Tuesday, October 6, 2020

Nova v Dow: A Radical Departure from Established Law

Nova Chemicals Corporation v Dow Chemicals Company 2020 FCA 141 Stratas JA: Near JA / Woods JJA dissenting, aff’g 2017 FC 350, 2017 FC 637 Fothergill J

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

The decision of Stratas JA for the majority in Nova v Dow raises a panoply of important conceptual and practical issues related to an accounting of profits, and non-punitive monetary remedies more broadly: whether the test for causation is the same in an accounting and damages; whether an accounting should provide a disincentive to infringe; the nature of the non-infringing alternative in the “but for” world; whether the hypothetical world should be constructed on subjective or objective terms; the role of opportunity costs; and the deductibility of fixed costs. And this is not to mention other less foundational but nonetheless significant issues of disgorgement of springboard profits and the date for currency conversion. Unfortunately, this decision is not likely to settle any these issues; on the contrary, it is likely to increase uncertainty in this entire area of law.

In a series of decisions, following from the seminal decision of the SCC in Schmeiser, 2004 SCC 34, the FCA has built up what I consider to be a sound and principled approach to non-punitive remedies: see Lovastatin Damages 2015 FCA 171; Venlafaxine s8 2016 FCA 161; Perindopril Accounting #1 2017 FCA 23; Cefaclor Damages 2018 FCA 217; Perindopril Accounting #2 2020 FCA 60. While some important loose ends remain outstanding, the foundations were solid. Stratas JA’s decision for the majority in Nova v Dow has thrown a grenade into that edifice and rattled it down to those foundations. The general statements of principle set out by Stratas JA constitute, in my view, a radical and unsound departure from established principles, including those set out by the SCC in Schmeiser. Some specific holdings might be seen as refinements of prior law, but even then, because of the unorthodox reasoning, it is not clear whether even these points can now be considered settled.

A modified version of the facts illustrates the main issue. Ethylene is a feedstock for a variety of products. It is not patented, and is sold on the market as a bulk chemical. Mova (I’ll change the names since the facts are not exact) makes ethylene in a very efficient plant in Alberta, so that it has substantially lower costs than other ethylene producers. This is the “Alberta Advantage.” Historically, Mova used its ethylene feedstock to make commodity grade “pail and crate” plastic. The market for pail and crate plastic is highly competitive, and it all sells for the same price. For most producers, the economic profit is zero. Mova sells for the same price as everyone else, but it makes a supra-competitive profit, Pa, equal to the Alberta Advantage, due to its uniquely low cost of production.

After many years of selling pail and crate plastic, Mova developed a new product, a specialty film-grade polymer—plastic wrap— in the class of metallocene linear low-density polyethylene or “mLLDPE.” mLLDPE is also made from ethylene feedstock. The market for mLLDPE is not very competitive—there is only one other producer, Dov—so it is much more profitable. Mova switches all its ethylene capacity from pail and crate production to mLLDPE. The cost of production is the same—primarily the cost of the ethylene feedstock—but prices for mLLDPE are much higher than for pail and crate plastic. When Mova enters the market, it matches Dov’s price. Because its costs are the same, Mova makes profits, Pm, that are substantially higher than previously.

Unfortunately for Mova, it turns out that Dov holds a patent on mLLDPE. Dov sues Mova for infringement, prevails, and elects an accounting of profits.

It is uncontroversial that the purpose of the causation requirement is to ensure that the profits disgorged are only those which are causally attributable to the invention: Schmeiser [101]; Perindopril Accounting #1 2017 FCA 23 [28]; Nova v Dow [46], [80]. The correct approach to causation in an accounting was set out by the SCC in Schmeiser: “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]. This can be broken down into three steps:

(1) determine the infringer’s actual profits earned in the real world;

(2) determine what the infringer’s profit would have been in the hypothetical world in which it used the best non-infringing option;

(3) subtract (2) from (1) to arrive at the profits to be disgorged

Mova’s actual profits are Pm. Had it not infringed, its best non-infringing option would have been to go back to pail and crate plastic, and it would have made Pa. The difference is Pm - Pa. That is, the profits to be disgorged are the excess profits from selling patented mLLDPE, less the Alberta Advantage in ethylene production.

This result seems intuitively correct, at least to me. I trust that it is clear that the extra profit, Pa, that Mova was making before it started making infringing product is not due to the infringement, but rather to its efficient ethylene plant. As I see it, that remains true when Mova starts using its efficient plant to make feedstock for infringing product.

We can also look at the issue from Dov’s perspective. As the FCA noted in Perindopril Accounting #1 2017 FCA 23 [28] “the value of a patent lies in the ability of the patentee to exclude competitors and competition.” So, the value of the patent is the profit made by Dov due to its ability to exclude others from the mLLDPE market. What were Dov’s profits before Mova entered the market? Prior to Mova’s entry, Dov sold at the same price as Mova did when it entered, but Dov is not a uniquely efficient producer of ethylene, so its costs are higher by an amount equal to the Alberta Advantage (which, recall, is equal to Mova’s profit in the bulk ethylene market). So Dov’s profit prior to Mova’s entry was Pm - Pa. This confirms the result, flowing from the application of the Schmeiser test, that the profits caused by the invention are Pm - Pa.*

However, it seems clear that on these facts that Stratas JA would require Mova to disgorge the entire actual profits, Pm. This is because Stratas JA did not apply the Schmeiser test. He held instead that “In an accounting of profits, courts must work in the real world, not the hypothetical. Courts care only about actual revenues and actual costs. What ‘could have’, ‘should have’, or ‘would have’ happened is of no moment” [37]. In effect, he eliminated part (2) of the Schmeiser test. This is very surprising, given the authority of the SCC in Schmeiser, which has, moreover, been affirmed and applied in the series of FCA decisions noted above. In the remainder of this post I support this assertion in more detail, by showing that Stratas JA’s decision departs from established law both in statements of principle and in the result.

When a court departs from established law, it is usually from a sense that the law will not do justice on the facts. Why did Stratas JA feel that it was necessary to depart from the established test in this case? My sense is that there were two considerations driving his decision.

First, Stratas JA emphasized throughout that an accounting must disgorge only the profits caused by the infringement: [37], [47], [48], [50], [80]. He evidently had a strong intuition that in this case, Nova’s profits from its efficient ethylene plant were caused by the infringement, and that intuition was sufficiently strong to lead him to depart from established law. As just discussed, I do not share that intuition. I can only speculate as to why Stratas JA felt the way he did, but I suspect it was due to the particular way in which the case was argued. As I will discuss below, Nova apparently did not argue the three-step Schmeiser test outlined above. Rather, it ran together the first and second steps. This was perhaps understandable, but with hindsight it was an unfortunate choice, as its approach was wrong in law and, perhaps worse, it concealed the true issues.

Second, Stratas JA was strongly and explicitly of the view that an accounting of profits should provide a disincentive to infringement. If Mova is entitled to keep the profits from its efficient operation, it will be no worse off as a result of the infringement, and there would be, according to Stratas JA, no incentive not to infringe. While I understand this perspective, it is, in my view, contrary to established law and wrong as a matter of policy. It is well-established that damages aim for “perfect compensation”: the plaintiff is restored to the position it would have been in but for the infringement, no better, and no worse: Lovastatin Damages 2015 FCA 171 [42]. An accounting is also non-punitive; it should restore the infringer to the position it would have been in but for the infringement: no better and no worse. So I agree with Stratas JA that an accounting, administered on the Schmeiser test does not provide an incentive not to infringe; rather, it leaves the infringer indifferent between having infringed and not. Where we part company is that I believe this is sound policy, whereas he evidently does not.

Principles

Now to the details, starting with the principles. The established law relating to an accounting of profits was summarized by the FCA in Perindopril Accounting #1 2017 FCA 23 (my emphasis)

[40] I begin my analysis with the observation that when considering the availability of non-infringing alternatives one enters the world of the hypothetical – in the real world the defendant used an infringing product. This Court considered the nature of the hypothetical world in Lovastatin. Notwithstanding that Lovastatin concerned a claim for compensatory damages, not an accounting of profits, again I believe that the Court’s commentary in Lovastatin has equal application to this case. In both situations the Court is to consider a hypothetical world where the infringing conduct did not take place.

There are two key points: first, the accounting inquiry involves a comparison between the real world and a hypothetical world; second, an accounting is like damages in this respect.

The FCA in Nova v Dow appears to have repudiated both of these foundational principles:

[40] In order to do this properly in an accounting of profits, courts must avoid the hypothetical, “but for” world. Questions of what the parties could, would, or should have done are irrelevant to the analysis.

[76] The use of hypotheticals and the “but for” test. . . has no place in an accounting of profits.

The irrelevance of the hypothetical world is stressed throughout: see [37], [41], [65], [67], [78], [94], [101], [102], [157].

The importance of the hypothetical world is so well entrenched in the remedial context that Stratas JA could not reject it entirely. Instead, he held that it is irrelevant in an accounting, and that there is is a difference between damages and an accounting in this respect:

[45] To be clear, ‘but for’, hypothetical reasoning applies when courts award compensatory damages for patent infringement. . . . But this is not the case in an accounting of profits. . . .

See similarly [67], [76].

Stratas JA cited no authority for the proposition that an accounting and damages are different in this respect, no doubt because there is none. The prior FCA cases consistently hold that the same principles are applicable in both contexts: see Perindopril Accounting #1 2017 FCA 23 [42] (approving the ‘could have’ and ‘would have’ test in the accounting context); Cefaclor Damages 2018 FCA 217 [47] (noting that consideration of the non-infringing alternative “was based on general principles of Canadian common law”); Perindopril Accounting #2 [42]-[60] (relying on the damages decisions in Cefaclor, Venlafaxine and Lovastatin as stating the principles applicable in the accounting context); Lovastatin Damages 2015 FCA 171, [60] (holding that the principles relating to a non-infringing alternative set out in Schmeiser, an accounting case, are also applicable in assessing damages). Indeed, in Lovastatin Damages, the FCA, relying on Schmeiser, an accounting case, mandated the use of a hypothetical world in assessing damages precisely because doing so in the accounting context was established by the SCC in Schmeiser, and there is no principled difference between an accounting and damages in this respect: Lovastatin Damages 2015 FCA 171. [51], [59-60]; see also Cefaclor Damages 2018 FCA 217 [47] (noting that the use of the non-infringing alternative analysis in the damages context was based on general principles as applied in Schmeiser). To assert that a hypothetical world is relevant only to damages, and not to an accounting, is to rewrite history.

Moreover, in the Perindopril litigation the FCA explicitly endorsed the use of a hypothetical reasoning on the facts, in the context of an accounting. The defendants, Apotex, had made infringing product in Canada and exported it. They argued that in the hypothetical world, they would have had the product manufactured abroad and would therefore have made the same profits (or even more) even if they had not infringed in Canada: Perindopril Accounting#1 FC 2015 FC 721, [79-80]. The trial judge rejected this argument [121], and the FCA reversed and remanded to the trial judge for reconsideration (my emphasis):

[65] For clarity, the issue I would remit to the Federal Court is whether Apotex would have and could have obtained quantities of non-infringing perindopril from any of [potential foreign suppliers] and, if so, whether Apotex would have and could have used non-infringing perindopril for sales to its affiliates in the United Kingdom and Australia.

This is not consistent with the view expressed in Nova v Dow [45] that “[w]hat ‘would have’ and ‘could have’ happened does not matter” in an accounting of profits.

On remand in Perindopril, the trial judge found that in the “but for” world, the defendant would have chosen to manufacture abroad, but only after a two-year delay to allow it to source from its own foreign affiliate rather than a third party: 2018 FC 346 aff’d 2020 FCA 60. This is a clear example of hypothetical reasoning applied in the context of an accounting of profits. On the facts, the two-year delay meant that Apotex would not have been in a position to supply the foreign market until after the expiry of the patent, so that Apotex was not entitled to make any deduction from its actual profits for this non-infringing alternative. Nonetheless, it is abundantly clear from Perindopril #1 FCA that the FCA would have been willing to allow a deduction had the foreign supply been available earlier in the “but for” world: that is exactly why the FCA remanded the issue to the trial judge. As the FCA stated in Perindopril #2 FCA (my emphasis):

[49] Of necessity, this is a hypothetical exercise, as the Court must reconstruct the market to determine what would have happened in the hypothetical situation where, rather than infringing a plaintiff’s patent, the infringer instead chose to compete with the plaintiff using a non-infringing product.

And then of course there is Schmeiser, the foundational case establishing the need to consider a hypothetical world in assessing non-punitive remedies. Schmeiser was an accounting case. As Stratas JA himself explained (my emphasis), in Schmeiser “[t]he Supreme Court awarded no profit because the infringers’ ‘profits were precisely what they would have been had they planted and harvested ordinary canola’ and, therefore, the infringers ‘obtained no premium’ nor ‘gain[ed] any agricultural advantage’ from the patented seeds.” On its face, this is hypothetical reasoning, looking to what the profits “would have been.”

Rather than providing authority, Stratas JA justified his view by saying that “In an accounting of profits, courts must work in the real world, not the hypothetical. Courts care only about actual revenues and actual costs. What ‘could have’, ‘should have’, or ‘would have’ happened is of no moment. The only thing that matters is what did happen” [37, original emphasis]. Stratas JA’s statement echoes the words of Lord Halsbury in United Horse Shoe (1888), LR 13 AC 401, 409 (HL): “what does it matter if it is ever so much established that the loss which the pursuers have sustained by the unlawful act of the defenders might also have been sustained by them under such circumstances as would give the pursuers no right of action? Your Lordships have to deal with the facts as they exist.” The authority of United Horse Shoe was, however, expressly rejected by the FCA in Lovastatin at [61]-[68], immediately after the FCA explained that the principles of causation are the same in both an accounting and damages. My next post will discuss why Stratas JA was of the view that this logic should not apply in assessing damages; for now, it is enough to say that the FCA has previously held that the same principles apply in both contexts.

Thus, the decision in Nova v Dow is a rather astonishing volte-face, particularly for a decision that purports to be clarifying existing principles rather than departing from them [10].

Disincentive to Infringe

What is driving this? In part, Stratas JA was concerned that allowing the deductibility of opportunity costs would encourage infringement, because the infringer would be no worse off than if it had not infringed (original emphasis):

[41] For example, infringers cannot deduct opportunity costs (i.e., what they would have done but for the infringement). Opportunity costs are hypothetical costs occurring in the “but for” world. . . . When hypothetical revenues and costs are used it distorts the picture of what the infringer actually earned making it impossible to extract the actual value derived from the infringement.

[42] In its oral submissions, the respondent Dow offered a simple, apt example to illustrate this point: if bank robbers steal $50 and are liable to return it, they cannot seek to deduct $20 because they would have earned $20 in the workplace had they never robbed the bank. To allow the $20 deduction is to incentivize bank robbery: in the best case scenario, where their robbery is not detected, the bank robbers would keep all of their ill-gotten gains and, in the worst case scenario, the bank robbers would keep what they would have earned had they not robbed the bank. Given this, why wouldn’t the bank robbers decide to rob a bank every time? Likewise, why wouldn’t infringers decide to infringe every time? Under this approach, they are in effect given a free lottery ticket—there is only upside, no downside.

This example is strained. In the first place, patent infringers are not bank robbers, a point I will come back to. Secondly, if bank robbers steal $50 and the bank wants it back, it can sue in damages and get it all; the robbers’ costs, of any type, are irrelevant, and the incentive problem is solved. But let’s put that aside and suppose an accounting is the only remedy.

Stratas JA’s discussion raises two distinct issues: does the accounting remedy provide a disincentive to infringe? Does allowing deduction of opportunity costs in particular affect that balance?

To separate the issue of incentive to infringe from that of opportunity costs, suppose that in order to rob the bank, the robbers had spent $20 to rent a jackhammer for the day. If the bank elects an accounting, then the $20 is a actual cost, which can clearly be deducted, and the robbers will be required to disgorge only $30. Stratas JA’s “free lottery ticket” remarks apply equally in this situation. In the best case scenario, the robbers would keep all of their ill-gotten gains and, in the worst case, they are no worse off—they disgorge $30, but keep $20 that they had spent on the jackhammer, for a net profit of zero.

This is exactly the same as the “perfect shelter” argument that was raised in the Perindopril litigation. The defendants, Apotex, had made infringing product in Canada and exported it. They argued that in the hypothetical world, they would have had the product manufactured abroad, and would have made the same profits (or even more) even if they had not infringed in Canada: Perindopril Accounting#1 FC 2015 FC 721, [79-80]. The trial judge rejected this argument, holding that if this was accepted, “they would never have to disgorge any profits from infringing a Canadian patent,” as they would only have to prove that they could and would have manufactured abroad [121].

The Perindopril example illustrates several points. First, the argument that an accounting provides a “free lottery ticket” or a “perfect shelter” has nothing to do with deduction of opportunity costs; it flows directly from the non-punitive nature of an accounting.

Second, the argument has been twice rejected by the FCA. In the Lovastatin litigation, Snider J accepted this argument at first instance, 2013 FC 751 [119]-[120] and was reversed on appeal, Lovastatin Damages 2015 FCA 171 [71]. The same argument was also rejected on appeal in Perindopril itself: Perindopril Accounting #1 2017 FCA 23 [34]. And there is also the authority of Schmeiser itself, in which the infringing farmer, Percy Schmeiser, was liable for “nothing”: 2004 SCC 34 [105]. While the SCC did not address the “free lottery ticket” argument directly, it must implicitly have rejected it, given that the SCC actually gave Schmeiser the “perfect shelter.”

Why did the FCA previously reject this argument? In the passage quoted above, Stratas JA asks rhetorically, “Given this, why wouldn’t the bank robbers decide to rob a bank every time?” The obvious answer is that if they get caught, they will go to jail. This is not a flip answer; it is essentially the answer given by the FCA in Lovastatin Damages at [71], where the FCA pointed out that there are a variety of other remedies that disincentivize infringement, including punitive damages, elevated costs and the prospect of injunctive relief for the remainder of the patent term. While the FCA in Lovastatin referred to “elevated costs” as a deterrent, I’d point out that even ordinary costs provide a disincentive because, at the very least, the loser will bear its own costs, as well as some portion of the prevailing party’s costs. In this case, for example, Nova has already been ordered to pay over $10 million to Dow, in addition to its own costs which are likely on the order of $20 million, if its costs are similar to Dow’s: see 2017 FC 759; 2016 FC 91 aff’d 2017 FCA 25 . This is in addition to the burden of business disruption and internal managerial time and effort which are not compensable.

In other words, an accounting is not intended to disincentivize infringement; that is left to other remedies. There is a very good reason for this. As the FCA explained in Lovastatin Damages 2015 FCA 171 (citations omitted, emphasis added):

[42] The Act as a whole is intended to advance research and development, and to encourage broader economic activity. The Act coaxes inventive solutions to practical problems into the public domain through the promise of a limited monopoly for a limited period of time. At the heart of this bargain with the inventor, and at the heart of the Act, is the concept of balance between the benefit conferred on the public through the disclosure of a new and useful invention, and the benefit conferred on the inventor through the grant of a monopoly. Thus, in the event of infringement, under-compensation of an inventor discourages research and development, and the disclosure of useful inventions. Equally, over-compensation of an inventor chills potential competition to the extent that a potential infringer is uncertain about the scope and validity of a patent. The balance at the heart of the Act requires perfect compensation.

If the bounds of the patent monopoly could be determined precisely ex ante, in the way that a property surveyor might determine the bounds of a property prior to purchase, then we would not have to worry about punitive remedies, as any potential infringer would be able to either licence or take steps to avoid infringement. But the boundaries of the patent monopoly are notoriously uncertain. In some fields, such as pharmaceuticals, is it often possible to at least identify the relevant patents, but validity and claim construction / infringement are often hotly contested and remain uncertain until after a court decision. In other fields, such as the tech sector, even identifying all the relevant patents can be next to impossible. To compound the problem, independent invention is not a defence; unlike with copyright, the fact that a party developed a technology entirely on its own does not shield it from a potential infringement action. As a practical matter, in many fields it is effectively impossible to know whether one is infringing. Not even Google knows which— or even how many—patented technologies are embodied in one of its own Pixel phones.

For that reason, it is important that the standard non-punitive remedies do not make the infringer worse off. If it is not possible to determine in advance whether you are infringing a valid patent, the best we can hope for is that if a party does infringe, it will not be worse off than if it had managed to avoid infringement. Otherwise, the risk is that investors and business-people will avoid innovative fields that are a minefield of patent risk, and invest instead in a relatively patent-free field, like real estate. The effect will be that the patent system will stifle innovation rather than encouraging it.

Because of the uncertainty related to patent validity, even discouraging intentional infringement can be problematic. Many granted patents are invalid, which is why the presumption of validity is so weak in Canadian law: see Diversified Products (1991), 35 CPR (3d) 350, 357–359 (FCA). A party who invalidates a patent in litigation does a valuable public service by eliminating an unjustified monopoly and such service should not be discouraged by the threat of punishment if their initial view that the patent is invalid turns out to be wrong. Indeed, as noted, there is already a considerable incentive to avoid infringement in terms of the costs of litigation, and adding to that burden with the prospect of punitive remedies would only add to the chilling effect. This is why the point of an accounting is to ensure that the infringer is no better off, but also no worse off: see Perindopril Accounting #1 2017 FCA 23; [29]; Lovastatin Damages 2015 FCA 171 [60].

This is also why the bank robber analogy is inappropriate. It invites us to think of infringers as bad actors, who are morally culpable in the sense of having intentionally done something which they know to be wrong, and who are therefore deserving of punishment. Yes, there are infringers who are indeed morally culpable bad actors, who may have copied a technology from the patentee, believing the patent to be valid, but other remedies, such as punitive damages, are available in such cases, after a due inquiry. But many infringers are not bad actors.

Let me give another scenario. Sumona is a clever young woman from a poor family. She is working at a low paying job, but she has an idea for a great new technology product. She quits her job, borrows some money from her family, who scrimp to put together every penney they can to support her dream. She works long, grueling days for two years, and ultimately builds her business—call it BlueBerry—into a successful firm. Then, out of nowhere, a patent assertion entity emerges—call it NTQ—with a patent on her basic technology. NTQ never brought any product to market itself, so she would never have come across their technology in the market. In theory Sumona might have done a search and found the patent, but tech patents are notoriously difficult to search, and in any event, she was busy building her company on a shoestring. If NTQ elects an accounting, Sumona will be required to give up all the profits from her company, but she argues that had she not founded the company, she would have kept working at her menial job, and that opportunity cost should be deducted from the profits. The answer, according to Stratas JA, is no, she should not be allowed to deduct those opportunity costs, to ensure that she does not have an incentive to infringe.

Opportunity costs

Is there anything different about opportunity costs? While Stratas JA’s statements of principle expressly rule out consideration of any hypothetical world, he is specifically critical of allowing deduction of opportunity costs (eg [[41]-[43], [71]), and I wonder if his objection to the use of the hypothetical world is motivated by an objection to the deduction of opportunity costs.

In my view, framing the issue as a matter of opportunity costs is unfortunate. In a sense is it true that opportunity costs are not and should not be deductible, but in a more important sense, it is irrelevant.

The issue is illustrated by the facts of the case itself. The patented product is mLLDPE, as described above. Nova sold infringing mLLDPE. The feedstock used to make it is ethylene. Ethylene itself is unpatented. It has many uses and it is apparently available on the market as a commodity chemical. If Nova had bought ethylene on the market, it is uncontroversial that it would have been able to deduct the market price of the ethylene as an actual cost in determining the profit to be accounted for.

The tricky point is that Nova did not buy ethylene on the market, but made it at its own facility in Alberta, at a price significantly below the market price. This was referred to as the “Alberta Advantage”: [FC 137]. Ultimately, the question was whether Nova should be permitted to deduct market price of ethylene (thus reducing the quantum of profit), or only its actual cost of production. The putative foregone sales of ethylene on the open market are an opportunity cost, so this raises the issue of whether opportunity costs can be deducted.

However, the details of how the argument was developed are crucial. To repeat the standard three-step test:

(1) determine the infringer’s actual profits earned in the real world;

(2) determine what the infringer’s profit would have been in the hypothetical world in which it used the best non-infringing option;

(3) subtract (2) from (1) to arrive at the profits to be disgorged

As I understand it, at first instance Nova argued that at step (1) it was not its actual cost of producing ethylene that should be deducted from its actual profits, but rather the opportunity cost, namely the foregone market value of the ethylene feedstock: [FC 134]. Fothergill J rejected this argument (my emphasis, citations omitted):

[138] In an accounting of profits, the aim is to provide the plaintiff with all of the profits made by the defendant resulting from the infringement. For this purpose, one must take the infringer as one finds them.

[139] An accounting of profits should be based on actual revenues and costs. Here, Nova enjoyed an economic advantage with respect to the cost of ethylene, the benefit of which must now be passed on to Dow. Nova did not pay a market price for the ethylene it used to manufacture the infringing products. . . . Put simply, Nova’s position that the market price should be applied is based upon a theoretical cost that it did not incur.

In my view, this is correct, so long as we understand the highlighted reference to “an accounting of profits” to mean “determination of actual profits in step (1),” which, as I understand it, was the specific question at issue in this aspect of Fothergill J’s decision. That this was the issue is also consistent with the authorities cited by Fothergill J. He cited Tye-Sil at 32 CPR(3d) 385 at [6], which says that if a defendant was unusually efficient and so unusually profitable, it must account for that extra profit, and Athey v Leonati, [1996] 3 SCR 458 at [34], which notes that under the “thin skull” rule, it is the actual damage suffered by the plaintiff that must be taken into account, not the damage that would have been suffered by an average person in the same circumstances.**

With that said, there is a certain logic to Nova’s argument. To simplify, suppose Nova’s only cost is the cost of ethylene; suppose further that it can be established that had Nova not infringed, it could and would have sold the ethylene that it actually used to make the infringing product on the open market at the market price, thus reaping the benefit of the Alberta Advantage. As just noted, on the standard approach, that hypothetical would not be taken into account at the first step of the analysis; but it would be taken into account at the second step of the analysis.

Reducing the issue to a formula shows why, with the following notation:

            Nova’s actual cost of producing ethylene, which is also its “but for” cost: Ca

            Market price of ethylene: Re

            Actual revenues from selling infringing product = RLL

Under the standard approach:

(1) Actual profits: Pa = RLL - Ca

(2) Hypothetical “but for” profits: Ph = Re - Ca

(3) Difference: Pa - Ph = (RLL - Ca) - (Re - Ca) = RLL - Re

So, on the standard analysis and above factual assumptions, the amount to be disgorged is indeed equal to Nova’s actual revenue less the market price of ethylene, which is to say the opportunity cost of the forgone market sales of ethylene. In effect, Nova’s proposed approach (if I understand it correctly) was simply to run all the steps together, and assess the profits to be disgorged by subtracting the market price from the actual revenues.

This gives the same result in this scenario, which is presumably why Nova proposed it. Nonetheless, Fothergill J was right to reject it. My simple example, in which Nova’s approach gives the same result as the standard approach, assumed that it could be established that Nova could and would have sold the ethylene on the open market. The standard analysis and Nova’s simplified analysis only give the same result if this assumption is true. I made the assumption in my example to illustrate the point, but in reality it is a factual issue and cannot be assumed, and must be proven. By running the steps together, Nova’s simplified analysis elides the crucial question of whether Nova could and would have sold its ethylene at the market price in the “but for” world. If the market for ethylene is very liquid, the answer would be “yes.” On the other hand, if ethylene is a rare chemical, with no other uses, only produced by Nova in its Alberta plant for this very purpose, then the answer would clearly be “no.” Intermediate cases are also possible. If ethylene has a number of uses, but the market is thin, with ethylene made to order by specialty suppliers with a long leadtime, it might be said that there is a market price for ethylene, yet at the same time, it would not have been possible for Nova to sell its product in the hypothetical world because it did not have established links with customers. The crucial point is that it is a factual issue that must be addressed on the facts, and Nova’s simplified analysis avoids that factual issue.

Now, I must caution that the specific arguments raised by Nova are only briefly discussed in the trial decision ([FC 134]-[140]), and I am not sure they played out exactly as I described above. In particular, Nova also argued that if it had not made infringing mLLDPE it would instead have made other products, such as pail and crate plastic, presumably from the same ethylene feedstock [FC 149]. Presumably, it would have made a supra-competitive profit in that market as well, as a result of its Alberta Advantage. If so, to deduct the “opportunity cost” of the foregone sale of ethylene in the open market would double count the ethylene advantage, and would be wrong for that reason as well. It is not clear to me how the “opportunity cost” argument and the “but for” world were linked in Nova’s argument. Moreover, the arguments were cast differently on appeal [84].

Nonetheless, the key point here is that it is understandable, but wrong, to deduct opportunity costs from actual profits. The standard “but for” analysis works perfectly well without any reference to opportunity costs, and introducing that concept simply confuses the issue.

Majority decision in FCA

Finally we can turn to Stratas JA’s decision on this point.

[90] NOVA’s first apportionment argument concerns its manufacture of ethylene, a major component of metallocene linear low-density polyethylene and its SURPASS product that infringed Dow’s patent. NOVA says that had it not infringed Dow’s patent, it would have produced ethylene anyway and would have made lawful profits from that. It says that it should be credited for these hypothetical profits in the accounting of profits—or, in other words, there should be apportionment of the profits—because it “would have earned…ethylene profits even without infringing”: NOVA’s revised memorandum of fact and law at paras. 24, 55, 56.

Stratas JA held that the argument failed on the facts, because “In the Federal Court, NOVA did not demonstrate that it would have been able to sell ethylene to third parties if it did not use it to make infringing SURPASS polyethylene” [92]. So far as I can tell, Fothergill J made no factual finding one way or the other on this point, presumably because of the way the argument was run (as just discussed). Stratas JA did not suggest otherwise. In effect he held that the absence of a finding on that point meant that it had not been established; he also referred to confidential evidence that “stood in the way” of the argument that NOVA could have sold ethylene to third parties [92].

In such circumstances, an appeal court might normally have remanded the issue to the trial judge for a factual determination, but that was unnecessary because Stratas JA also held that the argument failed as a matter of law:

[94] Further, in law, NOVA’s argument fails. NOVA says that it would have earned ethylene profits from ethylene sales without infringing Dow’s patent. But this is purely hypothetical and, thus, is legally irrelevant to the accounting of profit. What a party could have done or would have done in a hypothetical non-infringement scenario is no part of the accounting-of-profits exercise. As explained in paragraphs 37-45 and 65-79, above, only actual revenues, costs and profits are relevant.

Thus, even if it were proven that Nova could have and would have sold ethylene on the market at the market price had it not infringed, it still would not be permitted to deduct those profits, on the view that they are hypothetical and therefore irrelevant. I should point out that while I have speculated that Stratas JA was driven by an understandable desire to prohibit deductibility of opportunity costs, his holding is framed more broadly: all hypothetical costs and revenues, not just opportunity costs, are said to be irrelevant. To the extent that Stratas JA did not want to allow deductibility of opportunity costs, then I sympathize with his decision. But to say that the hypothetical world is irrelevant as a means of excluding opportunity costs (if that is what was motivating Stratas JA), is to throw the baby out with the bathwater.

Stratas JA then went on to justify this result with the incentive based arguments already discussed. I won’t repeat my critique except to say that I have not found any suggestion in the history of this litigation to suggest that Nova was a bad actor, who knowingly copied Dow’s technology, and might therefore be deserving of punishment. So far as we can tell from the facts, Nova might well have invented it independently. Nova appears to have been in the position of myriad small and medium sized Canadian companies today, who are working to develop a new and innovative product while negotiating the minefield of possible patent infringement. Today, we are all infringers—everyone with a smartphone is using patented technology, much of which is not licensed. Some very respectable commentators argue as a result that the patent system is an impediment to innovation and should be abolished: see Bessen & Meurer, Patent Failure. I do not count myself among that group, but their concern is undoubtedly real and important. That is why it is crucial that patent remedies are properly tailored, to avoid chilling innovators.

In any event, it doesn’t matter whether Nova in particular was a bad actor: we don’t know, because the issue never arose, precisely because it is irrelevant to liability and to the accounting remedy. Punitive remedies, in particular punitive damages, should be reserved to the context where some kind of blameworthy conduct has been established on the facts.

Stratas JA also referred to the concept of a non-infringing baseline, which I will discuss in my next post, holding that this could not aid Nova [97].

Finally, I want to return to the point about opportunity costs. Stratas JA stated:

[95] The Federal Court looked at this very issue and specifically rejected it. It held that “[a]n accounting of profits should be based on actual revenues and costs” and added that a “market price” for ethylene was “a theoretical cost [NOVA] did not incur” (at para. 139). These findings are correct in law and, for good measure, are supported by the evidence.

As discussed above, as I read it, Fothergill J did not state that what would have happened but for the infringement is irrelevant. Rather, he held that hypothetical costs cannot be deducted from actual profits.

This is where the characterization of the problem as involving “opportunity costs” obscures the issue. Stratas JA stated that “Opportunity costs are hypothetical costs occurring in the ‘but for’ world” [41]. This is not quite right. Opportunity costs are hypothetical profits occurring in the “but for” world—profits that might have been made by choosing a different path. Economists characterize such foregone profits has an “opportunity cost,” because by choosing one course of action in the real world, you are giving up whatever profit you would have made from the alternative course of action: if I go to see a movie instead of driving for Uber, the cost is not just the $15 for the ticket, but also the $30 I would have made from driving.

It is true that the profits that Nova would have made from selling ethylene on the market, rather than infringing, can be characterized in economic terms as an opportunity cost, and by characterizing them as such, it seems plausible that these costs should be deducted from the actual profits to arrive at the profits due to the infringement. Plausible, but wrong, for the reasons discussed above.

The proper characterization of the profits that Nova would have made from selling ethylene on the market rather than infringing, is exactly that: the defendant’s profit “had he used the best non-infringing option.” In that case, it is straightforward that the profits the defendant would have made in the hypothetical world are relevant, just as it was relevant in Schmeiser that Schmeiser would have hypothetically have made the same net profits by growing unpatented canola rather than infringing canola.

This may seem like a technical legal distinction between concepts that are economically the same. But the distinction is not merely technical: it makes the important point, discussed above, that the profits using the best non-infringing option must be proven on the facts, and cannot be assumed.

As discussed above, in my view Fothergill J was correct to hold as he did. The narrowest possible reading of Stratas JA’s decision, based on [95], is that he simply affirmed the holding that opportunity costs cannot be deducted from actual profits at the first step of the Schmeiser test. I would like to be able to read the decision that way. But unfortunately I cannot read it so narrowly, given Stratas JA’s repeated express statements that the hypothetical world is irrelevant to an accounting of profits. On the whole, Stratas JA’s decision is, in my view, inconsistent with established authority, including that of the SCC in Schmeiser. Consequently, it is also wrongly decided on the facts: given that the hypothetical world is relevant, the issue of whether Nova could and would have sold ethylene on the market rather than infringing, should have been remitted to the trial judge for a factual determination.

*To spell this out more explicitly, first consider the ethylene market. For most producers, the economic profit is zero, which is to say that the market price for ethylene is equal to the cost of production: Pe = Re - Ce = 0. (The accounting profit allows a return on capital.) Mova’s cost of production is lower than everyone else’s by an amount equal to the “Alberta Advantage,” Caa. So, Mova’s profit on bulk ethylene is Pa = Re - (Ce - Caa) = Re - Ce + Caa = Caa. This makes sense: everyone else makes zero economic profit, and Mova makes a supra-competitive profit equal to its cost advantage.

Now consider the mLLDPE market. When Mova enters the market, it matches Dov’s price, RLL. Because its costs are the same, namely the cost of ethylene, Mova makes profits that are substantially higher than previously: Pm = RLL - (Ce - Caa).

Now consider Dov’s profit in the mLLDPE market before Mova entered the market. Dov sells at RLL. Its cost is equal to the market price of ethylene, Re, if it buys on the market. If it makes its own ethylene, its cost of production is the same as everyone else (except Mova), namely Ce. Since the two are equal, it doesn’t matter whether Dov buys or manufactures. So, Dov’s profit before Mova enters the market is PD = RLL - Ce. This is the profit attributable to its ability to exclude. Without the ability to exclude, competition would drive the price down to equal cost, with an economic profit of zero, as in the ethylene market.

To compare with Dov’s profit due to the invention with Mova’s profit, recall that Mova’s profit, Pm = RLL - (Ce - Caa), which means RLL = Pm + (Ce - Caa). Substitute this in, and then recall that Mova’s profit in the ethylene market is equal to its cost advantage, which is to say Pa = Caa. So,

PD = RLL - Ce = [Pm + (Ce - Caa)] - Ce = Pm - Caa = Pm - Pa.

That is, Dov’s profit from its ability to exclude is equal to Mova’s profit in the mLLDPE market, less Mova’s cost advantage.

**Fothergill J also cited Schmeiser at paras 103-105, apparently for the general differential profits approach, as well as NV Siebrasse & AJ Stack, “Monetary Relief – Quantum” in Dimock, Intellectual Property Disputes: Resolutions and Remedies (Toronto: Thomson Reuters, 2016) at 19-73. Unfortunately, while I was a co-author, I can’t obtain that volume in the current circumstances, so I’m not sure what it says.

I would also note that there is another aspect to the “Take the infringer as you find him” maxim which is relevant to the “but for” world. When constructing the “but for” world, one must consider what the actual infringer would have done, not what an omniscient infringer would have done; so, the non-infringing alternative should be an alternative that was actually known to the defendant, and not a superior but obscure alternative that the infringer would never have considered: see Siebrasse et al, Accounting of Profits, 24 CIPR 83 (2008) at 89. This is reflected in the “would” branch of the standard approach.

No comments:

Post a Comment