior decision, without adding any substantive analysis [8].

Second, in Dow v Nova (No 2) [165], Fothergill J had allowed Nova to deduct a “proportional amount” of fixed costs, which he considered properly attributed to the production and sale of the infringing products. But what is “proportional”? Dow proposed an allocation based on relative production volumes [10], which Forthergill J accepted, noting that while Nova’s alternative proposals relied on “potentially valid distinctions,” Nova had not adduced evidence to support them [13]. As I suggested in my previous post on fixed costs, the theoretically ideal method is probably to allow full deduction of opportunity costs. Allowing a deduction of a portion of fixed costs is a half measure between that and refusing deduction entirely. Even though not ideal, that approach may be justified on the basis of practical administrability as compared with a full opportunity cost approach. If it is correct to say that deduction of proportional fixed costs is justified on that pragmatic basis, it is not surprising that there is no single theoretically appropriate basis for the apportionment.

The third issue is the timing of the conversion of certain capital expenditures from Canadian dollars to U.S. dollars for the purpose of calculating annual capital depreciation [15]. Nova sold most of its infringing product in the US and retained its profits in US dollars (Dow v Nova (No 2), [189]), and Fothergill J held the profits to be disgorged should be calculated in US dollars and converted to Canadian dollars as of the date of judgement (Dow v Nova (No 2), [176]). Nova was entitled to deduct some part of the depreciation on the plant that made the infringing product, but that plant was built in Canada and paid for in Canadian dollars. The question was how the depreciated amount should be converted to US dollars to be deducted from the annual profit, which was calculated in US dollars. Dow’s position was that the appropriate date for conversion was when the expenditure was incurred, so that the full capital amount would be converted into US dollars at the exchange rate for that year, then annual depreciation would be calculated in US dollars [16]. Nova’s position was that the annual depreciation should be calculated in Canadian dollars, and then converted to US dollars on an annual basis, using the exchange rate for the year of the deduction, on the view that the point of depreciation is to recognize that capital-related expenses are multi-year expenses [16], [18]. The issue was important because the exchange rate had fluctuated substantially over the relevant period, and “Dow complains that [Nova’s] approach results in a greater deduction for depreciation in USD than the costs Nova actually incurred in USD,” [17] I’m not sure I’m entirely persuaded by Nova’s point, since no one was taking issue with the depreciation itself, and even if it is true that capital-expenses are multi-year expenses, it does seem strange that the amount of the depreciation should be greater than the actual cost incurred. On the other hand, the conversion rate when the plant was built was unusually low [20], and Forthergill J was concerned that using it as the single point of conversion “would be punitive,” and he therefore adopted Nova’s approach, which was reasonable and supported by the evidence.

In the result, Fothergill J awarded Dow just under $645 million including pre-judgment interest to April 7, 2017, the date of judgment in Dow v Nova (No 2), plus pre-judgment interest to the date of this judgment. This seems to be a record award in reported Canadian patent cases.

Entitlement to Punitive Damages, Revisited, Revisited. . .Revisited

Eurocopter v. Bell Helicopter Textron Canada Limitée 2013 FCA 219 Mainville JA: Noël, Trudel JJA [FCA Liability] aff’g 2012 FC 113 Martineau J [FC Liability]

Airbus Helicopters SAS v Bell Helicopter Textron Canada Limitée 2017 FC 170 Martineau J [FC Damages]
            2,207,787 / helicopter landing gear

I have to admit, I’m not doing very well with this case. Monday’s post tried to correct a misinterpretation in my post on the 2013 FCA Liability decision affirming that Airbus / Eurocopter was entitled to punitive damages, and only a few hours after posting, I added an addendum to reflect another new interpretation. Maybe I should declare defeat and move on. But quantum of punitive damages and entitlement are intimately linked – entitlement to punitive damages turns on the court’s outrage at the infringer’s conduct, and the quantum must be proportionate to that same outrage – which is why I had to reexamine the Liability decision in order to blog on the punitive damages aspect of the Damages decision. And the issue is an important one. Punitive damages have traditionally been awarded only in exceptional circumstances in Canadian patent cases, and Bell is the case which comes closest to typical infringing behaviour. (Contrast the previous leading case, Lubrizol, in which the infringer breached an interlocutory injunction). Because punitive damages are a quasi-criminal sanction (Whiten 2002 SCC 18, [36]), parties must be able to predict, and so avoid, conduct that will attract punitive damages. So in this post, I will try, once again, to what substantive rules governing entitlement to punitive damages emerge from FCA Liability 2013 FCA 219 aff’g FC Liability 2012 FC 113. Consider this the post that I should have written on the punitive damages aspect of the 2013 FCA Liability decision.

Punitive Damages and Infringer’s Belief that Patent is Invalid

Airbus Helicopters SAS v Bell Helicopter Textron Canada Limitée 2017 FC 170 Martineau J
            2,207,787 / helicopter landing gear

This Bell Damages decision follows the decision of Martineau J in the liability phase of this bifurcated trial, Bell Liability 2012 FC 113, aff’d 2013 FCA 219, in which he had held one of the claims of the 787 patent, owned by Eurocopter (now the helicopter division of Airbus), to be valid and infringed. In the Damages decision, Martineau J awarded $500,000 in compensatory damages and $1,000,000 in punitive damages, plus pre-judgment and post-judgment interest [2], [108]. After writing two long posts on the compensatory damages award in March (see here and here), I procrastinated on writing a third post, which I knew would be just as long, on the punitive damages issue. Now that I have returned to the case, I realize that I didn’t deal properly with the FCA Liability decision on the issue of whether the infringer will avoid punitive damages if it believes that the patent is invalid. In this post, I will deal with that issue.

To recap my post on the Bell FC Liability decision, Martineau J held that Bell’s conduct was sufficiently egregious to justify awarding punitive damages to Eurocopter, in large part because Bell had intentionally copied the landing gear, when it knew or should have known that it was patented:

[433] This is a case of wilful blindness or intentional and planned misappropriation of the claimed invention. Eurocopter has proven that the infringement of the ‘787 Patent by the making and use of the Legacy gear was not innocent or accidental.

Now, we can certainly imagine a more innocent infringement. A infringer might have independently invented the invention, and practiced it in complete ignorance of the patent until approached by the patentee. Or it might have independently invented it, and then discovered the patent, and chosen to go ahead nonetheless. It might have copied the technology without realizing it was patented. But, as I argued in my posts on both the FC and FCA decisions, even copying and intentional infringement of a technology which the infringer knew to be patented, is not necessarily objectionable. I was (and remain) concerned with the chilling effect on validity challenges if that conduct alone were sufficient to justify punitive damages. Suppose a pharmaceutical patentee has tried to evergreen a valuable drug patent by product-hopping to a new version protected by a formulation patent that claims and enteric coating or extended release, and, after a detailed assessment of the patent, a generic company reasonably concludes that the patent is probably invalid for obviousness — after all, such patents are regularly, though not invariably, held to be invalid by Canadian courts. The generic then launches a competing version which it knows will infringe. Surely, this infringement, though planned and deliberate, does not constitute misconduct worthy of punitive damages, even if the generic's reasonable and good faith opinion that the patent was invalid ultimately turns out to be wrong. On the contrary, the generic is doing a public service by attacking a patent that is wrongly increasing the burden on the health care system.

The situation is arguably quite different if the defendant infringes a patent which it believes to be probably valid. That is the position taken by the FCA is the Bell FCA Liability decision. This is where I misinterpreted the FCA decision. In my post on the FCA decision, I said the main reason the FCA held punitive damages were justified was “simply the factual finding that Bell intentionally infringed.” While the FCA did refer to Martineau J’s finding of intentional infringement [186], the FCA summarized by saying (my emphasis) that

[192] Where a person infringes a patent which it knows to be valid, appropriates the invention as its own, and markets it as its own knowing this to be untrue, punitive damages may be awarded. . .

In my post I said that the FCA reasoning “does imply that punitive damages would not be available if the infringer had had a good faith belief that the patents it knew it was infringing were invalid.” Saying it was an “implication” is much too weak. The Court began its discussion by saying this (my emphasis):

[185] [Bell] submits that infringing a patent is not conduct which intrinsically merits punishment where the infringer did not know of the existence of the patent or reasonably held that the patent was invalid. I agree that it would be difficult to uphold a punitive damages award in such circumstances. However, these are not the circumstances applicable to this case.

Moreover, [192] came at the end of the FCA’s summary of the evidence in this case, so the reference to a person infringing a patent which it knows to be valid is not a general statement about possible scenarios, but rather a statement about this particular case.

Thus, contrary to the statement in my earlier post, according to the FCA, intentional infringement of the patent alone will not normally justify punitive damages; what is required is intentional infringement of a patent which the infringed believes to be valid. In this case, according to the FCA, punitive damages were justified because Bell deliberately infringed a patent which it knew to be valid, and marketed the invention as its own.

Here is my excuse for my misreading. The evidence reviewed by the FCA at [186] as supporting the award of punitive damages, establishes that Bell intentionally copied the landing gear which it knew or should have known was patented (see FC Liability [431-34]), but I do not see any finding by Martineau J that Bell believed the patent to be valid, or that it’s belief to the contrary was unreasonable, either in the cited paragraphs, or anywhere else in either the FC Liability decision or the FC Damages decision. So far as I can see, the question of whether Bell believed the patent to be invalid during the time it infringed was never addressed by Martineau J; indeed, it seems not to have been raised at all. My statement that the main reason the FCA approved the award of punitive damages was simply intentional infringement, was based on the evidence reviewed by the FCA, rather than the statements in [185] and [192].

I’m not really sure to make of this. Maybe the FCA read the cited paragraphs of the Liability decision differently from me, and sees them as finding that Bell believed the patent was valid. Maybe there is something else in the record that isn’t reflected in the FC Liability decision which establishes that Martineau J did indeed find that Bell believed the patent to be valid, though it would be surprising for a key factual issue not to be referred to. [Addendum: At [185] the FCA said that this is not a case in which it was established that the infringer reasonably believed the patent was invalid. It has just occurred to me that this is strictly true on a straightforward reading of Martineau J's Liability decision, since there was no finding one way or the other on that point. That implies that in order for the infringer to escape punitive damages, the onus is on the infringer to establish that it reasonably believed the patent was invalid (or, presumably, that it reasonably believed that its product would not infringe). On that view, the FCA decision is consistent with Martineau J’s finding. If that is the correct interpretation, it would have been helpful for the FCA to have spelled this out more explicitly.]

In any event, the point to emphasize is that, according to the FCA, an award of punitive damages will rarely be justified in a case in which the infringer reasonably believed the patent was invalid.

The argument for awarding punitive damages is stronger when the infringer believed the patent to be valid, than in my hypothetical scenario of an objectively weak patent. Whether the infringer believed the patent to be valid is a distinction with a difference. Nonetheless, I am not persuaded that intentional copying of an invention which the infringer believes to be valid, in itself justifies punitive damages. A requirement that the infringer believed the patent to be valid is relevant and important, because it mitigates the chilling effect; but the chilling effect is not eliminated. If Canadians are spending millions of dollars annually on a product protected by a patent that is 70% likely to be valid, there is a 30% chance that those millions of dollars are being wasted on a product that should be available far more cheaply. The expected benefit to Canadians if the patent is invalidated may well outweigh the burden on the patentee of having to litigate. the rule that punitive damages are awarded on a showing that the infringer intentionally infringed a patent which it believed to be valid is essentially the US rule (treble damages for willful infringement), and that rule has not been an unqualified success. It adds to complexity of patent trials, by raising the question of the infringer’s belief as to validity, which is otherwise irrelevant; it adds legal costs by increasing the need for opinion letters, whose main purpose may be to defeat a potential claim for treble damages; it increases uncertainty as to potential liability, as the quantum turns on this difficult to prove consideration. factor; and it may deter innovators from reading and learning from issued patents, as a protection against a finding of willfulness. That doesn’t mean that punitive damages are never warranted in patent cases, but in my view there should be some egregious element beyond knowing infringement, even if the patentee believes the patent to probably be valid. Lubrizol Corp. v. Imperial Oil Ltd, 58 C.P.R. (3d) 167, (FCTD) rev’d [1996] 3 FC 40, 67 CPR(3d) 1 (FCA) is an example, in which the punitive damages were assessed primarily because the infringer had acted in callous disregard of an injunction. In this case, the fact that Bell presented the invention as its own is an additional factor, though whether that factor alone would have justified punitive damages if Bell had believed the patent to be invalid, is a different question, and I'm not sure that question was answered by either the FC or the FCA.

Update: see my more recent post for a more complete discussion of this decision.