Another Failed Attempt to Give Independent Meaning to Overbreadth

GreenBlue Urban North America Inc v DeepRoot Green Infrastructure, LLC 2023 FCA 184 Gleason JA: Woods, Mactavish JJA varg 2021 FC 501 McDonald J

2,552,348 / 2,829,599 / Integrated Tree Root and Storm Water System

DeepRoot’s 348 and 599 patents relate to a landscaping system to promote healthy urban trees using a subsurface structural cell system that supports the hardscape (eg sidewalk and paving). At trial, McDonald J held DeepRoot’s patents to be valid and infringed, in a straightforward decision which turned on the facts: see here.

At trial, GreenBlue argued that the patents were overbroad. This argument was dismissed by McDonald J on the facts after a brief discussion [FC 255]–[261]. The FCA decision further illuminates the nature of the argument (my emphasis):

[65] Finally, with respect to overbreadth, GreenBlue asks us to reach a different conclusion from the Federal Court and to find that metal rods, which were not claimed in the two Patents, are essential for the invention to work.

[66] This submission is premised on the assertion that the evidence established that without the metal rods, the invention would not function. In support of this contention, GreenBlue relies on the fact that the load testing report related to DeepRoot’s final commercial product showed that the product failed without the inclusion of the metal rods attached to the lid.

The FCA dismissed the appeal on this point on the straightforward basis that there was evidence to support McDonald J’s finding on the facts [69].

Of more general interest, this is another failed attempt to give some independent content to the concept of overbreadth. In Seedlings 2021 FCA 154, the FCA affirmed that “overbreadth is a distinct ground of invalidity that must be considered separately” [50]: see here. Ever since then, parties attacking a patent have been trying to give some content to overbreadth as a distinct ground of invalidity, largely without success. (The closest thing to success was in Rovi Guides v BCE Inc 2022 FC 1388: see here.) In this case, we see another failed attempt. This kind of argument would traditionally have been framed as inoperability, which is to say lack of utility. No doubt it would have failed if framed that way as well.

What Causation Concept Is to Be Used in Allocating Fixed Costs?

GreenBlue Urban North America Inc v DeepRoot Green Infrastructure, LLC 2023 FCA 184 Gleason JA: Woods, Mactavish JJA varg 2021 FC 501 McDonald J

            2,552,348 / 2,829,599 / Integrated Tree Root and Storm Water System

DeepRoot’s 348 and 599 patents relate to a landscaping system to promote healthy urban trees using a subsurface structural cell system that supports the hardscape (eg sidewalk and paving). At trial, McDonald J held DeepRoot’s patents to be valid and infringed, in a straightforward decision which turned on the facts: see here. McDonald J denied DeepRoot’s request to be granted an accounting because she was not satisfied that GreenBlue had in fact made a profit on infringing sales [FC 280]. GreenBlue appealed on liability and DeepRoot cross-appealed McDonald J’s refusal to grant an accounting. The FCA dismissed GreenBlue’s liability appeal on the basis that McDonald J had not made any material error of law and her findings of fact were supported by the evidence.

The FCA held that McDonald J erred in refusing to grant an accounting. The contentious issue was the deductibility of fixed costs. These are costs, such as the cost of lighting the factory or the cost of the financial department, which are necessary to the infringement—you can’t make infringing widgets if the lights are off—but which do not vary with the extent of the infringement and which would have been incurred even if there had been no infringement at all—the lights will be on whether the widgets the factory is making are infringing or non-infringing.

Deductibility of fixed costs is a tricky question. On the one hand, if the fixed costs are the same whether or not the product is infringing, this means that those costs are not caused by the infringement, and so should not be deducted. On the other hand, fixed costs are actual costs of production; the plant cannot make infringing products if the lights are turned off and a business that does not cover its fixed costs will not be profitable. In Nova v Dow 2020 FCA 141, Stratas JA was persuaded by the latter argument, and held that “[a]ny infringer, regardless of whether it is operating at full capacity, should be able to deduct a proportion of its fixed costs” [162]. This is referred to as the “full costs” approach, in contrast with an “incremental costs” approach in which only costs that vary with the infringement are deductible. See here for a more extended discussion. (The issue of deductibility of fixed costs was not addressed by the SCC on appeal in Nova v Dow 2022 SCC 43.)

The difficulty with a full costs approach is determining what proportion of the fixed costs can be deducted. Nova v Dow itself doesn’t help us answer that question. Fothergill J at trial had allowed deductibility of some fixed costs, but on the basis of an opportunity cost approach endorsed in Dart Industries [1993] HCA 54, according to which the proportion of fixed costs to be deducted turns on the proportion of fixed costs which sustained the foregone opportunity: 2017 FC 350 [161] But Stratas JA explicitly rejected Dart Industries[154], which Fothergill J had relied on, which means that we cannot rely on an opportunity costs approach to tell us what proportion of the fixed costs should be deducted. (While he rejected the reasoning underpinning Fothergill J’s deduction, he nonetheless approved the deduction itself, apparently on the view that Dart Industries is a kind of fixed cost approach, and since a full costs approach is not unsound, it was not an error for Fothergill J to use a full costs approach even though it was based on a wrong principle [163]. As discussed here, the Dart Industries approach is arguably a variation on the incremental cost approach, but that doesn’t really matter at this point.) The point of all this is that while we know the principle applied by Fothergill J in Nova v Dow itself, that does not help us decide what proportion of the fixed costs are to be deducted, because the principle applied in Nova v Dow was wrong.

Turning back to the case at hand, the discussion of this point at trial is brief and cryptic. McDonald J [FC 278] noted the statement in Nova v Dow 2020 FCA 141 that “the ‘full costs’ approach should always be available to an infringer” [145] and then remarked that “[h]aving accepted the financial evidence of GreenBlue, I am not satisfied that GreenBlue has in fact made a profit on sales of RootSpace” [FC 280]. Gleason JA’s decision is not much clearer as to the specifics, but it appears that the main fixed cost at issue was general overhead expenses [72] eg keeping the lights on, and it seems that McDonald J had allowed a deduction of fixed costs proportionate to the percentage of sales generated by the infringing product as a proportion of total sales [92]. (It was uncontested that incremental costs may be deducted [66].)

DeepRoot’s argument, which was accepted by the FCA, was that:

[72] the Federal Court erred in law by failing to conduct the necessary analysis to establish a causal connection between GreenBlue’s claimed overhead and the profits it earned through infringement, thereby allowing GreenBlue to wrongfully shield its profits from disgorgement. DeepRoot adds that there was no evidence to establish what portion of GreenBlue’s general overhead expenses were related to the infringing sales and that it was accordingly an error for the Federal Court to have accepted GreenBlue’s percentage allocation.

So, McDonald J’s error was a failure to establish a causal link between the fixed costs and the infringement.

In Nova v Dow, Stratas JA had implied that it is only the costs that are “caused by the infringement” that are disgorged: [153] (see similarly [157]). The requirement of a causal connection was emphasized even more strongly by Gleason JA in this case. In her review of the law, she stated that “An accounting of profits. . . requires that the defendant disgorge to the plaintiff the amount of profits earned by reason of the infringement” [79] and “With respect to both the remedy of damages and that of disgorgement, proof of a causal connection to the infringement is required” [81]. Similarly, she stated that the full costs approach allows deduction of “fixed overhead costs causally connected to the infringing sales” [85] and “fixed costs that are causally attributable to the infringing product” [88]. She concluded that:

[89] fixed non-incremental overhead costs may be deducted from sales to establish an infringer’s profit, but proof of causation is still required. In other words, the defendant must establish some link between the claimed portion of the overhead and the infringing sales.

This is all well and good. But what is the nature of the causal link? We need to know the causation concept for the parties to know what kind of evidence is relevant.

The most prominent causation concept in law is “but for” causation. On that view, costs are caused by the infringement if they would not have been incurred but for the infringement. In other words, if the causation concept is “but for” causation, then we end up at the incremental cost approach. Since Stratas JA rejected the incremental cost approach in Nova v Dow, “but for” causation can’t be the relevant causation concept. And indeed, Stratas JA also rejected “but for” causation fairly explicitly: see Nova v Dow FCA [148], [151], [153].

The other main causation concept known to law is material contribution. But the material contribution test is disfavoured and is confined to special circumstances, in particular where it is impossible to determine which of a number of negligent acts by multiple actors in fact caused the injury, but it is established that one or more of them did in fact cause it: Clements v Clements 2012 SCC 32 [13], [17], [42]. While the SCC has not ruled out the possibility that material contribution might be applied in other contexts, but no one has ever said this is appropriate in the context of deduction of fixed costs. Moreover, it is difficult to see how it could be applied, since material contribution goes to liability, not apportionment. If the defendant is found liable under a material contribution approach, the defendant is jointly and severally liable for the entire loss: Clements [12]. It is an all or nothing result, which provides no basis for an apportionment.

What other causation concepts might be used? While Stratas JA in Nova v Dow rejected “but for” causation, he did not specify an alternative causation concept. In the absence of a clear causation concept, we might then look to the facts to discover what approach is appropriate. Nova v Dow itself does not provide an answer. Stratas JA only referred in passing to the causation requirement and he did not expressly endorse any causation concept. Nor did he give any example of an appropriate deduction. The closest he came was to say that an infringer would only be entitled to deduct a proportion of its fixed costs: “For example, if an infringing product occupies 1% of a factory’s production capacity or volume, only 1% of the fixed costs will be deducted” [161]. This was only by way of a passing example to show that the full costs approach does not imply that the infringer would be able to subsidize its non-infringing products. More importantly, what is the causation concept which supports deduction of a proportionate amount by sales, or profits, or weight, or plant capacity, or whatever it might be? Without a principled causation concept, a deduction based on any of these factors is arbitrary and unprincipled.

As noted, it appears that in this case that McDonald J allowed deduction of fixed costs proportionate to the percentage of sales generated by the infringing product as a proportion of total sales [92]. Gleason JA gave a convincing example to show why this is not always appropriate [94]. But it is unlikely that a deduction proportionate to volume or capacity is always preferable.

In this case, it seems that the failure to call evidence was the key deficiency: see [98]–[101]. It is reasonable that neither volume, nor capacity, nor sales would always be the correct basis for apportionment, but any of them might be, depending on the evidence. But in the absence of a causation concept, what exactly is that evidence trying to prove? How can we know what constitutes evidence of causation, if we don’t know what “causation” means?

Gleason JA gave the example of the trial decision in Nova v Dow, 2017 FC 637, in which Fothergill J allocated fixed costs based on “billed volume” in light of evidence that “the fixed costs per pound were substantially the same for infringing and non-infringing products” [98]. I must admit that I don’t understand what it means to say that fixed costs were the same for infringing and non-infringing products. Fixed costs are costs which do not vary with production, so I don’t see how such costs can be the same for different products, except in a trivial sense that the cost of keeping the lights on is the same no matter what the plant is producing. Perhaps the expert report itself would be more illuminating, but frankly, I doubt it. If an expert has developed a new and appropriate causation concept, it is important as a matter of law that the court tell us what is. In the absence of a specific causation concept which is to be applied to the facts at hand, evidence purporting to tie fixed costs to one product or another is nothing more than a campfire story—a tall tale spun by an expert, but a tall tale nonetheless.

Gleason JA did not specify a causation concept. Instead, she concluded her discussion by saying that “What the foregoing examples demonstrate is that the approach to quantifying overhead costs for purposes of establishing profits earned through infringement is highly fact-dependent” [95]. She then remitted the matter to McDonald J. It is true enough that application of the appropriate causation concept to the facts is highly fact dependent. As the SCC said in Clements v Clements [9], “[t]he ‘but for’ causation test must be applied in a robust common sense fashion.” But the SCC in Clements did specify the causation concept—namely “but for” causation. It is not enough to ask the trial court to look to the facts and apply common sense. Yes, the court must look to the facts — but what are they supposed to be looking for? It is the job of the trial court to apply the law to the facts, and it is the job of the Court of Appeal to tell the trial court what the law is, and the nature of the causation concept is a matter of law.

In the absence of any causation concept, the expert witnesses will tell some kind of story as to why their preferred deduction has some ‘causal’ link to the infringement. McDonald J will have to accept one side or the other, and as long as there is some kind of evidence to point to containing the word “cause”, presumably the FCA will review her decision on a deferential standard and that will be the end of it. But unless and until the FCA tells us what the causation concept is, the deduction of fixed costs in this case and in all future cases will simply be arbitrary and unprincipled.

In my view the appropriate causation concept is “but for” causation. If the FCA wants to abandon “but for” causation in this context, it is incumbent on the Court to tell us what to replace it with. I know the FCA is just following in the footsteps of the SCC in Nova v Dow, in which the Rowe J insisted on the need for a causal link while steadfastly refusing to tell us the causation concept. But the failure of the SCC to define causation is all the more reason that the FCA has to step up.

Nova v Dow: The Value of the Invention

Nova Chemicals Corp v Dow Chemical Co 2022 SCC 43 Rowe J: Wagner CJ, Moldaver, Karakatsanis, Brown, Martin, Kasirer and Jamal JJ concurring; Côté J dissenting affg Nova Chemicals Corporation v Dow Chemical Company 2020 FCA 141 Stratas JA: Near, Woods JJA affg Dow Chemical Co v Nova Chemicals Corp 2017 FC 350, 2017 FC 637 Fothergill J

2,160,705 / film-grade polymers / ELITE SURPASS

The Intuition / The Legal Background / Causation as a Matter of Fact / The Concession / What Role for “But For” Causation in Identifying the NIO? / Summary of the Summary / Causation Concept in the Absence of an NIO / What is the NIO?

To this point, I have argued that Rowe J’s decision is unprincipled, in the literal sense that there is no unifying causation concept, and incoherent, in the sense that it does not provide any practical guidance as to how to identify the non-infringing alternative (NIO), which is at the center of his three-part test. In my first post in this series, I suggest that the reason for this is that Rowe J was driven by a powerful intuition that the value of the invention should not depend on fluctuations in unrelated markets. It is now time to return to that intuition and show why it is wrong, appealing though it is.

The basic problem with Rowe J’s decision is that he rejected but for causation, without providing any other causation concept in its stead. I suggest that his rejection of “but for” causation was driven by his commitment to the use of conventional seed as the appropriate NIO in Rivett 2009 FC 317, even though, as Rowe J specifically noted, conventional seed was in fact unavailable and so could not have been the appropriate comparator on strict “but for” causation. I suggest that the central intuition driving Rowe J’s decision is that the value of the patented invention is the difference between the profit that was made with the patented soybean and the profit that could have been made with conventional soybean, whether or not conventional soybean was actually available. The idea is that the value of the invention is an intrinsic property that cannot depend on the happenstance of whether conventional soybean is or is not in stock at the local Feeds ‘n Needs. As Stratas JA put it in his FCA decision:

[78] The patent is not valuable because it just so happens to be unavailable in a particular locale. It is valuable because it has some inventive quality that increases the infringing product’s profitability or marketability.

While Rowe J did not specifically approve this comment, I suspect that he was persuaded by the same intuition. This view is supported by Rowe J’s repeated insistence that the profits to be disgorged are those causally attributable to “the invention,” specifically rejecting Cote J’s emphasis on profits attributable to “the infringement” [61] and also rejecting Côté J’s suggestion at [91] that the focus was on the “value of the invention in the hands of the infringer” [60].

In the first post in this series I gave two examples to illustrate why the use of “but for” causation can lead to counter-intuitive results. Suppose it can be established that Nova would indeed have made $300m in profit in the market for commodity grade plastic. If we were to apply “but for” causation, this means that the profits to be disgorged would be $300m instead of $600m. Now put a twist on the facts, and suppose that while Nova was actually making infringing food-grade plastic film, a prominent Instagram influencer started wearing Crocs as a fashion accessory, with different colours to match different outfits, with the result that Crocs took the world by storm. The craze was so dramatic that the price of commodity grade plastic rose 20%. In that environment, but for the infringement, Nova could and would have made $400m in the market for commodity grade plastic. If we apply “but for” causation to this hypothetical, the amount to be disgorged would be only $200m.

This result is very counter-intuitive. What actually happened is that Nova made and sold infringing food-grade plastic. How can it be that some quirky fashion trend in an unrelated market could reduce the amount to be disgorged by $100m, even though the value of the infringing plastic remains exactly the same?

To take a more extreme example, what if Nova could show that it could have made a profit in a tight market for Chinese tea? Should that profit be deducted? The principle is no different than the example of commodity grade plastic, but surely the price of tea in China has nothing to do with the profit attributable to Nova’s infringement of a patent on plastic food wrap. As Dow put it, if there is no market substitute, there is no logical basis for limiting disgorgement of the actual profits simply because “the infringer might, in a but-for world, have hypothetically sold completely different products, in a completely different market, to completely different customers” (Dow Factum [73]).

The appeal of Rowe J’s approach is that it avoids these paradoxes. Under his approach, the profit to be disgorged remains the same, regardless of the vagaries of unrelated markets. What Nova might have done but for the infringement is irrelevant: if there is no NIO, then the amount to be disgorged is $600m in any scenario; and if there is an NIO, that NIO remains the same, regardless of what else Nova might have done, and the amount to be disgorged therefore remains the same, regardless of what might be going on with footwear fashion or the price of tea in China.

I’ll suggest that there are two aspects to this intuition. One aspect is that an invention has some inherent value that can be revealed by a proper comparison. We might call this the positive intuition. Another aspect is that even if the invention does not have any inherent value, at least we can say that the amount to be disgorged should not vary with the vagaries of entirely unrelated markets. We can call this the negative intuition.

The positive intuition can be addressed briefly (though I go on at more length in the article itself). A comparison between Schmeiser and Rivett, both of which were approved and relied on by Rowe J, shows that an invention does not have an inherent value. Both involved infringement of the same patent and the same comparator, namely conventional seed, but in Schmeiser the disgorgement was zero while in Rivett the disgorgement was substantial. This is because the nature of the infringement was different: Mr Schmeiser did not take advantage of the unique properties of the patented invention, while Mr Rivett did. Thus, the quantum disgorged turned not just on the invention, but on the infringement. With all due respect to Rowe J, it is perfectly clear, on the authorities approved by Rowe J himself, that the amount to be disgorged does indeed reflect the “value of the invention in the hands of the infringer” [60].

However, this does not address the negative intuition. It seems entirely reasonable that the quantum to be disgorged should vary according to the infringer’s use of the infringing invention itself, or fluctuations in the market for the infringing product. But it is a very different thing to say that the amount to be disgorged should vary with fluctuations in entirely unrelated markets, such as the market for Crocs or the market for Chinese tea. I suspect that this is the fundamental intuition that was driving Rowe J.

The general point is that on a “but for” accounting, the infringer will be permitted to deduct the profits it would have made in whatever other endeavour it would in fact have engaged in, no matter how unrelated to the invention itself; and to the extent that profits in that alternative market fluctuate, the profits to be disgorged will fluctuate, even if the market for the patented product itself is perfectly stable. How can that make sense?

This intuition is powerful but misguided. It is true that the value of the invention does not vary with fluctuations in an unrelated market. But in an accounting, it is the profit earned by the infringer as a result of the infringement that is disgorged. This turns on the value derived from the attributes of the invention, in combination with value derived from the attributes of the infringer. The value of the infringer’s unique attributes might very well change with fluctuations in other markets in which those attributes might also be exploited. So, in this case, Nova’s special attribute is its efficient ethylene production method. That is an advantage in making the patented plastic, but it is also an advantage in making commodity plastic. If the market for commodity plastic fluctuates, the value of the patented plastic will not change, but the value of Nova’s efficient ethylene production method does change. The profits to be disgorged will therefore also change because those profits reflect the infringer’s special attributes as well as the invention’s special attributes.

It may seem unfair to the patentee that the amount to be disgorged may vary with fluctuations in some unrelated market. Even if the infringer had special advantages in that other market, what is that to the patentee? And indeed, it would be wrong if the patentee’s incentive to invest in the patented technology were to vary with some unrelated market. The easy answer to this problem is that the patentee can seek damages, which will not vary with fluctuations in unrelated markets, whether or not the infringer has any special expertise in those markets. This is why damages are used to preserve the patent incentive.

The power of the paradox arises because it is true that the value of the invention itself does not change with fluctuations in unrelated markets, and it is also true damages awarded will not change with fluctuations in unrelated markets. But it is wrong to transfer this intuition from the damages context to the accounting context because the difference between damages and an accounting is precisely that the special characteristics of the infringer do matter in the accounting context, but do not matter in the damages context. In an accounting, the focus is on the infringer. The amount to be disgorged can fluctuate according to the market for an entirely unrelated product because a comparison with the profits the infringer would have made in the unrelated market is what allows us to distinguish the value of the infringement from the value contributed by the infringer’s specialized assets. The other market may be unrelated to the market for the infringing product, but it will be related to the infringer’s special advantages, and as the other market fluctuates, the value of the infringer’s advantages will fluctuate correspondingly.

This resolves the tea in China example I gave above. If the infringer did not have any special advantage in the Chinese tea market, it will not be able to prove that it would have invested in that market; and even if it does, it will not be able to prove that it would have made any extra profit. The market for Chinese tea is an efficient market and so the return that a naive investor makes will simply be the risk adjusted return to capital. There is nothing wrong with deducting that, as it reflects the cost of capital in the same way as borrowing or investing in bonds, which seem much less outlandish.

So, if we conceive of the purpose of an accounting as being to disgorge the value of the invention in the hands of the infringer, or, equivalently, as requiring disgorgement of profits caused by the infringement, then it makes perfect sense that the quantum may depend on what happens in markets that are unrelated to the market for the invention, but which are related to the infringer’s special attributes or expertise.

The real question is normative. Should we conceive of the disgorgement as focusing on the value of the invention in the hands of the infringer, or should we try to isolate the value of the invention in isolation from the value contributed by the infringer? That question is addressed in Part II of my article.

Nova v Dow: What is the NIO?

Nova Chemicals Corp v Dow Chemical Co 2022 SCC 43 Rowe J: Wagner CJ, Moldaver, Karakatsanis, Brown, Martin, Kasirer and Jamal JJ concurring; Côté J dissenting affg Nova Chemicals Corporation v Dow Chemical Company 2020 FCA 141 Stratas JA: Near, Woods JJA affg Dow Chemical Co v Nova Chemicals Corp 2017 FC 350, 2017 FC 637 Fothergill J

2,160,705 / film-grade polymers / ELITE SURPASS

The Intuition / The Legal Background / Causation as a Matter of Fact / The Concession / What Role for “But For” Causation in Identifying the NIO? / Summary of the Summary / Causation Concept in the Absence of an NIO

As described in previous posts, the heart of Rowe J’s approach to an accounting of profits is Step 2 of his three-step test, which asks the court to determine whether there is “a non-infringing option that can help isolate the profits causally attributable to the invention” [15]. The two key concepts are the “non-infringing option” (“NIO”) and profits “causally attributable to the invention.” If we knew the causation concept, it would allow us to determine whether there is an appropriate NIO in future cases. However, as discussed in previous posts, Rowe J did not provide any coherent causation concept and the causation concept at play in Step 2 is particularly obscure. But all is not lost. If we have a sufficiently precise definition of the NIO, we can still apply the three-step test, even if we don’t know what the causation concept is; and indeed, we might even be able to infer the causation concept from the definition of the NIO.

Unfortunately, Rowe J did not provide any clear definition of an NIO. The closest he came to a definition was to say that it is “any product that helps courts isolate the profits causally attributable to the invention” [58]. Without a definition of profits “causally attributable to the invention,” this definition is circular: the NIO is anything that helps identify profits caused by the invention, and the profits caused by the invention are the profits revealed by a comparison with the NIO. Rowe J also stated that whether an appropriate NIO exists is a question of fact [67], which is not helpful in the absence of a clear causation concept; if we knew the causation concept, then it would be a matter of fact to decide whether a proposed NIO reflected that causation concept, but if both the causation concept and the NIO are a matter of fact, then one cannot inform the other.

We might still try to extract some guidance from the specific examples approved by Rowe J as illustrating an appropriate NIO, namely Schmeiser 2004 SCC 34, Rivett 2009 FC 317 and its companion case Janssens 2009 FC 318 vard 2010 FCA 207 and Imperial Oil Ltd [1997] 2 FC 3 (CA). The first three all involve the same patent, relating to a gene conferring herbicide resistance in crops, and in all three the NIO used by the court and endorsed by Rowe J was a conventional crop of the same type. In Imperial Oil, the product was motor oil with a patented additive, and the appropriate NIO, according to Rowe J, was “motor oil without the patented additive” [57].

The most obvious suggestion is that an appropriate NIO is a true market substitute for the infringing product. This was the position taken by Dow. However, Rowe J expressly rejected this definition, saying the NIO “need not be a strict market substitute for the patented product” [67]. Reinforcing this, he also suggested that even commodity plastic might have been an appropriate NIO if the trial judge had made a factual finding to that effect [72]. Note however, that while Rowe J held that the NIO “need not” be a market substitute, he did not say that it is irrelevant whether the proposed NIO is a market substitute; that is, he left open the possibility that market substitutability might be considered as a factor.

Another possibility is that the NIO should be non-infringing but otherwise the same as the infringing product in a technical sense, which is to say, the same product without the patented feature. As noted, in discussing Imperial Oil Rowe J said the NIO was “motor oil without the patented additive” [57] and in discussing Schmeiser he said that all the profits earned were causally attributable to “non-patented features of the sold product—the canola seed itself” [52]. The idea is that we can reveal the value of the invention by comparing the profit actually made with the profit that would have been made by selling the same product without the patented feature.

However, this suggestion doesn’t work if we take Rowe J’s reference to the “patented” feature as meaning the feature claimed by the patent. In Schmesier, the canola seed itself was in fact claimed: Claim 22 and dependent claims were to the transformed plant cell and those claims were infringed: 2001 FCT 256 [127]. Similarly, in Imperial Oil, what was actually claimed was not the additive, but the motor oil with the additive. In both those cases, the product without the claimed feature is no product at all.

It might be suggested that Rowe J was speaking loosely, and he did not mean to refer to the feature that was patented in the strict sense of being claimed, but rather to the inventive concept. So, even though the motor oil with an additive was claimed in Imperial Oil, what was inventive was the additive itself. However, contrary to Rowe J’s suggestion, the motor oil without the additive is clearly not the appropriate comparator, because modern automobiles require some kind of detergent additive. The same motor oil without the additive would have been completely unsaleable. The value of the invention was not the value of the patented oil over the same oil without any additive, but the rather value of the patented oil over the oil with an additive that was previously known in the art. Thus, the appropriate comparator would not be the motor oil without the additive, but the motor oil with a different additive. Further, the inventive concept may not be defined, especially if obviousness is not litigated. Obviousness was not litigated in Schmeiser, for example, and we actually don’t know what was inventive about the claimed canola cell in Schmeiser; it might have been the gene, or the method of inserting the gene into the cell, or even the idea of the type of herbicide resistance gene to use, etc. In many cases, for example if the inventive concept is the method, it is difficult to even define what it means to be the same product without the inventive concept.

Consequently, an NIO cannot be defined either as a true market substitute, or as the same product without the invention, whether we take that to mean the claimed feature or the inventive feature. Nor, as we have seen, can it be defined as simply whatever the infringer would have done but for the infringement.

It is therefore impossible to extract any kind of clear guidance, either from Rowe J’s definition of the NIO or from the examples which he approved. However, it seems clear enough that it would be open to a trial judge to consider any or all of these factors—what the infringer would have done but for the infringement, whether the proposed NIO is a market substitute, whether the proposed NIO is technically similar to the infringing product—not as a definitive test, but rather as factors to be considered in determining whether a particular product is an NIO. (I don’t consider this to be an exhaustive list—the NIO is anything the trial judge finds helpful.) That is what I suggest will be the most likely outcome in practice: counsel will propose a comparator that seems intuitively reasonable on the facts, and the trial judge will accept or reject it as an NIO based on an ad hoc list of factors, each of which may have more or less weight depending on the case.

Nova v Dow: Causation Concept in the Absence of an NIO

Nova Chemicals Corp v Dow Chemical Co 2022 SCC 43 Rowe J: Wagner CJ, Moldaver, Karakatsanis, Brown, Martin, Kasirer and Jamal JJ concurring; Côté J dissenting affg Nova Chemicals Corporation v Dow Chemical Company 2020 FCA 141 Stratas JA: Near, Woods JJA affg Dow Chemical Co v Nova Chemicals Corp 2017 FC 350, 2017 FC 637 Fothergill J

2,160,705 / film-grade polymers / ELITE SURPASS

The Intuition / The Legal Background / Causation as a Matter of Fact / The Concession / What Role for “But For” Causation in Identifying the NIO? / Summary of the Summary

This post continues the series summarizing the key points in my two-part article on Nova v Dow which is forthcoming in the IPJ. So far, we have considered the causation concept at play at Step 2 of Rowe J’s three-part test when there is an appropriate non-infringing option (NIO). But what if there is no appropriate NIO? In that case, the entire actual profits—revenue minus costs—must be disgorged. Rowe J insisted repeatedly that only profits caused by the invention are to be disgorged and he did not limit this to cases in which there is an NIO. So what causation concept is at play when there is no NIO?

The best reading of the decision is that when there is no NIO, Rowe J considers that the entire actual profits are caused by the infringement. In one sense, it is clearly true that the entire actual profit, which is to say the actual revenue from the sale of the infringing plastic less the actual cost of making and selling that plastic, was caused by the sale of the patented plastic. However, this causation concept—which we might dub “actual causation”—is different from “but for” causation. This is illustrated by the crumbling skull doctrine. If the defendant, driving negligently, hit the plaintiff in the knee so that the plaintiff required a knee replacement, it is true in a straightforward and intuitive sense the entire harm to the plaintiff’s knee was caused by defendant hitting it, just as Nova’s entire actual profits are caused by its infringement. But if it can be established that the plaintiff had a pre-existing degenerative knee condition and would have required a knee replacement within two years in any event, the defendant will not be liable for the entire harm, but only for the loss of the two years that the plaintiff would have enjoyed before he would have required a new knee in any event. This is simply the application of “but for” causation: see Athey v Leonati [1996] 3 SCR 458 [35].

Dow argued that the “but for” logic would apply only if Nova sold a direct market substitute for the infringing plastic on the view that “a patent does not confer a complete monopoly if a defendant could make or sell a non-infringing version of the patented invention” Dow Factum FM020 [72]–[73]. This statement is true in the damages context: a defendant will only be able to deduct sales it would have made by selling a market substitute. This is because the patentee is claiming damages for its lost sales and the infringer would only have been able to take some of those sales by selling a market substitute. But that does not reflect a principle that only profits caused by the sale of a market substitute are not caused by the infringement; rather, it reflects the way that “but for” causation plays out in the specific context of damages.

So, while actual causation is an intuitively reasonable causation concept, it is different from “but for” causation. Actual causation, which determines the amount to be disgorged when there is no NIO is obviously also different from whatever causation concept is at play when there is an NIO, simply because when there is an NIO, some amount is deducted from the actual profits.

A second issue addressed in Nova v Dow was whether the infringer is required to disgorge “springboard” profits. These are profits earned after the expiry of the patent which the infringer only earned because the pre-expiry infringement gave the infringer a head start in building sales capacity and market share. Rowe J relied on “but for” causation as a justification for holding that springboard profits must be disgorged:

[81] [A]fter the patent expires, the infringer can use this sales capacity and market share to earn profits that it would not have earned but for the infringing activity that occurred during the life of the patent. Thus, a portion of such post-expiry profits may be causally attributable to infringement of the invention.

So, the three-step test uses two different causation concepts, namely actual causation when there is no NIO, and some other causation concept when there is an NIO, and both of these are different from “but for” causation, which is used in the context of springboard profits. The use of different causation concepts in different contexts is not simply loose language or an oversight; Côté J pointed out the inconsistency explicitly and at length at [178]–[182].

Further, while Rowe J set out a three-step test to “conceptualize” an accounting [15], he did not use that three-step test in the context of springboard profits. This means that it is not clear what approach should be taken to various categories of damages that were not directly addressed in the decision, such as convoyed goods (unpatented goods normally sold with the patented product), fixed costs, and the cost of capital.

This means that there are three distinct causation concepts at play in Nova v Dow— “but for” causation in the context of springboard damages, actual causation, when there is no NIO, and whatever causation concept is at play when there is an NIO—and Rowe J gave no reason why one rather than the other is used in the different contexts. Further, there are also two different frameworks for assessing the amount to be disgorged, namely the three-step test and “but for” causation, and Rowe J gave no reason why one rather than the other is used in different contexts.

Nova v Dow: Summary of the Summary

Nova Chemicals Corp v Dow Chemical Co 2022 SCC 43 Rowe J: Wagner CJ, Moldaver, Karakatsanis, Brown, Martin, Kasirer and Jamal JJ concurring; Côté J dissenting affg Nova Chemicals Corporation v Dow Chemical Company 2020 FCA 141 Stratas JA: Near, Woods JJA affg Dow Chemical Co v Nova Chemicals Corp 2017 FC 350, 2017 FC 637 Fothergill J

2,160,705 / film-grade polymers / ELITE SURPASS

In Nova v Dow, Rowe J, writing for an 8-1 majority, addressed the proper method of calculating an accounting of profits in the patent context. I had started a series of posts summarizing my forthcoming two-part article on Nova v Dow which will appear in the IPJ, but I got distracted over the summer. The series resumes with this post and will continue on Fridays.

My article is very long, so I will start by emphasizing my conclusion: the Nova v Dow decision is so bad, both doctrinally and in policy terms, that the accounting remedy has been fatally damaged. If an accounting continues to be granted routinely Nova v Dow will almost certainly have a chilling effect on innovation and legitimate competition. There is no way to interpret Nova v Dow that would restore the remedy to a sound basis. The best response to Nova v Dow is therefore to abolish the accounting remedy entirely. That may not be practical, as it would require legislative intervention. The next best response is to restrict the accounting remedy to cases in which the infringer was trying to game the system by avoiding ex ante licensing, ie was engaging in holdout. Restricting the availability of an accounting is consistent with the equitable nature of the accounting remedy; it is consistent with the existing caselaw on the entitlement to an accounting; it is consistent with Nova v Dow, which did not address entitlement to an accounting; and it is supported by the extra-judicial writing of Edelman J of the High Court of Australia.

My previous summary posts are also very long, so in this post I will recap the facts and provide a summary of my summaries.

The Facts

Dow had a patent on specialized plastic film, used for items such as food packaging. Nova made and sold a competing film that was found to be infringing. Dow was awarded an accounting of Nova’s profits from sales of infringing film, with a quantum of $644 million. The major input to the infringing plastic is ethylene, an unpatented bulk commodity. Nova had a very efficient process for making ethylene—the “Alberta Advantage”—and so could make ethylene for far less than it would have cost to buy on the open market. On the facts, if Nova had not infringed, it would have used its ethylene to make commodity grade “pail and crate” plastic [FC 158]. Nova argued that because of the Alberta Advantage, it would have made approximately $300 million in profit on the sale of that plastic, despite the competitive nature of that market [FCA 187]. The question was whether Nova was entitled to deduct the amount it would have made in the pail and crate market, on the view that that part of the overall profit was caused by its efficient ethylene production process and not by the infringement.

Rowe J set out a three-step test to “conceptualize” an accounting of profits [15]:

Step 1: Calculate the actual profits earned by selling the infringing product — i.e., revenue minus (full or differential) costs.

Step 2: Determine whether there is a non-infringing option that can help isolate the profits causally attributable to the invention from the portion of the infringer’s profits not causally attributable to the invention — i.e., differential profits. It is at this step that judges should apply the principles of causation. . . .

Step 3: If there is a non-infringing option, subtract the profits the infringer could have made had it used the non-infringing option from its actual profits, to determine the amount to be disgorged.

On the facts, Rowe J held that Nova was not entitled to deduct the amount it would have made in the commodity plastic market, because commodity plastic was not a “non-infringing alternative” (“NIO”) under Step 2, so the entire actual profits were to be deducted.

Rowe J stated that it is a "fundamental principle" that the infringer must disgorge "all profits causally attributable to infringement of the invention" [4] and at the same time he emphasized that the profits to be disgorged in an accounting are "only the profits causally attributable to the invention" [48] (original emphasis). Rowe J repeated the need for a causal connection more than two dozen times.

The Intuition

A key question is what constitutes an appropriate non-infringing option at Step 2. One view, taken by Nova and Côté J, is that the NIO reflects the application of “but for” causation, so that the amount to be disgorged is the difference between the infringer’s actual position and the position the infringer would have been in but for the infringement. On this view, the profit to be disgorged is the profit caused by the infringement, which takes into account both the value contributed by the invention and the value contributed by the infringer. This means that profits that the infringer might have made in other unrelated markets—such as the market for commodity plastic—are deducted because if the infringer could have earned excess profits in an unrelated market this reflects the unique value contributed by the infringer in consequence of its special expertise or manufacturing advantage.

The other view, taken by Rowe J for the majority, focuses on profits caused by the invention as opposed to profits caused by the infringement. The intuition seems to be that an accounting reflects the value of the invention, as opposed to the value of the infringement, and the value of the invention cannot depend on the vagaries of unrelated markets such as the market for commodity plastic. On that view, there is something special about certain non-infringing options such that a comparison with an appropriate NIO reveals the value of the invention itself. Whether something is an appropriate NIO does not simply turn on what the infringer would have done but for the infringement; the NIO turns on the nature of the invention, not on the what the infringer might have done with it. On this view, the value of the invention cannot depend on what happens in entirely unrelated markets.

The Legal Background

The prior leading case on an accounting was Schmeiser 2004 SCC 34, in which the SCC held that the preferred method for calculating an accounting of profits is the “differential profit” approach, in which “[a] comparison is to be made between the defendant’s profit attributable to the invention and his profit had he used the best non-infringing option” [102]. The subsequent FCA caselaw held that the same approach applies in the context of damages and elaborated a “could and would” test for identifying an appropriate NIO. In my view, these cases support the view that the differential profits approach is simply the reflection of “but for” causation. However, there were a couple of passages suggesting that an NIO must be considered “a true substitute by the consumer”: see Cefaclor Damages 2018 FCA 217 [54]. Based on this authority, Dow argued that an appropriate NIO had to be a true market substitute for the infringing product. Accordingly, Dow argued that the profits Nova would have made in the commodity plastic market could not be deducted because it was undisputed that commodity plastic is not a market substitute for the specialized patented plastic. While Rowe J held in Dow’s favour, he explicitly rejected the market substitute test for an NIO [67]. (The market substitute test was not raised in the FCA.)

Causation as a Matter of Fact

Rowe J stated more than two dozen times that there must be a causal link to the profits to be disgorged. The most prominent causation concept in law is “but for” causation and both Nova and Côté J, dissenting, took the position that “but for” causation was the appropriate causation concept. Remarkably, despite his emphasis on the importance of causation, Rowe J never identified the legal causation concept at all; instead he asserted that causation is “a practical question of fact which can best be answered by ordinary common sense” [15] and similarly, “[w]hether there is a non-infringing option that can assist courts in isolating the profits causally attributable to the invention is a question of fact” [67]. This is a remarkable novel proposition; there are many well-known SCC decisions in which the central question of law was as to the nature of the correct causation concept, including decisions which overruled the trial judge on a question of law for having applied the wrong causation concept. The cases cited by Rowe J for the proposition that causation is a matter of fact actually stand for the proposition that applying the legal causation test to the facts requires common sense.

The Concession

While Rowe J refused to articulate any causation concept, we might have expected some insight as to the principle at work from the application of his framework to the facts. Nova had argued that had it not infringed, it would have made approximately $300m in profits in the market for commodity plastic. There was no finding at trial that Nova would in fact have made such profits, as the trial judge, affirmed by the Court of Appeal, had refused to allow the deduction as a matter of law. Since Rowe J affirmed, one might normally infer that any profits from the commodity market cannot be deducted as a matter of law. However, Nova had conceded at trial that there were no non-infringing options available for the purpose of applying the differential profits approach [70]. Relying on this concession Rowe J held that there was nothing to be deducted at Step 2 and so Nova was required to disgorge its entire actual profits. Thus, while the FC and FCA had treated deductability of the commodity plastic profits as a matter of law, the SCC affirmed on the basis of a concession of fact. As a result, Rowe J did not address the nature of the causation requirement.

What Role for “But For” Causation in Identifying the NIO?

A central part of Rowe J’s three-step approach is the identification of the NIO at Step 2. The closest Rowe J came to providing a definition of the NIO is that it is “any product that helps courts isolate the profits causally attributable to the invention from the profits which arose at the same time the infringing product was used or sold, but which are not causally attributable to the invention” [58]. But Rowe J did not identify the causation concept, nor did he define or even discuss what it means for profits to be “causally attributable to the invention.” Combined with his holding that whether something is an NIO is a matter of fact, this means that an NIO is anything that the trial judge finds helpful. This implies that it remains open to a trial judge to apply “but for” causation to identify the NIO at Step 2, or as one consideration in identifying the NIO, if the trial judge is of the view that doing so would be helpful. Indeed, at one point Rowe J implied that commodity grade plastic might have been an appropriate NIO if the trial judge had made a factual finding to that effect [72]. At the same time, it is clear that the NIO is not determined solely by finding what the infringer would in fact have done but for the infringement; if that were the case, Nova would have won.

Column IV Costs for Summary Trial

Eli Lilly Canada Inc v Mylan Pharmaceuticals ULC 2023 FC 780 St-Louis J

2,226,784 / tadalafil / CIALIS / NOC

There are a couple of points worth noting in this costs decision following from the summary trial in Lilly v Teva 2022 FC 1398*, in which the defendants, Apotex, Mylan, Teva and Pharmascience, prevailed: see FC “Or” And FC Overbreadth. Only Mylan and Lilly filed written submissions in regards to costs [1].

The first point is that while Rule 407 provides that costs are normally assessed on column III, the jurisprudence has established that the upper end of column IV is often appropriate in patent trials due to their greater than average complexity. Lilly argued that notwithstanding this caselaw, column III should apply in respect of summary trials. St-Louis J rejected this submission: “Accepting Lilly’s argument of some sort of a norm being fixed at column III for summary trials in intellectual property would entail, in addition to encroaching on the Court’s discretion, assuming that summary trials in intellectual property are, by default, of average complexity. While summary trials may mitigate the considerations highlighted by the Chief Justice in Allergan [2021 FC 186], I have not been convinced that they necessarily or always completely diminish the complexity of a pharmaceutical patent litigation to the point that these 400(3) factors cannot be considered and found to be present” [37].

Secondly, St-Louis J confirmed that an offer to settle that does not meet the conditions of Rule 420 may nonetheless be considered as one factor in awarding costs under Rule 400. This point is by now well-established [32]. (In this case, Mylan’s offer did not meet the timing conditions in subsection 420(3)—ie presumably it was made less than two weeks prior to trial.)

(I will have a post on the FCA decision in Benjamin Moore 2023 FCA 168, addressing the difficult issue of patentable subjet-matter, but the post may be delayed as I am teaching a new course in the fall and I have an article deadline coming up.)

*For some reason, the trial decision is still not available on the FC website, so I’ve linked to a copy made available by Smart & Biggar.

Are Successive NOAs Permitted in Respect of the Same Patent?

Janssen Inc v Apotex Inc 2023 FC 912 Manson J

2,655,335 / paliperidone regimens / INVEGA SUSTENNA / NOC

This summary judgment motion raises a tricky question on the application of res judicata / abuse of process. Apotex sought to launch a generic form of paliperidone and had to address Janssen’s 335 patent. In its initial NOA, Apotex alleged it would not infringe; it did not allege invalidity [12]. That led to a summary trial in which Apotex alleged non-infringement without challenging the validity of the 335 patent. Apotex lost, but an appeal is ongoing: see Janssen v Apotex 2022 FC 107 (see here) [14]–[16]. Apotex then amended its NOA to allege invalidity, without alleging non-infringement [17]. Janssen brought this motion for summary judgment, on the basis that Apotex was precluded from raising the defence of invalidity, on the basis of res judicata or abuse of process.

Manson J held that the issue was suitable for summary judgment as it turned solely on the application of the law to settled facts [38]. However, Manson J dismissed the motion. His key holding was that

[69] The Regulations, when viewed contextually and purposively, do not preclude successive dual NOAs in respect of the same patent, one alleging non-infringement and another alleging invalidity.

There are two branches to res judicata, namely cause of action estoppel and issue estoppel. Manson J held that neither of these was engaged.

Both branches require that the prior decision is final [42], [52]. Manson J noted that there is conflicting caselaw on whether a decision is final when there is a pending appeal, but he held that the weight of authority is to the effect that a decision is not final if an appeal is pending [48]–[51]. It was not disputed that the appeal in Janssen v Apotex 2022 FC 107 is pending, and consequently Manson J held that the matter is not res judicata [51], [53].

I must say that I don’t find the authorities relied on by Manson J to be strong support for his conclusion that a decision is not final if an appeal is pending. He quoted the statement in CUPE 2003 SCC 63 [46] that “[a] decision is final and binding on the parties only when all available reviews have been exhausted or abandoned.” But the thrust of that paragraph as a whole is that a party who is dissatisfied with a decision can attack it through an appeal or judicial review, but it is an abuse of process to attempt to impeach a decision in a different forum. This could just as well support the view that the only mechanism for attacking a decision is by way of appeal or judicial review. Manson J also relied on Lilly v Teva 2018 FCA 53 [83]. That case held that a decision was final even though there was no opportunity for appeal because the issue was moot. This is entirely consistent with saying that the only mechanism for attack is by way of appeal. In both these cases, the prior decision was held to be final. Finally, Manson J relied on Canada v MacDonald 2021 FCA 6 [15], in which Stratas JA remarked that “If an appeal is brought, the appeal court can interfere with the order or judgment. Thus, an order or judgment under appeal is not final for the purposes of the doctrine of res judicata.” Again, this is consistent with saying that the only mechanism for attack is by way of appeal. As I read it, Stratas JA was saying that taking an appeal is not precluded by res judicata, but this does not mean that it is not final for other purposes. Indeed, Stratas JA held that res judicata did apply to bar re-litigation on the facts [24] (albeit in obiter, as the matter was disposed of on other grounds [12]).

I am not saying that Manson J is necessarily wrong in his conclusion that a decision is not final for the purposes of res judicata if it is under appeal, but I would suggest that the point is not settled by the authorities he referred to. In my view, those cases stand primarily for the proposition that any attack on a decision other than by way of appeal or judicial review is an abuse of process.

While Manson J would have dismissed the issue estoppel argument on finality alone, he also found that the issue of infringement is not the same as the issue of validity, so that branch of the test for issue estoppel was not satisfied either [57]–[63]. For issue estoppel, it is not enough that the argument might have been raised; it must actually have been raised and decided (albeit possibly implicitly).

This left abuse of process, which is a supplementary doctrine that “has applied to bar needless litigation that may not meet the technicalities of issue estoppel or cause of action estoppel” [65]. There is no test as such for abuse of process, which turns on the court’s assessment of whether the proceedings are oppressive or vexatious: CUPE 2003 SCC 63 [35]–[43].

Janssen’s main argument was that:

[66] Apotex “testing a non-infringement allegation before alleging invalidity” constitutes an abuse of process. Janssen argues that allowing these actions to continue would lead to the absurd result that generic drug manufacturers could deliver one NOA making just one allegation, fully litigate that issue, only to then deliver another NOA with a fresh allegation in respect of the same patent and drug.

I must say that strikes me as a strong purposive argument. If the PM(NOC) system did not exist and a generic brought a declaratory action alleging non-infringement in order to “clear the way,” I think it would have to be an abuse of process to bring a section declaratory action alleging invalidity if the non-infringement action failed.

Now, this is not a declaratory action, so the text and context of the PM(NOC) Regulations are crucial [68]. Manson J stated that “Nowhere in the text of the Regulations does it state that a [generic] can serve only one NOA in respect of each patent” [70]. That’s true, but nether it say that multiple NOAs are permitted; the Regs simply do not directly address this situation. He continued “On the contrary, the Regulations require that an NOA is served on a patentee each time an allegation is made leading to, in some cases, successive NOAs and successive litigation” [70]. I’m not sure what examples he had in mind. To the extent that a generic serves successive NOAs because it is seeking ANDS for different products, that is clearly not an abuse of process, but in such a case the underlying facts are different, because the products are different, so res judicata or abuse of process is clearly not engaged. That is going on here.

Manson J relied heavily on the fact that allegations of invalidity and non-infringement are contemplated in different subparagraphs of the Regs, namely 5(2.1)(c)(ii) and (iv) [72]–[73]. I have difficulty reading as much into this as did Manson J. It looks to me like this paragraph is simply trying to set out an exhaustive list of allegations in a clear format. Manson J placed considerable emphasis on the fact that invalidity and non-infringement found in separate paragraphs, noting that

[72] There is, however, no such distinction between individual grounds of invalidity. If a generic drug manufacturer were attempting to serve a second NOA in order to pursue different grounds of invalidity it did not pursue in previous litigation that would almost certainly be abusive of the process set out under the Regulations.

This strikes me as being somewhat formalistic. Yes, they are in two separate paragraphs. But what function or purpose is served by allowing separate allegations of non-infringement and invalidity, while prohibiting separate allegations of different grounds of invalidity? Invalidity attacks are not generally more closely linked than validity and infringement; eg an attack based on patentable subject matter will have little evidence or argument related to an obviousness attack.

Janssen also relied on the statement in the 2017 RIAS that “[the amendments eliminate] the need for separate proceedings to address all claims in a single patent” [66]. Manson J dismissed this on the basis that “the new Regulations were implemented to avoid the dual track litigation that occurred under the old Regulations where limited issues of patent validity or infringement were addressed summarily via applications for prohibition orders under the Regulations, and then re-litigated as actions for infringement under the Patent Act” [75]. No doubt that was one specific objective of the new Regs, but the amendments were not limited to curing that problem: the RIAS states that “[a]s a general statement, the amendments seek to facilitate the just and expeditious resolution of patent infringement and validity disputes under the Regulations.” The need for an efficient and expeditious resolution of the disputes is emphasized throughout in a variety of contexts.

Moreover, the 2017 RIAS contemplates that non-infringement and invalidity will normally be litigated in the same proceeding. So, when the patentee brings an infringement action, “[i]n addition to defending such an action, a second person may commence a counterclaim seeking to invalidate the patent.” This is directly reflected in s 6(3)(a), which contemplates that when a patentee brings an action for infringement in respect to the NOA, the generic may respond with a validity attack. The RIAS also remarks that because the proceedings “will focus on underlying questions of patent validity and infringement” mootness is no longer an issue. The RIAS explains that “[n]on-infringement allegations need not be as detailed as invalidity allegations; a second person will be free to choose how detailed a non-infringement allegation may be. However, the second person is required to serve, along with its NOA, any portions of its submission or supplement that could be relevant for determining whether a listed patent would be infringed. The second person must comply with this requirement even if it makes no allegation of non-infringement.”

Overall, I would say that purposive and contextual factors relating to the desirability of promoting efficient and expeditious resolution of the proceedings weigh quite strongly against permitting separate allegations of invalidity and non-infringement. A sufficiently clear text will dominate purposive and contextual considerations, but the only textual point invoked by Manson J was that that invalidity and non-infringement allegations are mentioned in separate subparagraphs of 5(2.1)(c); this strikes me as a fairly weak textual point, which is not sufficiently clear to warrant an interpretation contrary to the purposive and contextual emphasis on streamlining the litigation process.

The issue will be slightly different if an appeal had been decided in the earlier proceeding, or if the time for taking an appeal had passed. If a generic argues non-infringement, and, for example, does not appeal, can it then allege invalidity in a new NOA? According to Manson J, doing so would not be an abuse of process, as nothing in his abuse of process analysis turns on the finality of the prior decision. This strikes me as counter-intuitive. Perhaps the answer is that in such a case cause of action estoppel might apply. In that scenario, the finality requirement would not be an obstacle, so the question would be whether an allegation in an NOA is a “cause of action.” The answer isn’t entirely clear, because strictly the cause of action is for infringement, brought by the patentee in response to the NOA. But given the emphasis in the 2017 amendments on efficient dispute resolution, I find it very difficult to believe that the generic would be permitted to file successive NOAs alleging that the patentee is not in fact the owner or licensee, then after losing in a final decision, and then another alleging invalidity, and another alleging that the patent at issue was not properly listed, and yet another alleging non-infringement. If that is not permitted if the prior decision is final, I find it difficult to see why it should be permitted simply because the prior decision is under appeal.

Misleading the Court as a Basis for Punitive Damages

Fromfroid SA v 1048547 Ontario Inc 2023 FC 925 Grammond J

2,301,753 / rapid cooling cells

This decision is of interest primarily for the award of punitive damages, granted because the infringer had intentionally infringed the patent rather than buying the patented system from the patentee in order to save money and then tried to cover up the infringement by presenting misleading evidence to the court [106]–[107]. As a matter of law, it is interesting that attempting to mislead the court was used as one factor—and for one defendant, the only factor—for awarding punitive damages, rather than sanctioning that conduct through elevated costs. There is precedent for this in the trademark context: see Chanel v Lam Chan Kee Co 2016 FC 987 [76] affd 2017 FCA 38 [11].

Fromfroid’s 753 patent relates to a cooling system for food products. The system consists of “a tunnel lined with tarpaulins that inflate with air blown by a fan. When they inflate, the tarpaulins press against the pallet of products to be refrigerated, which is placed in the centre of the tunnel. In so doing, they force the stream of cold air to pass between the containers of products instead of going around the pallet, which ensures more rapid cooling” [3]. The complete system consists of a number of “cells” each with its own fan.

The defendant 1048547 Ontario Inc dba Skotidakis, operates a dairy products business [6]. Skotidakis was interested in acquiring a cooling system for its dairy products. In 2014, Skotidakis obtained a trial cell from Geosaf, Fromfroid’s agent, for a period of about six weeks. Fromfroid then submitted a bid for a 24-cell system, but Skotidakis did not follow up [6]. In October of 2018, approximately three months after the 753 patent expired, a representative of Geosaf attended Skotidakis’s facilities in order to service unrelated equipment. The Geosaf representative saw and photographed a 24-cell cooling system that bore a strong resemblance to the patented system [7]. The only live issue at trial was whether the system was built before or after the expiry of the patents [10]. (There was a feeble attempt to argue non-infringement on the basis that the tarps did not press against the sides of the pallet, but Grammond J had no difficulty dismissing this on the facts [74]–[93].)

The difficulty for Fromfroid was that it had no direct evidence of when the system was built. Instead, it had to rely on indirect evidence, such as wear marks apparent in the photograph taken by the Geosaf representative and the difficulty of constructing a system in only 3 months [71].

Against this circumstantial evidence was the direct testimony of four persons, all more or less closely associated with Skotidakis, who stated that, “to their personal knowledge, the cells were built [in the three month period after expiry], particularly since this testimony is supported by documentary evidence” [49]. As counsel for the defendants put it, “[a]ccepting Fromfroid’s position would entail that these four witnesses lied to this Court and that the documents they submitted are forgeries” [49].

While Grammond J did not explicitly find that the witnesses lied or that the documents were forgeries, he did reject their evidence as being defensive and unresponsive [52], deliberately evasive [53], “implausible” [57] and imprecise [59]. He found that while the documents Skotidakis tendered in evidence “show no obvious signs of falsification, there is also no basis for excluding the possibility that they are forged, that the date on them could have been altered (e.g., by replacing ‘2016' with ‘2018') or that they refer to a different project” [61].

With little weight given to the defendants’ evidence, Grammond J found that the circumstantial evidence was sufficient to establish “a serious, precise and concordant presumption” (per CCQ Art 2849) that the cells were built before the 753 patent expired [71]. While Grammond J did not expressly find that the Skotidakis witnesseses had lied or that the documents were forged, he did find that “Skotidakis sought to conceal the infringement by presenting various pieces of evidence intended to mislead the Court as to the date the cells were made” [106].

Grammond J awarded lost profit damages on the basis that but for the infringement, Skotidakis would have purchased its system from Fromfroid [97]–[102].

He also awarded punitive damages, on the basis of two factors. First, he found that Skotidakis knew of the existence of the patent and infringed knowingly [105]. This alone would not be sufficient to justify punitive damages [105], but as an additional factor, Grammond J found that in attempting to mislead the Court, Skotidakis had engaged in “highly reprehensible misconduct” [106]. He stated that “the attempt to mislead this Court is serious and should be severely denounced” [109]. Accordingly, he ordered Skotidakis to pay $200,000 in punitive damages, on a top of $150k in compensatory damages [102], [111].

Grammond J also awarded punitive damages against the defendant, Frimasco, a construction company that had built the system on behalf of Skotidakis, even though Frimasco was not aware of the patent at the time it constructed the system [112]. The sole basis for awarding punitive damages against Frimasco was its role in concealing the infringement [112]. While Frimasco did not benefit directly from the infringement, “Skotidakis and Frimasco have a close and longstanding business relationship” [37]. Mr. Robert, the president of Frimasco at the time, gave implausible testimony relating to the purchase of parts for the system [36] and as to the construction speed [47], and as to the nature of the system when testifying regarding infringement [78]. Grammond J found that Frimasco’s “involvement in the cover-up is just as reprehensible as that of Skotidakis” [113]. (Note there are a couple of typos in paras [112], [113], where Grammond J refers to the patentee, Fromfroid, when he evidently intends to refer to the defendant Frimasco.)

It is interesting to see such conduct used as a factor—and in the case of Frimasco, the sole factor—supporting an award of punitive damages. I would have expected to see some kind of costs sanctions instead. But I’m not an expert in sanctions for giving misleading evidence or litigation misconduct generally. Chanel v Lam Chan Kee Co 2016 FC 987 [76] affd 2017 FCA 38 [11], cited by Grammond J at [106], does use attempts to mislead the court as a factor in awarding punitive damages (though in Chanel the attempt to mislead the court was only one of nine factors supporting the award of punitive damages). Further, it is established in Ontario law that litigation misconduct can be an independent actionable wrong that could give rise to punitive damages: McCabe v Roman Catholic Episcopal Corp 2019 ONCA 213 [43]–[48]. At the same time, litigation misconduct can also be sanctioned by costs: Render v ThyssenKrupp 2022 ONCA 310 [87]–[91]. Costs submissions remain to be made in this case [121]. It will be interesting to see if Grammond J takes the misconduct into account in awarding costs, or whether that would be considered double counting. Absent the award of punitive damages, I would be inclined to think that a very substantial costs award would be appropriate, well beyond 40% lump sum which is the higher end of the usual range, and approaching full indemnity. It is apparent from Grammond J’s reasons that Skotidakis did not have any good faith defence and should have settled rather than raising a defence premised on trying to mislead the court.