Wednesday, May 24, 2017

Dead Application is Dead

University of Alberta v. Canada (Attorney General) 2017 FC 402 Russell J
            Application 2,804,560

S 37 of the Patent Rules provides that a patent application “must contain. . . a declaration that the applicant is the legal representative of the inventor,” if the applicant is not the inventor. The ‘560 application, filed in Feb 2013 by an agent on behalf of TEC Edmonton, had no such declaration. Two weeks after the 560 Application was filed, CIPO issued a requisition to TEC’s agent requiring compliance with s 37 within 12 months [3]. TEC’s agent did not respond. A year later, CIPO issued a Notice of Abandonment, which stated that the application could be reinstated within a further 12 months pursuant to s 73(3) [4]. TEC’s agent did not respond. Two months later, TEC Edmonton assigned its right in the 560 application to the University of Alberta. A new agent was appointed. The new agent did not respond to the original requisition or to the Notice of Abandonment. (You can see where this is going.) In Feb 2015, two years after the requisition was issued, the 560 application was marked as dead, meaning that the time for reinstatement had passed. A year after that, in Feb 2016, the University of Alberta filed a petition for a correction of the patent records to show the ‘560 Application was in good standing [8]. CIPO replied that the 560 application was beyond the period of reinstatement.

On this application for judicial review, the University of Alberta in effect sought to have the 560 application reinstated. Russell J refused. To get around the clear legislative scheme the University advanced “esoteric” arguments [58], [80], more or less to the effect that there was no obligation to respond to the requisition because there was really nothing wrong with the application in the first place. The statutory interpretation argument, which I won’t go into in detail, turned on a distinction between substantive and formal compliance with the requirements of the Act. Russell J pointed out that, whatever the merits of such a distinction in theory, it had no basis in the Act [86]. He also noted that the notion that there was nothing substantively wrong with the application in the first place was a purely after-the-fact rationale, which wasn’t really the reason TEC’s agent had not responded [61], [80]. More importantly, even if the applicant had been of the view that the requisition was improper, the proper recourse would have been an application for judicial review of the decision to send the requisition, in a timely manner [54]. Allowing this collateral attack on that original decision would circumvent those procedures and limitations. Moreover, there was no unfairness in this result in principle, as the statutory scheme provided ample opportunity for the applicants to correct the record by simply replying to the requisition, or rectifying within the generous legislative time-limits [89]. Nor was there any unfairness on the facts, as the record was clear that the requisition was sent and noted on the prosecution register and there was no evidence it was not received or understood by the then patent applicants and/or their then agents [54].

Monday, May 8, 2017


I'll be on vacation for a couple of weeks. I'll resume blogging sometime in the week of May 22nd.

Friday, May 5, 2017

Can Fact Witnesses Provide Evidence on the “But For” World?

Eli Lilly Canada Inc v Teva Canada Ltd 2017 FC 88 O’Reilly J
            2,041,113 / olanzapine / ZYPREXA

In this s 8 NOC case, Teva is seeking damages from Lilly as compensation for having been prevented from coming to market with a generic version of olanzapine.* O’Reilly J’s decision did not arrive at a final quantum, but simply made the factual findings necessary for those calculations [6]. Most of the issues—when would Teva’s product have been listed, what would its market share have been, how much trade-spend—turned on the particular facts. The issue of most interest and general importance was an evidentiary question as to the admissibility of opinion evidence proffered by a fact witness. O’Reilly J also clarified a point regarding pipefill.

Opinion evidence proffered by fact witness
In order to assess damages, it is necessary to determine what would have happened in the “but for” world. That is necessarily a hypothetical matter. Teva wanted its fact witness to testify as to what they would have done or what would have happened in the but-for world [14], for example whether Teva would have launched with material produced by Process 1 or Process 2 [15]. The idea was evidently that the Teva executives who were actually responsible for making those types of decisions, and who would have made the decision in question had the statutory stay not been triggered, were best placed to address that aspect of what would have happened in the but for world. Lilly objected on the basis that the opinions of fact witnesses are not admissible, and O’Reilly J agreed [13].

O’Reilly J did not cite any authority for this proposition. While I’m not an expert on evidence law, I’m not sure that there is a rule prohibiting fact witnesses from giving opinion evidence. The issue is discussed at length in Graat v. The Queen, [1982] 2 SCR 819, in which Dickson J, for the Court, concluded (836):

Except for the sake of convenience there is little, if any, virtue, in any distinction resting on the tenuous, and frequently false, antithesis between fact and opinion. The line between “fact” and “opinion” is not clear.

He then says that the question at issue should be resolved on the basis of general principles:

Admissibility is determined, first, by asking whether the evidence sought to be admitted is relevant. This is a matter of applying logic and experience to the circumstances of the particular case. The question which must then be asked is whether, though probative, the evidence must be excluded by a clear ground of policy or of law.

Ultimately, the SCC held that the witnesses in question “had an opportunity for personal observation. They were in a position to give the Court real help” (836). Consequently the evidence was admissible. Without more details than are evident from the decision, I don't know how those principles would have applied to the evidence in question in this case, but it seems to me at least arguable that it would have been admitted.

O’Reilly J noted that there was a way around the problem:

[13] During the trial, I suggested to counsel that the best way to provide the relevant evidence to the Court would be to explore with fact witnesses what they did in the real world. The witnesses could then be asked whether they knew of any reason why they would have acted differently in the but-for world. This would confine fact witnesses to their own knowledge and experience, as opposed to asking them, in an open-ended fashion, what they would have done or what they thought would have happened in the but-for world.

While this would be a getting a way around the putative rule against allowing fact witnesses to give opinion evidence, the point of Graat, as I read it, is that these kind of indirect methods are not necessary.

The specific issue in Graat, and many of the cases discussed therein, was whether lay witnesses, such as police officers, could testify as to whether the accused’s ability to drive was impaired by alcohol. As many of the decisions emphasized, this is a situation where an ordinary person who has actually observed the accused has sufficient experience to form a helpful opinion. In a passage adopted by Dickson J at 839, Lord MacDermott in Sherrard v Jacob [1965] NILR 151, 162, stated:

The driving of motor vehicles is now so much a matter of everyday experience for ordinary people that I find it difficult to see how inferential or opinion evidence as to being (a) under the influence of drink and (b) thereby unfit to drive a car can be placed in different categories for the purpose of determining admissibility. The one as much as the other seems to be within the capacity of the non-expert to form a reasonable conclusion

While drunken driving might seem far removed from pharmaceutical production processes, the parallel seems to me to be close nonetheless. Whether a pharmaceutical company might have used one process or another is not a matter of everyday experience for ordinary people, but, as I understand the facts, it was a matter of everyday experience for the particular fact witnesses in this case. The point from Graat is that whether or not he had been qualified as an expert, the witnesses in question had sufficient relevant knowledge and experience to provide helpful evidence on the issue at hand.

A broad reading of Graat is that opinion evidence is admissible if it would be helpful and there is no clear ground of policy or of law for excluding it (835-36). A narrower reading is that fact witnesses may give opinion evidence when that is a method of making a compendious statement of facts (840). Even on the narrow reading, it seems to me that there is a good argument that the evidence at issue in this case would be admissible (again with the caveat that it is not possible to take a firm view with the limited details in the opinion). In some decisions reviewed by Dickson J, the court had taken a position very similar to that implicit in O’Reilly J’s suggestion, to the effect that the fact witness could state all the factual circumstances which would lead her to form the opinion that the accused was intoxicated, but the witness could not state the opinion itself (828). This position was clearly rejected by the SCC in Graat. As I understand, the point of allowing the opinion evidence is largely that such stratagems are unnecessary.

I don’t want to say that the evidence at issue in this case should have been admitted, as the nature of the evidence was described only in general terms. And, again, I am not an expect in evidence law, so there may well be a rule or development that I have missed. But, at least on the authority of Graat, it seems to me that there is no strict rule prohibiting fact witnesses from giving opinion evidence.

Teva claimed that its losses should include an amount for pipefill— “that is, the quantity of sales Teva would have made to distributors in the but-for world, an amount that would not be captured by retail sales figures” [90]. The idea is that the manufacturer ships its product from the factory, it sits a while with the wholesaler until ordered by the retailer, and sits a while with the pharmacy before being dispensed to an individual [93]. Teva wanted its lost profits to include losses on product that had left the factory, but had not yet been dispensed. O’Reilly J held that Teva was not entitled to lost profits on pipefill [103]. The basic rule for s 8 damages is that “but for” causation, in a world in which the patentee had not applied for an order of prohibition, except that losses made after the compensable period are not compensable, even if they were caused by the statutory stay: 2009 FCA 187, [92]-[102]; 2011 FCA 149. This is a statutory exception to the general principle that losses caused by the wrong are recoverable. O’Reilly J observed that pipefill product is actually sold, but it is sold outside of the compensable period [92]. Consequently, it is not a recoverable loss. This holding is significant because there are several other cases, reviewed in detail by O’Reilly J [96]-[101], which arguably did allow pipefill. He held that these cases were not determinative. In some the holding was ambiguous, and “In none of them was the issue seriously contested or a quantum specifically calculated” [101]. To the extent they differed, he therefore declined to follow them [103]. Given that O’Reilly J gave full consideration to the point, his holding is likely to be influential.

*The procedural history is long. Hughes J ruled against Lilly in 2007 FC 596, the NOC proceeding that ultimately gave rise to this s 8 action [4]. After Novopharm (now Teva) launched, O’Reilly J held the patent invalid Olanzapine (No 1) 2009 FC 1018 which was reversed and remanded by 2010 FCA 197. On remand, in was Olanzapine (No 2) 2011 FC 1288, O’Reilly J again concluded the patent was invalid and this was affirmed 2012 FCA 232 [2] (blogged here).

Wednesday, May 3, 2017

Should 36(1)(4)(b) of the Federal Courts Act be Repealed?

Dow Chemical Co v NOVA Chemicals Corp 2017 FC 350 Fothergill J
            2,160,705 / film-grade polymers / ELITE, SURPASS

In the liability phase of this action, Dow Chemical Co v NOVA Chemicals Corp 2014 FC 844 aff’d 2016 FCA 216, O'Keefe J held Dow’s 705 patent related to advanced film-grade “mLLDPE” polymers, to be valid and infringed by Nova, and he also held that Dow was entitled to damages under s 55(2) of the Patent Act for pre-grant ‘infringement’, and that Dow was entitled to elect between damages and an accounting for post-grant infringement. Dow subsequently elected an accounting [107]. Fothergill J’s decision in the remedies phase addresses various issues to allow the parties’ accountants to calculate the actual sums owed by Nova to Dow [6]. (For more background see last Wednesday’s post.) The question of pre-judgment interest arose both in the context of interest on the award of a reasonable compensation under 55(2), and also in terms of interest on the accounting. There are two basic questions as to interest: what is the rate, and should it be compounded?

In respect of the interest on reasonable compensation under s 55(2), [101]-[105], the only issue was the rate, as O’Keefe J had held that interest should not be compounded [101]. The Federal Courts traditionally tended to award interest at the annual average Bank of Canada bank rate, which may in practice be undercompensatory. The Federal Courts Act, s 36(3) does, however, give the court discretion to award a different rate. There seems to have been a trend over the last few years for parties to ask for a different rate (see here for an overview), and the question then arises as to what rate is appropriate. In this case, Dow asked for, and Fothergill J awarded, pre-judgment interest at a rate equal to Dow’s annual cost of borrowing [102]. This basis was largely undisputed [104]. A similar rate was also used in, for example, Cefaclor Damages 2014 FC 1254 (discussed here). This suggests that there is a trend towards using the plaintiff’s annual cost of borrowing as the appropriate pre-judgment interest rate. While I hesitate to express a firm view, interest assessed on this basis strikes me as sound, as it best reflects the actual loss to Dow. With that said, considerations of administrative efficiency might support a different rate, particularly when the quantum at stake is relatively modest. Roy Epstein, Prejudgment Interest Rates in Patent Cases: Don't Compound an Error, 24(2) IPL Newsletter (2006), has an interesting discussion suggesting that data on average actual short-term market interest rates paid on commercial and industrial loans might be a less expensive way at arriving at a reasonably reliable assessment of the plaintiff’s loss. (Hat tip to Professor Tom Cotter for bringing this article to my attention.)

Turning to the issue of interest on the award of Nova’s profits, O’Keefe J had left the question entirely open, in terms of rate and compounding [166]. Nova argued for prime rate + 1%, not compounded [167], while Dow argued that the applicable rate should be Nova’s weighted annual cost of borrowing, compounded [168].

Fothergill J held that the appropriate rate was Nova’s weighted average annual cost of borrowing [168], [173]. This is consistent with his holding in the context of reasonable compensation.

He also held that interest should be compounded [173], relying, inter alia on Reading & Bates Construction [1995] 1 FC 483, 486, stating “ the awarding of compound pre-judgment interest as deemed earnings on the profits is the rule,” and Beloit Canada [1995] FCJ No 733, 61 CPR(3d) 271 (FCA) reemphasizing the same point [171]. The rationale is compelling: as a result of the infringement, the infringer had use of money that it would not otherwise have had, and in light of “the modern reality that interest paid or earned on deposits or loans is compound interest,” the need to “achieve equity in the accounting of profits“ requires that interest be compounded, or the infringer to profit from its wrong: (Reading & Bates, ibid).

Given that compound interest is established as the norm in an accounting, it is curious that compound interest has only recently begun to be awarded in the context of damages. The rationale is equally compelling; the patentee was deprived of money it would otherwise have had, and in reality that money would have earned compound interest. The obstacle has been s 36(1)(4)(b) of the Federal Courts Act, which on its face prohibits compound interest. Bank of America Canada v Mutual Trust Co, 2002 SCC 43 generally, and Eli Lilly / Cefaclor Damages, 2014 FC 1254 in the context of damages, have gotten around this by awarding compound interest as being compensation, rather than interest on interest as such: see here. For reasons that are not clear to me, s 36(1)(4)(b) and its predecessors were not seen as posing the same problem in the context of an accounting. The provision prohibits “interest. . . on interest” accruing under 36(2), which refers to “an order for the payment of money,” and so on its face would also apply to an accounting.

In any event, notwithstanding s 36(1)(4)(b), compound interest is now available on both an award of damages and an accounting of profits, and this is sound in principle. It would appear best to repeal this provision entirely, at least so far as patent law is concerned.

Monday, May 1, 2017

Fixed Costs and the Differential Profits Approach

Dow Chemical Co v NOVA Chemicals Corp 2017 FC 350 Fothergill J
            2,160,705 / film-grade polymers / ELITE, SURPASS

In the liability phase of this action, Dow Chemical Co v NOVA Chemicals Corp 2014 FC 844 aff’d 2016 FCA 216, O'Keefe J held Dow’s 705 patent, related to advanced film-grade “mLLDPE” polymers, to be valid and infringed by Nova: for more background see last Wednesday’s post. Fothergill J’s decision in the remedies phase addresses various issues to allow the parties’ accountants to calculate the actual sums owed by Nova to Dow [6]. Dow elected an accounting [107], and one of the issues requiring clarification was deduction of fixed costs [141]-[165].

The infringer’s profits are its revenues less its costs. The direct costs of producing the infringing goods are clearly deductible, but fixed costs such as rent and general overhead present a difficult problem. The argument against deduction is that the fixed costs would have been incurred in any event; the argument in favour is that a business is not profitable if it doesn’t cover its fixed costs. While this case raised the problem in the context of an accounting of the infringer’s profits, the problem is the same when assessing lost profit damages.

I’ll begin with an important issue of terminology. As noted by Fothergill J [143], the differential profit approach is the preferred approach, but it was not used in this case because:

[146] In this reference, Nova concedes that there were no “direct non-infringing alternatives” available for the purpose of applying the “differential profits” approach.

I would quibble with this statement, at least to emphasize the importance of the word “direct.” It’s not quite right to say that the differential profit approach can’t be applied when there are no non-infringing alternatives. In the differential profit approach, “[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” (Schmeiser, 2004 SCC 34, [102]). Alternatively, the differential profit approach says that the infringer’s profits are the difference between its actual profits and the profits it would have made in “a hypothetical world where the defendant’s impugned conduct did not take place” 2016 FCA 161 [47]. These are equivalent formulations, which means that in a broad sense, the “non-infringing alternative” is whatever the infringer would have done had it not infringed. In this case, on the facts, Nova would have made lower grade products, such as “pail and crate” grade plastics [158]. In the broad sense, that is the NIA. But when Nova conceded that there were no “direct” non-infringing alternatives, what it was saying is that there were no alternative products that it could have made that would have taken any share of the market from the patented product: the pail and crate grade plastics do not compete with mLLDPE. That is a narrower use of the term “non-infringing alternative.”

The term “non-infringing alternative” tends to be used in the narrower sense, as it was used in this case. That is understandable, because when the infringer’s alternative would take market share from the patented product, that raises a host of issues, such as the size of the NIA’s market share, that do not arise otherwise. However, we shouldn’t lose sight of the fact that there is only one basic question: what is the difference between the infringer’s actual profits (or patentee’s profits, when damages are assessed) and the profits it would have made had it not infringed.

Now, turning to the issue at hand, in US law of damages fixed costs are generally excluded when determining profits: Paper Converting Mach Co 745 F.2d 11, 22 (Fed Cir 1984). The English and Australian courts, on the other hand, will allow deduction of some part of the overhead if it can be shown on the facts that, but for the infringement, the infringer would have used its capacity to manufacture non-infringing product. The leading case is Dart Industries Inc v Decor Corporation Pty Ltd [1993] HCA 54, in which the High Court stated the following:

[14] In calculating an account of profits, the defendant may not deduct the opportunity cost, that is, the profit forgone on the alternative products.

The opportunity cost is the profit that the defendant would have made if it had made a non-infringing alternative, so this is, strictly, a rejection of the differential profit approach. The High Court continued:

[14] But there would be real inequity if a defendant were denied a deduction for the opportunity cost as well as being denied a deduction for the cost of the overheads which sustained the capacity that would have been utilized by an alternative product and that was in fact utilized by the infringing product.

[15] Where the defendant has forgone the opportunity to manufacture and sell alternative products it will ordinarily be appropriate to attribute to the infringing product a proportion of those general overheads which would have sustained the opportunity. On the other hand, if no opportunity was forgone, and the overheads involved were costs which would have been incurred in any event, then it would not be appropriate to attribute the overheads to the infringing product.

If, on the facts, the infringer would have simply cut back production, and let the plant sit idle, then there is no deduction; but if it can be proven that there were indeed other opportunities that were forgone in order to make the infringing product, then the fixed costs may be deducted accordingly, though there is no deduction for the foregone profits. In effect, fixed costs are used as a partial proxy for true opportunity costs, except that the opportunity is limited to cost recovery, and excludes profit that would have been made in the “but for” world: see Duff & Phelps Group, Siebrasse & Stack, “Monetary Relief – Quantum,” in Dimock, IP Disputes 19-103-04.

Even though the Dart Industries approach is, strictly, a departure from the differential profit approach, it may nonetheless be justified on the basis of administrability. That was basically the view of McHugh J, concurring in Dart. But I wonder if there are circumstances in which a full absorption cost approach, which would deduct a portion of fixed costs even if there were no other opportunity, might make sense? All costs are variable in the long run, and suppose the infringement took place over a long period in a purpose built plant that was not suitable for any other product. (Perhaps the infringer had independently created the invention.) We might say that but for the infringement, the infringer would have let the plant sit idle, in which case the costs of the plant would not be deducted; but we might also say that but for the infringement, the infringer would not have built the plant at all, in which case a full absorption approach might be consistent with the differential profit approach. In any event, the Dart Industries approach is clearly preferable to the US approach, as opportunity costs are real costs, which should not be ignored.

In summary, the current Canadian approach, following Dart Industries, seems basically sound, though there is an argument for the pure opportunity costs approach; but the problem is inherently difficult, and different results might be warranted, particularly on unusual facts. For a discussion of the different approaches, see generally Cotter, Comparative Patent Remedies, 206-07.

Idiosyncratic costs [134]-[140]
On a different issue, Nova produced ethylene, the basic feedstock for the production of the patented plastic, at its own facility, at a significantly lower cost than the market price [137]. The question was whether the cost to be deducted should be the market price, or Nova’s actual costs [134]. Fothergill J held that Nova’s actual costs should be used, on the principle that “one must take the infringer as one finds them” [138].