1 Availability and Standard

Most major patent systems recognize that a patentee’s lost profits are an appropriate measure of damages, although there are differences regarding both the standard for awarding lost profits as well as the implementation of this methodology. Here, we first summarize the availability and (where ascertainable) the standard for proving entitlement to lost profits in key jurisdictions and then offer several recommendations.

In the United States, Section 284 of the Patent Act provides that a “court shall award the [prevailing patentee] damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer.”Footnote ¹¹ Courts recognize two types of compensatory damages under Section 284: (1) “the patentee’s lost profits” and (2) “the reasonable royalty [the patentee] would have received through arms-length bargaining.”Footnote ¹²

The classic type of lost profits damages are from lost sales of a product or service that practices the patent.Footnote ¹³ “Lost sales constitute sales that the patent owner failed to make due to the infringement, as well as sales the infringer made that the patent owner would have made but for the infringement.”Footnote ¹⁴ For instance, in Seymour v. McCormick (1853), the Supreme Court of the United States held that a prevailing patentee “is entitled to the actual damages he has sustained by reason of the infringement, and those damages may be determined by ascertaining the profits which … he would have made, provided the defendants had not interfered with his rights.”Footnote ¹⁵ However, the Court rejected the trial court’s presumption “that if the [infringer] had not made and sold machines, all persons who bought the [infringer]’s machines would necessarily have been compelled to go to the patentee and purchase his machines.”Footnote ¹⁶ Instead, the Court required proof the patentee would have actually made these sales absent infringement – a burden that it was unable to carry.Footnote ¹⁷

Modern U.S. case law follows a broadly similar approach, holding that lost profits are not presumed,Footnote ¹⁸ and instead requiring the patentee to “show a reasonable probability that, ‘but for’ the infringement, it would have made the sales that were made by the infringer.”Footnote ¹⁹ This is most commonly achieved using the four-factor Panduit test,Footnote ²⁰ which requires “the patent owner to prove: ‘(1) demand for the patented product, (2) absence of acceptable non-infringing substitutes, (3) his manufacturing and marketing capability to exploit the demand, and (4) the amount of the profit he would have made.’”Footnote ²¹ The first three of these requirements are best viewed as proxies to establish causation in fact. The first, demand for the patented product, demonstrates that at least some consumers would have preferred the patentee’s product because of the patented technology. The second, which will be discussed in more detail below, asks whether consumers would have been willing to substitute a noninfringing alternative for the patentee’s product.Footnote ²² If so, the substitution effect will make it more difficult for the patentee to obtain supra-competitive profits.Footnote ²³ The third asks whether the patentee would have been able to increase production in order to make (at least some of) the sales that the infringer actually made.Footnote ²⁴ The fourth and final element encompasses the “but for” market reconstruction – i.e., what would have been the patentee’s profits absent infringement? Courts in the United States have viewed this as a relatively demanding element, requiring “reliable economic proof of the market” that would have developed “‘but for’ the infringement” to establish the amount of lost profits with sufficient accuracy.Footnote ²⁵

U.S. courts also allow recovery for other foreseeable profits lost by the patentee due to the infringement.Footnote ²⁶ These may include, for instances, losses due to price erosion,Footnote ²⁷ lost sales of unpatented products sold by the patentee that directly compete with the infringing product,Footnote ²⁸ and (explained in more detail below) lost sales of unpatented components and products that are “functionally associated” with the patented item.Footnote ²⁹ In addition, if a patentee can prove entitlement to lost profits for only some of its lost sales, it can “recover a mixed award of lost profits on some sales and an established or reasonable royalty on other sales.”Footnote ³⁰

Despite this, awards of lost profits are increasingly uncommon in the United States. A recent study by consulting firm PricewaterhouseCoopers found that lost profits alone represented 26 percent of patent damages awards from 1997 to 2006, compared to nearly 60 percent that awarded damages based exclusively on a reasonable royalty.Footnote ³¹ This trend continued from 2006 to 2015, where only 21 percent of patent damages awards were based solely on lost profits.Footnote ³² In addition, the study’s authors note that “price erosion claims have become almost nonexistent in recent years.”Footnote ³³ Several explanations have been offered for this development. First, lost profits are available only in a subset of patent disputes – namely, cases where both “the patent owner and infringer actively compete in the same market.”Footnote ³⁴ Thus, cases brought by nonpracticing patentees – which represent a considerable share of patent infringement lawsuits filed in the United StatesFootnote ³⁵ – are ineligible for a lost-profits recovery. Second, in the context of complex, multifunction products, it may be difficult for a patentee to demonstrate that the infringer’s inclusion of a patented feature caused it to lose sales.Footnote ³⁶ Third, some patentees who might be eligible to recover lost profits appear to be eschewing them in favor of reasonable royalty damages.Footnote ³⁷ This may be the case for several reasons: because the patentee is skeptical that it can satisfy Panduit’s rigorous requirements; because the patentee wishes to avoid disclosing detailed financial information regarding its business to a competitor; or because the patentee believes that it can obtain at least as large of an award using the more flexible reasonable royalty approach.

The UK and Commonwealth countries, like Canada and Australia, similarly permit awards of lost profits damages. Section 61 of the UK Patent Act authorizes the patent owner to claim “damages in respect of the infringement,” and like in the United States, case law in the United Kingdom has explained that the objective of patent damages is to restore the patentee to the position it would have occupied but for the infringement.Footnote ³⁸ In general, this extends to all losses by the patentee (including lost profits and price erosion) that are: (1) foreseeable, (2) caused by the infringement, and (3) not excluded from recovery by public or social policy.Footnote ³⁹ In practice, this standard appears to be more flexible and less demanding than Panduit; the High Court of Justice has specifically noted that although the burden of proof is on the patentee, “[d]amages are to be assessed liberally.”Footnote ⁴⁰ Both Canada and Australia permit the recovery of actual damages suffered by the patentee due to infringement as well,Footnote ⁴¹ including lost profits and price erosion subject to the foreseeability principle.Footnote ⁴² In Canada, for example, to recover lost profits, the patentee “must show on a balance of probabilities that ‘but for’ the defendant’s wrongful conduct, [it] would not have suffered loss.”Footnote ⁴³

One significant area of divergence between the United States on one hand, and the United Kingdom and Australia on the other, is that the latter may decline to award damages (including lost profits) against an infringer who was not aware, and had no reason to believe, that the patent existed.Footnote ⁴⁴ In other words, in these jurisdictions, an unwitting infringer may only be subject to injunctive relief. In contrast, direct patent infringement in the United States is a strict liability offense, and damages can be awarded against even an innocent infringer.Footnote ⁴⁵ This distinction may not be as sharp in practice, however, because “in the typical lost profits case, the defendant is a competitor of the plaintiff and thus unlikely to qualify as an innocent infringer.”Footnote ⁴⁶ In addition, in these countries the fact that a patentee marked its productsFootnote ⁴⁷ with the patent number(s) or an Internet link containing patent information can undermine a claim of innocent infringement.Footnote ⁴⁸

Lost profits are similarly available as a matter of principle in every EU country,Footnote ⁴⁹ although in practice they appear to be considerably less common than in the United States and the United Kingdom.Footnote ⁵⁰ For example, in Germany, patent owners may recover the difference between the profit they would have earned absent infringement and their actual profits.Footnote ⁵¹ This may include “both profits lost on sales lost to the infringer and damages for price erosion,” as well as “more remote harms” like market confusion provided that such harms were likely caused by the infringement.Footnote ⁵² Similarly, in France, the patentee can recover lost profits on lost sales of patented goods as well as price erosion.Footnote ⁵³ However, the amount of lost profits awarded in France appears to be considerably lower than in the United States, even after accounting for the larger size of the U.S. economy.Footnote ⁵⁴

In Asia as well, lost profits are available in most major jurisdictions, although many also require some degree of culpability by the infringer. For example, in its 1998 amendments to Japan’s patent law, the Diet “intended for awards of lost profits … to be the general or default remedy in patent infringement matters.”Footnote ⁵⁵ As a result of these changes, the owner or exclusive licensee of a Japanese patent can claim damages “against an infringer … sustained as a result of the intentional or negligent infringement of the patent right.”Footnote ⁵⁶ Japan’s Patent Act also presumes that the amount of the patentee’s lost profits is the same as the infringer’s profits.Footnote ⁵⁷ Despite these changes, one study found that lost profits represented the minority approach to compensating the patentee: about 20 percent of all patent damages claims in Japan were based on lost profits, and a mere 17 percent of successful claims.Footnote ⁵⁸ Korea’s patent law is highly similar to Japan’s, providing that lost profits shall be awarded only for intentional or negligent infringement.Footnote ⁵⁹

In China, Article 65 of the Patent Law establishes a statutory preference for awarding “the patentee’s actual losses caused by the infringement,” although it also permits use of the infringer’s profits as a proxy for the amount of the patentee’s loss if it is difficult to determine.Footnote ⁶⁰ In addition, while “article 65 does not expressly condition damages liability on the defendant’s intent or negligence …, as a general matter, Chinese law accepts the principle that damages liability is conditional upon the defendant’s fault.”Footnote ⁶¹ Despite this preference, lost profits are rarely awarded in China; in over 90 percent of cases, statutory damages are awarded instead.Footnote ⁶² Several empirical studies also have found that the amount of Chinese patent damages awards tend to be low, particularly by U.S. standards.Footnote ⁶³

India’s law regarding patent damages has been influenced by that of the United Kingdom.Footnote ⁶⁴ The Indian Patent Act authorizes the patentee to elect to recover either actual damages or an accounting of the infringer’s profits.Footnote ⁶⁵ Similarly, India precludes any damages, including lost profits, in cases where the infringer “was not aware and had no reasonable grounds for believing that the patent existed.”Footnote ⁶⁶ However, there appears to be little precedent from Indian courts that provides specific guidance regarding damages.Footnote ⁶⁷

In light of the foregoing discussion, we make several recommendations regarding the availability and standard for awarding lost profits damages. First, we recommend that lost profits (including from lost sales and price erosion) should be the preferred measure of damages when a patentee can establish harm in a product market due to the infringement. As the Supreme Court of the United States explained in Aro Manufacturing v. Convertible Top Replacement Co., the patentee’s loss from patent infringement is “‘the difference between [the patentee’s] pecuniary condition after the infringement … and what his condition would have been if the infringement had not incurred.’”Footnote ⁶⁸ Lost profits best serve this make-whole objective by compensating the patentee’s actual losses caused by the infringer’s market entry. Although available in all major jurisdictions, in practice lost profits are less commonly awarded than other methodologies for determining damages, such as a reasonable or established royalty or an award of the infringer’s profits. In particular, we are concerned about authority suggesting that a patentee could potentially obtain greater than its lost profits under an alternative measure of damages (such as a reasonable royalty), as this would tend to overcompensate the patentee.Footnote ⁶⁹

Second, we recommend that lost profits should be awarded whenever a practicing patentee can demonstrate “but for” causation by a preponderance of the evidence.Footnote ⁷⁰ The focus on causation is central to the lost-profits analysis, but some jurisdictions such as the United States have articulated more detailed standards or requirements as part of this inquiry (i.e., the Panduit factors). Rigorous adherence to such standards might make it more difficult in practice for a patentee to establish entitlement to lost profits.Footnote ⁷¹ We recommend that jurisdictions instead focus on “but for” causation as the central inquiry for lost profits claims.

Third, as previously noted, jurisdictions differ on whether some degree of fault or culpability is required to support an award of lost profits. We were unable to reach a consensus about whether lost profits should be available regardless of the infringer’s degree of fault. However, this may be a worthwhile topic for further research.Footnote ⁷²

Fourth, we were unable to reach a consensus regarding some jurisdictions’ (rebuttable) presumption that the amount of the patentee’s loss is equal to the amount of the infringer’s profits. Arguments in favor of this approach include that it may simplify the damages calculation and thus reduce adjudication costs,Footnote ⁷³ and that the patentee may prefer to rely on the infringer’s profits because it would not require disclosure of the patentee’s sensitive financial information (such as net revenue, fixed costs, variable costs, and research and design costs) to a competitor. Arguments against this approach include that the infringer’s profits may represent a poor proxy for the amount of the patentee’s lost profits, potentially resulting in over- or under-compensation. For example, if the infringer is more efficient than the patentee (and thus has a higher per-unit profit on its sales), then using the infringer’s profits as a basis for determining the patentee’s loss will result in overcompensation.Footnote ⁷⁴ In contrast, if the infringer is less efficient than the patentee (and thus has a lower per-unit profit on its sales), then using the infringer’s profits will result in under-compensation unless the patentee can overcome this presumption. It is unclear, however, to what extent this presumption actually results in over- or under-compensation. We propose further research on this issue from both theoretical and empirical perspectives.

2 Noninfringing Alternatives

A specific issue worth further discussion is the role that noninfringing alternatives play in the lost-profits analysis. If the infringer could have competed against the patentee just as effectively by offering a noninfringing alternative to the patented invention, the patentee would have lost just as many sales (and thus profits) absent the infringement. In such a case, the patentee has not lost any profits caused by the infringement, since it would have lost those profits anyway, and it should recover at most a reasonable royalty reflecting some portion of the value of the patented technology to the infringer (e.g., its profit-enhancing or cost-reducing advantages over the next best alternative). Put another way, an award of lost profits if a noninfringing alternative exists would render the patentee better off than it would have been “but for” the infringement, and thus would enable the patent owner to reap a reward in excess of the economic value of its invention.Footnote ⁷⁵ Courts in the United StatesFootnote ⁷⁶ and FranceFootnote ⁷⁷ have long recognized the relevance of noninfringing alternatives in these contexts, as more recently has Canada,Footnote ⁷⁸ while courts in the United Kingdom and Germany have not.

The United Kingdom in particular continues to abide by an 1888 decision of the House of Lords, United Horse-Shoe & Nail Co. v. John Stewart & Co.,Footnote ⁷⁹ which rejected the relevance of noninfringing alternatives to damages calculations.Footnote ⁸⁰ With all due respect to the House of Lords, we think that United Horse-Shoe fails to grasp the economic logic embodied in the noninfringing alternative concept, and that there is little reason for contemporary patent systems to continue adhering to the decision. We therefore recommend that United Horse-Shoe and other similar decisions elsewhere be overruled, and that courts explicitly recognize the importance of considering noninfringing alternatives to the accurate calculation of patent damages.Footnote ⁸¹

Relatedly, to the extent that domestic law permits the recovery of the infringer’s profits attributable to the infringement,Footnote ⁸² we also recommend that courts or legislatures explicitly define the term “profit” to mean the benefit derived from the infringement over the next best alternative. Although that benefit most commonly takes the form of an increase in the infringer’s profits compared to what it would have earned using a noninfringing alternative, in a case in which the infringer winds up losing money (for example, because it did not sell enough infringing products to cover costs) the infringer nonetheless benefits if its losses would have been even greater absent the infringement. Thus, to the extent disgorgement is permitted, the infringer also should be required to disgorge the cost saving it enjoyed as a result of the infringement, even if it earned no “profit” in an accounting sense.

Another matter is what qualifies as a noninfringing alternative to the patented technology. For example, U.S. case law establishes that an alleged alternative must have similar functionality and a comparable price to the patented technology.Footnote ⁸³ This definition, however, fails to recognize that substitution is a matter of degree in product markets,Footnote ⁸⁴ particularly for multifunctional products where consumers may value certain features more than others. Even an imperfect substitute that provides some, but not all, of the functionality of the patented invention can nonetheless affect both the price of the patented product as well as consumer choice.Footnote ⁸⁵ As a result, we recommend that courts focus on the substitutability of noninfringing alternatives in evaluating how many of the infringer’s sales the patentee would have made in the “but for” analysis.

In addition to these recommendations, there remain several issues regarding non-infringing alternatives that deserve further research. First, as discussed in Chapter 1, to date there has been little discussion in the legal and economic literature of how courts should proceed when the next best alternative itself is patented.Footnote ⁸⁶ In the reasonable royalties context, the principal question raised by the presence of patented alternatives is whether one should assume that the owners of the two patents engage in Bertrand competition – which ultimately could drive the price of both patents down to zero, if neither is better than the other – or whether such an assumption threatens to undermine the patent incentive. In the present context, the question that arises from the presence of patented alternatives is whether courts should presume that the patented alternative was not available to the infringer, or instead should require proof that the patent covering the alternative was either invalid or would have been licensed (and if so, at what price). The problem with the latter option is that it risks greatly increasing the cost of adjudication for comparatively little benefit.Footnote ⁸⁷ Further research to address the issue would be welcome.

Second, further research on which party should be required to prove the absence of noninfringing alternatives would be helpful, as there is an apparent conflict among jurisdictions on this issue. In the United States, the case law is a bit muddled, but the patentee generally must make this showing as part of the Panduit test for lost profits.Footnote ⁸⁸ This effectively requires the patentee to prove a negative – namely, that there was no feasible noninfringing alternative during the period of infringement. In contrast, in Canada, the infringer bears the burden of demonstrating the existence of a noninfringing alternative.Footnote ⁸⁹ To our knowledge, there is little discussion in the legal or economic literature addressing which of these approaches is optimal. One might speculate that the infringer often would be better placed than the patentee to propose and substantiate the existence of noninfringing alternatives, particularly if, as in Grain Processing, the infringer had the capacity to create and implement a noninfringing design around without much difficulty. But perhaps patent owners have unique insights into the matter that are not apparent at first blush, or maybe the allocation of the burden of proof on this issue does not matter much in practice because both parties have sufficient motivation to present the evidence that best favors their position.

Third, the degree of certainty needed to establish that a noninfringing alternative was in fact available to the infringer is not always clear. The U.S. decision in Grain Processing, for example, held that a noninfringing alternative was available during the period of infringement – even though it was not actually on the market – because it would have been simple for the infringer to develop a noninfringing (but slightly costlier) process to produce the (unpatented) end product. Other cases addressing this issue turn on their unique facts,Footnote ⁹⁰ and of course having to establish the availability of an alternative that was not actually on the market at the time of infringement poses some risk of increasing adjudication and error costs.Footnote ⁹¹ Nonetheless, we are inclined to agree with the Grain Processing framework on the basis that the increase in accuracy justifies the cost, though further research might help to structure this analysis so that courts can apply it in a consistent, predictable, and cost-efficient fashion.

3 Lost Profits on Sales of Related but Unpatented Products

Another important issue is whether a prevailing patentee can recover lost profits damages for unpatented products that are related to sales of the patented product.

In the United States, courts have applied three (at least partially overlapping) doctrines to determine which kinds of potential lost sales can be compensated for in a lost profits award. Generally, these doctrines apply respectively to (1) sales of products that incorporate both infringing and noninfringing components, (2) additional contemporaneous sales of distinct but related items, and (3) anticipated future sales of replacement or repair parts.

First, in the context of complex products, courts have applied the so-called entire market value rule to define the scope of the primary lost “sale” for which profits may be owed. Though the majority of modern case law on the entire market value rule has come in the context of reasonable royalty awards,Footnote ⁹² courts have also discussed the doctrine in relation to lost sales of infringing products or assemblies that have both patented and unpatented components. In such cases, the entire market value rule dictates that lost profits damages may be recovered for lost sales of all components that operate as part of the same “functional unit” as the infringing component or part, such that they are “analogous to components of a single assembly or parts of a complete machine.”Footnote ⁹³ Thus, for example, a patentee that sells paper winding equipment can recover lost profits for lost sales of the entire line of equipment – including the unpatented stand, loader, embosser, and sealer – because all three work together with the infringing rewinder as part of a single assembly that the patentee virtually always bundled into a single sale.Footnote ⁹⁴

Second, courts have also developed the related concept of “convoyed” or “collateral” sales. Convoyed sales are sales of items that, though physically separate from the infringing product, are nonetheless typically sold along with the infringing product. Though case law in this area is muddled, courts have sometimes suggested that patentees may recover damages for lost sales of items simply because they were traditionally purchased at the same time as the infringing device.Footnote ⁹⁵ For example, in Golden Blount, Inc. v. Robert H. Peterson Co., the Federal Circuit affirmed an award of lost profits damages that included profits lost on aesthetic artificial logs and grates in addition to profits lost on infringing gas fireplace burners because it was “standard practice in the industry” to sell all three items together.Footnote ⁹⁶ However, in most cases involving convoyed sales, the Federal Circuit has required that the patentee demonstrate that the unpatented component must be “functionally associated” with or related to the patented product in some way.Footnote ⁹⁷

Third, at least some case law differentiates between convoyed sales and sales of repair or replacement parts, sometimes called “derivative sales,” which are often made in the future after the original infringing sale.Footnote ⁹⁸ However, in principle, the same rules applicable to convoyed sales appear to apply in this context as well. As the Federal Circuit stated in King Instrument Corp. v. Otari Corp., lost profits damages are recoverable for spare parts when the patentee “normally would have anticipated the sale of the spare parts” but for infringement.Footnote ⁹⁹ Thus, for example, in Leesona Corp. v. United States, the patentee was allowed to recover damages reflecting lost sales of replacement anodes for a patented battery where in “a normal ‘life cycle,’ it was anticipated that the 22 anodes for each battery would each be replaced 50 times.”Footnote ¹⁰⁰

At their core, all three doctrines focus on drawing a line between what sales were, and were not, foreseeably lost due to infringement. We recommend that losses associated with all three categories of sales should generally be recoverable provided that the patentee can demonstrate both (1) “but for” causation and (2) proximate causation, which is established by demonstrating that sales of the unpatented component, part, or good was “reasonably foreseeable by an infringing competitor in the relevant market.”Footnote ¹⁰¹ Such a change would appear to be consistent with the law in several other countries, which permit the recovery of lost profits on lost sales of collateral goods subject to normal principles of proximate causation.Footnote ¹⁰²

There is one possible qualification to the argument that patentees generally should be able to recover lost profits for lost sales of unpatented products that they would have earned, but for the infringement, as long as those losses are proximately caused by the infringement. Suppose that the patentee is not making, using, or selling any products covered by the patent in suit, but rather is enforcing the patent to maintain its position in the market for the unpatented product. By their nature, patents tend to suppress competition with respect to use of technology. But it is debatable whether the enforcement of a patent merely to eliminate the use of a newer and possibly superior technology is consistent with the underlying purposes of the patent system. It is at least plausible that awards of lost profits (and injunctions) in such cases disserve the public interest. On the other hand, requiring patent owners to “work” their patents in order to recover lost profits would be at odds with the traditions of countries such as the United States, which generally has eschewed such requirements. In addition, it might be easy for patent owners to circumvent such a requirement by engaging in some token use of the technology covered by their patents. We therefore recommend further research on the frequency with which patent owners seek to enforce “idle” patents, and of the appropriate legal rules for addressing such conduct.Footnote ¹⁰³

4 Apportionment

One topic that is particularly significant for complex, multifunctional products is whether the patentee is required to quantify the portion of its lost profits that are attributable to the patented feature(s), as opposed to unpatented aspects and other components of the larger product. This is known as apportionment.

Historically, apportionment was an important issue in early U.S. patent cases.Footnote ¹⁰⁴ For example, in Seymour v. McCormick,Footnote ¹⁰⁵ the Court distinguished between a patent directed to an entirely new machine and a patent that merely claimed an improvement on an existing machine. In the former situation, a patentee “would be entitled to the entire lost profit on any sales to an infringer because those sales would have necessarily gone to the patentee.”Footnote ¹⁰⁶ In contrast, the Court held that it would be a “grave error” to award similar damages for “an improvement of small importance when compared with the whole machine.”Footnote ¹⁰⁷ Thus, Seymour “recognized that if patent damages were not calculated after apportioning value between the patented invention and the prior art,” it would overcompensate the patentee.Footnote ¹⁰⁸

Other courts and scholars have concluded that “apportionment is not required … because patentees need only show ‘but for’ causation to recover lost profits.”Footnote ¹⁰⁹ For instance, in W. L. Gore & Associates v. Carlisle Corp., the court held that “once the fact that sales have been lost has been proven, there is no occasion for the application of apportionment.”Footnote ¹¹⁰ In other words, the “but for” standard for lost profits largely obviates the need for further apportionment.Footnote ¹¹¹

The U.S. Court of Appeals for the Federal Circuit recently addressed the issue of apportionment for lost profits awards in Mentor Graphics Corp. v. EVE-USA, Inc.Footnote ¹¹² Both parties in Mentor Graphics made and sold emulation and verification systems, which are software programs that allow one computer system to act like another, ordinarily noncompatible, system.Footnote ¹¹³ These emulation systems, which are used by chipmakers like Intel to test semiconductor designs, are highly complex and expensive. The patented technology at issue covered a method and apparatus for debugging chip designs by inserting “test probes” with the ability to measure “intermediate values” in a series of logic gates.Footnote ¹¹⁴ This feature was later incorporated into the infringing emulators.Footnote ¹¹⁵

At trial, the jury awarded the patentee over $36 million in lost profits.Footnote ¹¹⁶ On appeal, the infringer asserted that the verdict should be overturned because the district court had failed to apportion the amount of lost profits “to cover only the patentee’s inventive contribution.”Footnote ¹¹⁷ While agreeing that “apportionment is an important component of damages law generally” and that it is “necessary in both reasonable royalty and lost profits analysis,” the Federal Circuit rejected the infringer’s argument, holding that “apportionment was properly incorporated into the lost profits analysis … through the Panduit factors.”Footnote ¹¹⁸ Specifically, it explained that Panduit’s first two requirements – “that patentees prove demand for the product as a whole and the absence of non-infringing alternatives” – appropriately “ties lost profit damages to specific claim limitations and ensures that damages are commensurate with the value of the patented features.”Footnote ¹¹⁹ Under the facts in Mentor Graphics, the lost profits analysis under Panduit was straightforward; the relevant market contained two suppliers (the patentee and the infringer), there was one purchaser (Intel), and there were no acceptable noninfringing alternatives, so each sale made by the infringer necessarily resulted in a lost sale to the patentee.Footnote ¹²⁰ In addition, the infringer did not dispute any of this evidence on appeal.Footnote ¹²¹ As a result, the Federal Circuit held that “satisfaction of the Panduit factors satisfies the principles of apportionment.”Footnote ¹²²

We believe that the Federal Circuit’s approach in Mentor Graphics is correct. If the infringement caused the patentee to lose sales, the principle that patentees should be made whole requires that the patentee recover the profits it would have earned on these lost sales, even if the patented feature is only one aspect of a more complex product.Footnote ¹²³ By contrast, an infringing sale that does not displace a patentee’s sale should result in a reasonable royalty, where the value of other, non-patented features is considered in determining the royalty.Footnote ¹²⁴ As the Federal Circuit noted in Mentor Graphics, courts may grant mixed awards of lost profits and reasonable royalties in cases where some but not all lost sales are due to the infringement;Footnote ¹²⁵ this is particularly likely in cases involving complex, multicomponent products, where different customers value different features in their purchasing decisions. These mixed awards obviate the need for courts to engage in further apportionment of lost profits to cover only the value of the patented feature.

5 Potential Recovery for Other Harms

A final consideration is the availability of damages to compensate for other types of harms suffered by the patentee as a result of infringer’s unlawful competition that fall outside the categories previously discussed.

Although as a general matter tort law aims to restore the victim to the position it would have occupied had the tort never occurred, legal systems throughout the world often impose substantial limits on this restorative principle, both to reduce the costs of adjudication and to vindicate other social policies. Similarly, if the goal of the patent system were to fully restore the patent owner to the position it would have occupied but for the infringement, courts would permit the patentee to recover not only its lost profits on lost sales due to the infringement, but also compensation for any other proven and quantifiable harms so caused, such as: (1) future losses that the patentee may suffer due to the infringer’s accelerated entry into the market; (2) losses to the patent owner’s goodwill or reputation, or to the prestige of the goods embodying the patented invention, due for example to consumers confusing the infringer’s product with the patentee’s; (3) lost profits at subsidiaries of the patentee;Footnote ¹²⁶ (4) lost profits due to cost increases from lost economies of scale; (5) the opportunity cost of having to devote time to litigation, advertising or marketing expenses incurred in response to the infringement; and (6) emotional harms resulting from the infringement.

For a variety of reasons, however, it is unlikely that any legal system would award damages for all of these losses, even if the patent owner were able to prove them. For example, no legal system of which we are aware allows patent owners (or other tort victims, for that matter) to recover damages for their opportunity costs of having to devote time to litigation – a result that probably is sound, given the substantial difficulties that would surround the accurate quantification of such losses.

On the other hand, ordinary principles of proximate causation in Anglo-American jurisprudence and counterpart doctrines elsewhere do not necessarily preclude patent owners from recovering for some of these other losses where provable. For example, courts in the United States and elsewhere have approved awards of lost profits resulting from the infringer having gained an accelerated foothold in the marketplace as a result of its infringement, though such losses can be difficult to prove and awards do not appear to be common.Footnote ¹²⁷

As for injury to goodwill, reputation, or prestige, as well as emotional harms, Article 13(1) of the EU Enforcement Directive states that in setting damages for the infringement of IP rights, the judicial authorities of member states “shall take into account all appropriate aspects, such as the negative economic consequences, including lost profits, which the injured party has suffered, any unfair profits made by the infringer and, in appropriate cases, elements other than economic factors, such as the moral prejudice caused to the rightholder …”Footnote ¹²⁸ In its 2016 Liffers decision, the Court of Justice of the European Union held that under Article 13(1), an IP owner can recover damages for moral prejudice in addition to a reasonable royaltyFootnote ¹²⁹ – though what “moral prejudice” means in the context of patent infringement cases remains somewhat unclear. (Liffers itself was a copyright case, and it refers to the possibility that moral prejudice may include injury to the author’s reputation.) According to Fox et al. (Reference Fox, Berghuis, vom Feld and Orlando2015), in the patent context, the concept has met with varying interpretations throughout the EU:

Similarly, a recent patent infringement decision of the Court of Appeal of Madrid held that “moral damages” – including “psychological suffering or distress, which is considered to exist in a variety of situations such as psychological or spiritual shock or suffering, helplessness, worry (as a mental sensation of disquiet, sorrow, fear or foreboding uncertainty), anxiety, anguish, uncertainty, shock, affliction and other similar situations” – are in theory compensable (although the patent owner had not proven the facts alleged in support of them), as well as damages for “loss of prestige,” which were awarded based on evidence that the infringing products were of lower quality than the plaintiff’s, having been presented “in simple cardboard boxes as opposed to the luxury image attributed to the products of the complainant.”Footnote ¹³¹ In the United States, by contrast, while damages for harm to goodwill or reputation resulting from patent infringement are in theory compensableFootnote ¹³² (though again, apparently rare), emotional harms probably are not.Footnote ¹³³

A standard law-and-economics account of proximate causation suggests that infringers who have breached a duty of care should not be responsible for losses having a low ex ante probability of occurring because the imposition of liability in such cases would increase adjudication costs without materially decreasing the (already low) risk of harm.Footnote ¹³⁴ Whether this account (or other accounts of) proximate cause counsel in favor of more generous awards of damages for “moral prejudice” in patent cases has not been much addressed (to our knowledge) in the scholarly literature; neither has the related topic of the extent of proof that should be necessary to recover for and quantify such losses, assuming they are compensable at all. Further research on these issues may be warranted.

A final question related to this body of issues is whether courts should award damages, in patent or other cases, for “loss of chance” – i.e., for the profits that would have been earned on lost sales, discounted by the probability that those sales would have been made but for the infringement. For example, if the plaintiff can prove that there was a 30 percent chance it would have made ten more sales, under the loss of chance doctrine it would be entitled to recover 30 percent of the profit it would have earned on those sales. Courts in some countries award patent owners lost profits on this basis, but the United States does not. Rather, in the United States, the patent owner would recover no damages unless it could prove that it more likely than not suffered the loss (i.e., that the probability was greater than 50 percent).Footnote ¹³⁵

The principal argument in favor of awarding damages for loss of chance is that such a rule results in more accurate compensation to patentees in the aggregate. For example, if a patent owner could show that it had a 40 percent chance of making each of one hundred individual sales, under the U.S. rule it would recover no lost profits, even though it is likely that it would have made at least some of the one hundred sales. Conversely, if the owner can show that it had a 60 percent chance of making each of the hundred sales, it would recover lost profits for all one hundred, even though it is likely it would not have made all of them. On the other hand, one might question whether courts (or juries) are well positioned to make such finely grained probability determinations, and even if they are, whether the additional cost of adjudication would be justified by the marginal accuracy gains. Further research in this area might be of more than merely theoretical interest.

1 North America

In the United States, historically, recovery of the infringer’s profits was possible as an equitable remedy for patent infringement.Footnote ¹⁴⁸ The Patent Act of 1819 created federal jurisdiction for actions in equity under the patent laws, thus authorizing federal courts to issue injunctions and other equitable relief, including “an equitable account of the infringer’s illicit profits.”Footnote ¹⁴⁹ The Patent Act of 1870 explicitly extended disgorgement to actions at law as well, providing that “the claimant shall be entitled to recover … the profits to be accounted for by the defendant.”Footnote ¹⁵⁰

However, in the Patent Act of 1946, Congress dropped all references to the infringer’s profits.Footnote ¹⁵¹ The legislative history suggests that Congress was concerned by the time and expense needed to calculate the infringer’s profits, which in some cases took many years of litigation.Footnote ¹⁵² In Aro Manufacturing Co. v. Convertible Top Replacement Co., Justice Brennan of the Supreme Court concluded that the purpose of the 1946 amendment was “precisely to eliminate the recovery of profits as such and recovery of damages only,”Footnote ¹⁵³ in an opinion concurred in by a total of four of the nine Justices.Footnote ¹⁵⁴ Although it was “only a plurality opinion and arguably constituted dictum,”Footnote ¹⁵⁵ later court decisions have interpreted the amendment in the same manner.Footnote ¹⁵⁶ Therefore, most commentators believe that an award of the infringer’s profits is not possible for utility patents in the United States.Footnote ¹⁵⁷

For design patents, however, Section 289 of the Patent Act provides that the infringer shall “be liable to the owner to the extent of his total profit, but not less than $250.”Footnote ¹⁵⁸ The Patent Act of 1887 introduced a provision to the same effect,Footnote ¹⁵⁹ and when Congress abolished the recovery of the infringer’s profits in 1946, they retained the special “total profit” provision for design patents.Footnote ¹⁶⁰

While Section 289 makes it unlawful to manufacture or sell an “article of manufacture” to which a patented design or a colorable imitation thereof has been applied and makes an infringer liable to the patent holder “to the extent of his total profit,”Footnote ¹⁶¹ in the case of a design for a multicomponent product, a question arises how to identify an “article of manufacture”: whether it must always be the end product or it can also be a component of the product. In a case involving the infringement of designs for smartphones, the U.S. Court of Appeals for the Federal Circuit took the former interpretation and identified the entire smartphone as the only permissible “article of manufacture” for the purpose of calculating the infringer’s “total profit” because “[t]he innards of Samsung’s smartphones were not sold separately from their shells as distinct articles of manufacture to ordinary purchasers.”Footnote ¹⁶² However, in Samsung v. Apple, the Supreme Court recently reversed the Federal Circuit’s judgment and remanded the case.Footnote ¹⁶³ In a unanimous opinion written by Justice Sotomayor, the Court stated that “the term ‘article of manufacture’ is broad enough to encompass both a product sold to a consumer as well as a component of that product.”Footnote ¹⁶⁴ Thus, the Supreme Court adopted the interpretation under which a patent holder would “sometimes be entitled to the infringer’s total profit from a component of the end product.”Footnote ¹⁶⁵ The Court left it up to the lower courts to determine how to define the relevant “article of manufacture” and how to calculate the profit attributable to that article.Footnote ¹⁶⁶

In Canada, a successful patentee is entitled to damages, but may request an accounting of the infringer’s profits.Footnote ¹⁶⁷ The grant of an accounting is within the discretion of the court, though it is normally granted when sought, in the absence of some reason why it should not be permitted.Footnote ¹⁶⁸ There are no fixed criteria for denying an accounting, though traditional equitable criteria will be considered.Footnote ¹⁶⁹ In practice, “an accounting of profits has been the dominant monetary remedy for patent infringement in Canada,” with at least twelve reported decisions since 1990.Footnote ¹⁷⁰

2 Europe

In the EU, Article 13 of the IP Enforcement Directive (2004/48/EC) stipulates:

1. 1. Member States shall ensure that the competent judicial authorities, on application of the injured party, order the infringer who knowingly, or with reasonable grounds to know, engaged in an infringing activity, to pay the rightholder damages appropriate to the actual prejudice suffered by him/her as a result of the infringement. When the judicial authorities set the damages:

As the provision requires the judicial authorities of Member States to “take into account … any unfair profits made by the infringer,” a state “arguably would be in compliance with the Directive if it merely permitted courts to consider the defendant’s profit in estimating the plaintiff’s own ‘actual prejudice.’”Footnote ¹⁷² The Commission Staff Working Document “Analysis of Enforcement Directive” published in December 2010Footnote ¹⁷³ states:

In Germany, damages awarded for patent infringement are intended to be compensatory in nature and may be recovered for negligent or intentional infringement.Footnote ¹⁷⁷ Surrender of the infringer’s profits has been generally accepted as a method for calculating patent infringement damages for decades and, in 2008, its availability was codified in Section 139, para. 2 of the German Patent Act.Footnote ¹⁷⁸ Under this provision, a patentee may choose among three methods for calculating damages (i.e., lost profits (actual damages suffered), license analogy (royalty), and the infringer’s profits), but may not combine or cumulate them with respect to any single act of infringement.Footnote ¹⁷⁹

Traditionally, patentees rarely elected to calculate damages by reference to infringers’ profits, primarily because courts liberally allowed infringers to deduct production costs from the revenue they earned on sales of infringing products. However, the decision by Federal Court (BGH) in the Gemeinkostenanteil case in 2000Footnote ¹⁸⁰ brought about a fundamental change. In that decision, the Court held that, while the variable costs of manufacturing and marketing the infringing products may be deducted, an infringer’s fixed costs may no longer offset its revenue. Though this case dealt with infringement of design rights, courts have applied the same rule to damages awarded for patent infringement. Post-Gemeinkostenanteil, Germany has seen a marked increase in requests for damages based on alleged infringers’ profits.Footnote ¹⁸¹

In France, Code de la propriété intellectuelle Article, L. 615–7, provides for two different methods of calculation: actual damages and an analogy to licenses. It stipulates that “[t]o set the amount of damages, the court distinctly takes into account: the negative economic consequences of the infringement, including the loss of earnings and any loss suffered by the injured party, the moral prejudice caused to the latter and the profits made by the infringer, including the savings in intellectual, material and promotional investments that it achieved from the infringement.”Footnote ¹⁸² It is said that “[t]he place of the infringer’s profits in the calculation of damages is not yet clear in case law,” and “[s]ome decisions have considered that the claimant can be granted the infringer’s profits, while others take a different position.”Footnote ¹⁸³

In the United Kingdom, “it is standard to allow the successful patentee to elect for either an inquiry as to damages or an account of profits for past infringements.”Footnote ¹⁸⁴ The patentee is entitled to limited disclosure of the infringer’s financial information in order to choose between damages or profits.Footnote ¹⁸⁵ An account of profits is an equitable and restitutionary remedy whose purpose is to deprive the infringer of the profits that it has improperly made by wrongful acts committed in breach of the claimant’s rights and transfer those profits to the claimant.Footnote ¹⁸⁶ Requesting of profits is said to be a “much rarer choice than requesting damages because the outcome is much more uncertain.”Footnote ¹⁸⁷

3 Asia

Article 102 of the Japanese Patent Act provides for three special methods for calculating damages: methods using the patentee’s profit margin, the infringer’s profits, or hypothetical royalty. Paragraph 2 of that Article provides: “Where a patentee or an exclusive licensee claims against an infringer compensation for damage sustained as a result of the intentional or negligent infringement of the patent right or exclusive license, and the infringer earned profits from the act of infringement, the amount of profits earned by the infringer shall be presumed to be the amount of damage sustained by the patentee or exclusive licensee.” Though courts initially limited this form of damages to patentees that were practicing the patented invention, the Grand Panel of the Intellectual Property High Court softened this requirement in 2013, holding that disgorgement is available anytime the patentee lost profits as a result of infringement, even if those profits did not result from lost sales of goods or services covered by the patent-in-suit.Footnote ¹⁸⁸ For example, it is now generally accepted that disgorgement is an available remedy for patentees that sell unpatented products that compete with the infringing products supplied by the infringer. It is still unclear and disputed, however, whether patentees that only license the patent-in-suit to third parties may request disgorgement as a remedy. Moreover, when disgorgement is awarded, at least some courts have apportioned the infringer’s profits to reflect the percentage of infringing sales attributable to the infringing feature.Footnote ¹⁸⁹

Section 65 of the Patent Act of China provides for several methods for determination of the amount of damages: the actual loss incurred by the patentee, the infringer’s profits, and reasonable royalties – though, in practice, the vast majority of patentees in Chinese patent suits pursue statutory damages.Footnote ¹⁹⁰ Though disgorgement is rarely awarded, the Chinese Supreme People’s Court (SPC) has held that an infringer’s profits may be calculated by multiplying the profits per unit of infringing product and the quantity of the infringing products that have sold in the market.Footnote ¹⁹¹ The SPC also held in 2009 that disgorgement awards may be apportioned “to deduct profits led by factors other than the infringed patent from the whole amount of the infringing profits.”Footnote ¹⁹²

Australia follows a similar approach to the United Kingdom and Canada and permits disgorgement of an infringer’s profits.Footnote ¹⁹³ The leading authority makes clear that, like in the United Kingdom, a patentee may elect an accounting in lieu of seeking monetary damages.Footnote ¹⁹⁴

1 Availability

We were unable to reach consensus regarding the question of whether disgorgement should be available in all major patent systems. We acknowledge the divergence in approaches between jurisdictions on this issue,Footnote ¹⁹⁵ as well as the competing policy arguments for and against disgorgement. Group members who favor disgorgement point to several potential benefits. First, it creates an incentive for potential infringers to engage in ex ante licensing of patent rights, while not being as serious of a sanction as punitive damages. While an accounting will not incentivize a truly innocent infringer (i.e., an infringer that was unaware of the patent-in-suit and could not have reasonably identified all patents covering a product prior to market entry)Footnote ¹⁹⁶ into negotiating ex ante, it can incentivize a negligent or deliberate infringer to do so. Second, disgorgement may be advantageous for a patentee who wishes to avoid divulging financial information to a competitor in a damages assessment. This appears to be a motivation for pharmaceutical companies routinely electing an account in countries such as Canada, even when damages based on the patentee’s loss would be greater.Footnote ¹⁹⁷

In contrast, group members who are less enthusiastic about disgorgement as a remedy point to several potential drawbacks of the remedy. First, disgorgement may create a significant risk of over-deterrence, causing firms to be less willing to introduce new and innovative products, particularly complex products that incorporate numerous different technologies.Footnote ¹⁹⁸ In addition, non-compensatory damages awards like disgorgement threaten to exacerbate the holdup problem.Footnote ¹⁹⁹ Furthermore, there may be substantial litigation costs associated with calculating the amount of profit due to infringement that is to be disgorged.Footnote ²⁰⁰ Finally, disgorgement in the context of complex, multifunction products may amplify the risk of error in calculating a remedy, as courts will be required to determine the share of infringer’s profit due to infringement of patented feature(s), as opposed to unpatented or licensed components, as well as the infringer’s own contributions to the infringing product.

In light of this lack of consensus, the recommendations in the rest of this Section are oriented at jurisdictions where disgorgement is an accepted remedy. In countries like the United States that do not currently award disgorgement as a remedy, we propose further research on whether an accounting may be desirable as an alternative to enhanced damages to deter willful infringement.

2 Discretionary or As of Right

In jurisdictions where disgorgement of the infringer’s profits is available as a remedy, one question is whether it should be awarded automatically or at the trial court’s discretion. Though an accounting is an equitable remedy, in UK law, a successful patentee is entitled to an accounting if it so elects.Footnote ²⁰¹ Similarly, when the remedy was available in U.S. law, it appears to have been routinely granted, perhaps as of right. In Canadian law, in contrast, the remedy is clearly discretionary, though it is normally granted.Footnote ²⁰²

In view of the potential burden on the infringer in taking an accounting, particularly in complex product cases, we recommend that in jurisdictions that choose to allow an accounting, the grant of accounting be within the discretion of the court. One factor that should be considered is whether an accounting will place an undue burden on the infringer, recognizing that the costs of discovery in an accounting are likely to fall disproportionately on the infringer.Footnote ²⁰³ That is, an accounting should not be granted when it is used primarily as a tool to harass the infringer.

This is not to say that we are recommending that an accounting be granted sparingly. It should also be recognized that when damages are sought, particularly in the form of lost profits, the discovery burden will fall disproportionately on the patentee, and the desire to avoid that burden is an entirely legitimate reason for the patentee to elect an accounting in lieu of lost profits. As explained in more detail below, the patentee will often pay an implicit price in the form of foregone damages when electing an accounting,Footnote ²⁰⁴ and this is an inherent disincentive to abuse of the accounting remedy. The concern that an accounting will unduly burden the infringer will be further mitigated in jurisdictions with limited discovery. Therefore, an accounting should not be denied solely because of the burden it places on the infringer, but only when that burden is disproportionate to the amount at issue, as compared with the alternative of assessing damages.

We would also emphasize that the concern for the burden on the infringer is not necessarily the only consideration that should be taken into account in determining whether an accounting should be granted. We focus on this point because it is one lesson to be taken from the history of the accounting remedy in U.S. law. Because an accounting of profits is not available in the United States and it is not the primary remedy in most jurisdictions in which it is available, the broader question of when an accounting should be granted has not received sustained attention, either in the cases or in the literature. We propose further research on this issue.

3 Calculating the Infringer’s Profits