The Innocent Bystander Problem in the
Patenting of Higher Life Forms
Norman Siebrasse”
The innocent bystander problem is one of the key issues in the
controversy over the patenting of higher life forms. The problem arises
when a patented life form enters adventitiously onto the property of a
person who is unaware of its presence. At present it is clear that lack of
intent is not a defence to a claim of patent infringement; the questions
addressed in this aicle are whether the Patent Act should be amended
to provide an “innocent bystande ” defence in such circumstances, and
if so, what should be the bounds of such a defence.
The argument for amending the Patent Act turns on the issue of
innocent bystanders who benefit from their use of the invention. The
primary argument in favour of introducing an exemption from patent
liability for benefiting bystander turns on a concern for autonomy of
the bystanders, while the mar argument in favour of retaining liability
in such circumstances is based on the need to protect patent incentives.
The author analyzes the problem of famer autonomy using economic
theory and concludes that while there is a legitimate autonomy interest
at stake, it is not particularly strong. The article also assesses the
treatment of analogous problems in other areas of law, namely the law
of tangible property and unjust emichnent and notes that the law
generally does not protect the same autonomy interest in other contexts.
Turning to the problem of patent incentives, the article notes that
impairment of patent incentives resulting from truly innocent use is
likely to be muintal, but an innocent bystander exemption from
liability may nonetheless have significant adverse effects because
enforcement difficulties resulting from the need to establish intent may
facilitate intentional infrugement The article then describes in detail
the structure of an innocent bystander defence that would best balance
these competing concems. It concludes that the net benefits provided by
the innocent bystander defence would be minimal and would not justify
the additional legal complexity and costs of patent enforcement, which
even the best form of the defence would entail
Le problbrne du spectateu
innocent est l’un des problumes
cruciaux qui concement la question du brevetage des formes de vie
sup&ieures. Le problne se pose lorsqu’une forme de vie brevetdo
p6nire de fagon inopine sur la propr rilt d’une personne qui n’en
actuel des choses, il est
remarque pas la prsence. Dana l’6t
incontestable que l’absence d’intention ne constitue pas une d6fense
contre une requite en violation de brevet. La question que pose cet
article consiste A savoir si la Loi sur les brevets devrait dre arend6e
afin de permettre une d6fense de aspectateur innocent. dam de telles
circonstances et, si tel est le eas, quelles devraient en Ore les balises.
innocents repose sur le respect de
L’argurnent pour amender la Loi sur les brevets repose pluti stir
le cas den spectateurs innocents qui tirent bdndfice de lenr uilisation de
l’invention. L’argument primordial militant pour l’adoption d’une
exemption A la respousabilit6 pour violation de brevet qui protbge les
spectateurs
lenr autonomie.
L’argument principal en faveur de la position inverse se fonde sur la
nicessit6 de protdger les incitatifs inh&ents an brevet. L’auteur analyse
le problmne de l’autonomie du spectateur innocent sous l’angle de la
thiore 6conomique et conclut que s’il est vrai qu’il existe l& un int6rt
d’autonomie
ligitime, celui-ci est plutt faible. L’article 6value
6galement le traitement de problbmes similaires dans d’autres champs
du droit, N savoir le droit des biens tangibles t l’enrichissenrnt sans
cause, et monue que le droit ne protbge g6nlraleonent pas ces int6r&ts
d’autonomie similaires dam d’autres contextes. Quant 4 la question des
incitatifs, lauteur constate que le tort qui lenr semit causd par un usage
vdritablement innocent serait probablement minimal Cependant, une
exception exondrant les spectaeurs innocents pourrait tout de rinle
avoir d’inportants effats pervers, puisque les dimficultds d’application
d6coulant de la ndcessit6 d’6tablir ne intention faciliteraient les
violations intentionnelles.
L’article d crit ensuite en detail la strncure d’une defense basde
sur la th6orie du spectateor innocent qui 6quilibrerat le mieux possible
ces intb6rts divergents. L’article conclut que les bdnlfices conf6ris par
une defense do
innocent* seraent minimaux at ne
justifiemient pas la complexit6 juridique additionelle et les cots
d’application de ce regime de brevet modifi, ce que requrait namae la
meilleure version de la defense.
pectateur
“Associate Professor, Faculty of Law, University of New Brunswick. I would like to thank Mitchell
Mclnnes for his invaluable comments on the unjust enrichment section of this article, and Keith
Culver for many discussions on the subject of the article as well as comments on earlier drafts. The
author would also like to acknowledge the support of two Genome Canada funded projects: the
Genome Atlantic Potato Genome Project and the Transgenic Animals Project of the Canadian
Program in Genomics and Global Health.
McGill Law Joumal 2004
Revue de droit de McGill 2004
To be cited as: (2004) 49 McGill L. 349
Mode de r6f6rence : (2004) 49 R.D. McGill 349
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
Introduction
I. Anatomy of the Problem: Four Categories of Cases
I1. Potential Sources of Insight
A. CBAC Report
B. Remedial Innocent User Provisions in Other Jurisdictions
II1. The Problems
Introduction
A.
B. The Patent Incentive Problem
C. The Problem of Farmer Autonomy
IV. Innocent Bystander Defence
Introduction
A.
B. Availability of the Defence
C. “Depth” of the Defence
D. Substantive Limits of the Defence
E. Legal Scope: Tangible Property Claims
F Summary: Fashioning the Defence
Conclusion: Desirability of an Innocent Bystander Defence
351
354
360
360
361
364
364
365
368
378
378
378
381
384
387
389
389
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
Introduction
farmer, for
is suing Percy Schmeiser, a Saskatchewan
The innocent bystander problem is one of the key issues in the controversy over
the patenting of higher life forms. As the Federal Court of Appeal explained in its
recent decision in Monsanto Canada v. Schmeiser,’ the problem arises when a
patentee sues for patent
infringement when the patented plant “[has] come
fortuitously onto the property of a person who has no reason to be aware of the
presence of the characteristic created by the patented gene.”‘2 In the Schmeiser case,
Monsanto
infringing
Monsanto’s patent on its Roundup Ready (herbicide tolerant) canola by growing the
patented canola from seed that apparently originally blew onto his property.
Schmeiser himself was clearly an intentional infringer; though the seed may have
entered adventitiously, when Schmeiser discovered the patented canola on his land, he
intentionally selectively harvested the patented seed and used it in planting the next
year’s crop.’ Nonetheless, the court of appeal was sufficiently concerned about the
innocent bystander problem that it suggested that lack of intent should be made a
defence in a case in which the defendant had no knowledge of the nature of the crop,
even though it has long been clear law that intent is not an element of patent
infringement.’ Similarly, although the Canadian Biotechnology Advisory Committee
(“CBAC”) recommended the patenting of higher life forms, it proposed that the
Patent Ace be amended to provide protection from patent infringement claims for
such “innocent bystanders”.’ The innocent bystander problem was also cited as a
concern by the Supreme Court in its decision in Harvard College v. Canada
(Commissioner of Patents),’ which held that higher life forms per se are not
patentable!
Despite the holding in Harvard Mouse that higher life forms per se are not
patentable, the innocent bystander problem remains a live issue. In part this is because
of the possibility of legislative change to allow for the patentability of higher life
forms. More immediately, patents related to higher life forms, such as gene patents,
‘ (2002), [2003] 2 EC. 165, 218 D.L.R. (4th) 31, 2002 FCA 309, leave to appeal to S.C.C. granted,
[2002] S.C.C.A. No. 437 [Schmeiser (C.A.)].
2 Ibid. at para. 57.
‘Ibid. at para. 58.
4 See infra note 21 and accompanying text.
‘ R.S.C. 1985, c. P-4.
6 Canadian Biotechnology Advisory Committee, Patenting of Higher Life Forms and Related
Issues: Report to the Government of Canada Biotechnology Ministerial Coordinating Committee
(Ottawa: Canadian Biotechnology Advisory Committee, 2002) at 14 [CBAC Report].
‘(2002), 219 D.L.R. (4th) 577, 21 C.P.R. (4th) 417, 2002 SCC 76 [Harvard Mouse].
‘See especially ibid. at para. 171.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
are apparently valid,9 and such patents raise the innocent bystander problem just as
directly as patents for higher life forms per se.”
This article presents a policy analysis of the innocent bystander problem. In
particular, it analyzes the desirability of a substantive intent-based exemption to patent
liability, exemplified by the CBAC recommendation and the suggestion of the Court
of Appeal in Schmeiser. This article deals only with adventitious entry of patented life
forms that benefits the bystander, resulting in an action by the patentee against the
bystander. Escape that harms the bystander, resulting in a suit by the bystander against
the patentee or other party responsible for the escape, is a distinct question that would
be a matter for negligence, trespass, or nuisance law. In particular, the innocent
bystander defence should be distinguished from a possible duty to prevent escape.
The latter would be at issue in the case of escape resulting in harm and is not dealt
with in this article.”
This article assumes that the overarching goal of the law should be to deliver the
greatest net social benefits. This is, of course, the traditional approach to patent policy,
which considers patents to be justified only to the extent that the innovation they
foster brings benefits to society at large.’2 This article argues that farmer autonomy
should be treated in the same manner; farmer autonomy is not valued for its own sake,
but because respecting the autonomy of the farmer allows for the socially efficient
allocation of resources. Thus we will not ask whether the farmer’s right is inherently
9 While the Patent Office initiated the litigation in the Harvard Mouse case by refusing to grant a
patent for a mouse per se, it does allow patents on specific genes and individual cells containing those
genes. See e.g. the patent at issue in Can. Patent No. 1,313,830 for Glyphosate-Resistant Plants
(issued 23 February 2003). The Supreme Court accepted this distinction, as it accepted that a
genetically altered egg, and by implication, other individual cells, would be a patentable “composition
of matter”, even though the higher life form to which it gives rise is not: Harvard Mouse, ibid. at para.
162.
,0 For convenience this article will refer to “patents for higher life forms” to encompass both patents
related to higher life forms and patents for higher life forms per se, as there are no functional
differences between the two types of patents that are relevant to the innocent bystander problem.
Similarly, a plant or seed carrying a patented gene will be referred to as a “patented plant” or
“patented seed”.
” A strict duty to prevent escape, which could be relied upon by a fanner as a defence to a patent
action, would simply reduce to an innocent bystander defence, and need not be considered separately.
A negligence-based duty, which could be similarly relied upon, would not aid the farmer as most
escape is nonnegligent.
‘” This has been express in the law of patents since the earliest times: see e.g. Darcy v. Allin (1602),
Noy 173, 74 E.R. 1131 (sub nom. The Case of Monopolies, 11 Co. Rep. 84b) and The Clothworkers
of Ipswich Case (1615), Godb. 552, 78 E.R. 147 (K.B.), which expressed the view that the grant of a
patent monopoly is justified only by the good that the patentee brings to the country as a whole. The
same view is express in art. I, 8, cl. 8 of the United States Constitution, which gives the federal
government the authority to grant patents in order to “promote the Progress of Science and useful
Arts.”
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
353
superior to that of the inventor, but rather which combination of rights will bring
about the greatest net social benefit. More particularly, the analysis is primarily from a
law and economics perspective, buttressed by insights from related areas of law, in
particular property law and the law of unjust enrichment.’3
Because the innocent bystander problem is most likely to arise in the agricultural
context” this article uses farming as the main exemplar of an activity that implicates
patents on higher life forms. Similarly, the licencee or infringer of a patent related to
higher life forms is taken to be a farmer. It should nonetheless be understood that the
general discussion applies equally to horticulture, aquaculture, forestry, or any other
industry based on higher life forms.
This article proceeds as follows. The next Part sets out the anatomy of the
problem under existing law. In particular, we will see that the generally applicable
liability rule addresses the most pressing concerns regarding the escape of higher life
forms, since under existing law only benefiting innocent bystanders will be liable for
substantial damages. Thus the problem only really arises in respect of innocent non-
benefiting farmers. While there is an autonomy-based argument to be made that even
benefiting farmers should be relieved of liability if the use is innocent, the argument is
not nearly as compelling as it would be if non-benefiting farmers were also liable
under existing law. Part II deals with potential sources of insight, namely the CBAC
Report and remedial innocent user provisions in other jurisdictions, which turn out to
be unhelpful. Part HI sets out the fundamental tension raised by the problem of
innocent benefiting bystanders, namely the need to protect patent incentives on the
one hand while respecting farmers’ autonomy on the other. This sets the stage for Part
IV, which describes the innocent bystander defence that would best address this
tension. We will see that any innocent bystander defence would have to be relatively
limited in scope in order to reduce the potential for abuse. The article concludes by
addressing the ultimate question: is an innocent bystander defence desirable? Even
” Apart from the issue of the desirability of an innocent bystander exception to liability, the analysis
assumes that encouraging inventions related to higher life forms is desirable and that patents are
necessary to this end. If this were not true then the appropriate response would be to bar patents
related to higher life forms entirely and the innocent bystander problem would be moot. Of course,
individual inventions may be potentially harmful for health or environmental reasons, but patents are
only a right to exclude others and do not give a positive right to practice the invention. Issues related
to health and safety concerns are regulated under separate legislation: see e.g. Seeds Regulations,
C.R.C., c. 1400, s. 110, which regulates release of seeds of plant with novel traits into the
environment; and the Food and Drug Regulations, C.R.C., c. 870, Division 28, which regulates the
safety of genetically modified foods.
” Escape of patented animals is foreseeable, as in the case of escape of farmed salmon. The
innocent bystander problem, however, requires that the life form escape into the custody of someone
who benefits thereby, such as another farmer or aquaculturist. This is more likely in the case of plants
than in the case of animals, because the proximity of fields makes the escape of genetic material
through natural mechanisms, such as seed spread or cross-pollination, quite likely.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
the best possible innocent bystander defence is not necessarily the best possible
solution to the innocent bystander problem, as it raises some problems while solving
others. To make the case for introducing an innocent bystander defence, we must
assess whether the patent incentive problems the defence would create are worse than
the farmer autonomy problems it would solve. In other words, is the cure worse than
the disease? The article concludes that the net benefits provided by even the best
possible innocent bystander defence would be minimal and would not justify the
additional legal complexity and costs of patent enforcement that it would entail.
I. Anatomy of the Problem: Four Categories of Cases
A proposed innocent bystander exception to patent liability is, as the phrase
indicates, an exception to the general liability rule. In order to decide whether such an
exception is warranted we need to define both the exception and the general rule to
which it is an exception. A proposed innocent bystander defence must be assessed in
the context of the best possible general rule. Otherwise, the inadequacies of the
general rule rather than the merits of an intent-based defence will drive the analysis.
At present it is clear law that intent is not an element of patent infringement, so that
even an innocent bystander will be an infringer. The situation is more subtle at a
remedial level, since a technical infringer will not always be liable for substantial
damages.
This article takes the general remedial rule to be that the farmer will be liable, at
most, to the extent of the benefit she derives from the patented crop, regardless of
knowledge or intent. The benefit is determined by a test of “but for” causality, so that
the farmer is liable only to the extent of profit made that is over and above that profit
she would have made but for the use of the patented crop. Suppose, for example, an
entire field were planted with a patented crop. If the farmer would have made $100
per hectare profit if the field had been planted with non-patented seed, and she in fact
made $120 per hectare profit, the patentee would be entitled to no more than $20 per
hectare, as that is the amount of the profit causally attributable to the use of the
patented seed. 5 This will be referred to as the “‘but for’ benefit-based approach”, or
“benefit-based approach” to liability.
” The full $20 might be awarded if an accounting of profits were granted. Damages would normally
be less, as the patentee and a licencee would negotiate a licence fee that would split the profit
attributable to the use of the invention between them: see generally Norman Siebrasse, “A Remedial
Benefit-Based Approach to the Innocent User Problem” (2004) 20:1 C.I.P.R. 79. An award of
damages, rather than the harsher remedy of an accounting of profits, would usually be more
appropriate in the case of a innocent bystander. For the purposes of this article, however, this is not an
essential detail. The key point is that in any event, the patentee would not be entitled to more than the
profits that are causally related to the use of the patent.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
This benefit-based approach is taken as the general rule for two reasons. First, in
my view, it represents the best statement of existing law. It should be acknowledged
that considerable uncertainty exists on this point as a result of the decision of the
Federal Court of Appeal in Schmeiser. In that case, the Court of Appeal awarded the
patentee the entire profit made from the use of the invention, not just the “but for”
profit-$120 per hectare, rather than $20 per hectare, in our simple example. I have
argued at length elsewhere that the Court of Appeal’s decision is wrong as a matter of
existing law because it ignores the causation requirement that is fundamental to
private law liability.’6 Presumably, the current state of the law will be resolved when
the appeal in Monsanto v. Schmeiser is decided by the Supreme Court of Canada. In
any event, the second and more important reason for positing this general rule is that
it directly responds to the primary purpose of patent law, which is to provide
appropriate incentives to produce socially valuable inventions. A rule that gives the
patentee either more or less than the benefit which is caused by the use of the
invention will provide too much or too little incentive to invent.”
For its part, the proposed innocent bystander defence would relieve a user from
liability for patent infringement based on factors related to the intent or knowledge of
the user. For this reason it may be referred to as an “intent-based” approach to the
innocent bystander problem. The nature of the proposed defence will be defined more
precisely in Part V of this article. For now, these descriptions of the general liability
rule and the innocent bystander exception, though cursory, are sufficient to allow us to
organize the discussion by dividing cases according to whether the use is intentional
or innocent, and whether the farmer does or does not benefit. The possible scenarios
then fall into four broad categories, depending on the combination of intent and
benefit, as shown in Figure 1.
Two of these quadrants represent easy cases where the correct answer is
intuitively clear. It is important to recognize that the benefit-based general rule gives
the same result as an intent-based exception in both of these categories.
Quadrant I is the case where a farmer actively introduces the patented crop onto
his property without a licence and thereby benefits. In this case the result is the same
regardless of whether the innocent bystander defence is available; the general rule
would hold the farmer liable for the benefit gained, and the defence would not apply
because the use is intentional.
16 Siebrasse, ibid.
‘” See Part III(A), below.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
Intentional
Benefiting
Farmer
Intentional 4
Intentional
Non-Benefiting
Farmer
Benefiting
I
II
in
IV
Non-benefiting
Figure 1
Innocent
Benefiting
Farmer
Innocent
Non-Benefiting
Farmer
Innocent
Quadrant IV represents an equally easy case, but one that is much more germane
to our discussion. This is the case where a farmer’s field is contaminated with
patented seed but the fanner neither knows of, nor benefits from, the contamination.
This is likely to be a very important scenario in practice. Many types of patented
plants will require special treatment to take advantage of their properties. Herbicide
tolerant canola, such as in Schmeiser, for example, permits better weed control by
allowing the farmer to use herbicide to kill weeds after the canola has emerged. Using
this weed control technique on ordinary canola would kill the canola itself, and so this
method would never be used by a farmer who was unaware of the nature of the crop.’8
” Similarly, a pest-resistant plant will not provide any significant benefit to a farmer who is not
aware of its nature, as the farmer will use the usual amount of pesticides and get very little cost
reduction. So-called “Bt” crops are engineered to incorporate a gene derived from the micro-organism
Bacillus thuringiensis, which causes the production of an insecticidal protein: see e.g. U.S. Patent No.
5,608,142 “Insecticidal cotton plants”; Can. Patent Application No. 2,348,755 “Polypeptide
Compositions Toxic to Diabrotica Insects, Obtained from Bacillus Thuringiensis; CryET70, and
Methods of Use” (PCT Filing Date 29 October 1999). There may be some benefit from the extra pest
resistance provided by the gene.
20041
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
More generally, in any case in which the patented plant requires special treatment to
provide a benefit, an innocent bystander will also be a non-benefiting user, since she
would not know to adjust her farming techniques to take advantage of the patented
plant. And even inventions intended to provide an advantage without special treatment
are likely to provide little or no benefit when present as a small number of plants in a
field dominated by a different variety.”
Again, the general benefit-based remedial rule gives the same result as would the
intent-based defence in this category of cases. While the innocent bystander defence
would be available, it would not be needed, since under the benefit-based approach
the farmer would technically infringe, but would not be liable for more than nominal
damages, as the “but for” benefit from the use of the invention is zero.
The two remaining quadrants represent much harder cases. The case of the
innocent benefiting bystander (quadrant II) arises when the user did not intentionally
bring the patented product onto his land, yet the nature of the invention is such that
the user would benefit nonetheless. This type of problem is likely to arise in respect of
inventions that do not require special treatment to provide a benefit. An example
would be a crop engineered to give an unusually high yield under normal growing
conditions. This type of case presents a conflict between the need for patent incentives
and the desire to preserve the autonomy of the farmer. While there is a case for
relieving the farmer from liability, it is not nearly as intuitively compelling as is the
case of the innocent non-benefiting farmer.
Suppose that the invention is such that individual plants or animals have a high
value in a specific market. An example might be a pig genetically engineered to have
organs that are immunologically compatible with humans, so that the pig organs can
be transplanted into humans and are very valuable for this purpose.”0 At the same
time, the meat tastes like that of any other pig and is worth exactly the same in the
meat market. If the farmer never discovered the unique properties of the pig and sold
it in the ordinary market for meat, neither approach would hold him liable (quadrant
IV). But suppose the farmer had discovered the unique properties just at the time that
the animal was ready for market and sold it in the specialized market (for organ
“9 See text accompanying notes 42-43, below.
20 Adventitious entry of a pig might seem unlikely, but it might occur if the farmer inseminated a
sow with purchased semen inadvertently derived from a pig with a patented gene. Similar examples
may arise with plants engineered to produce commercially valuable molecules such as vaccines. This
is known as “plant molecular fanning”. See Canadian Food Inspection Agency, CFIA Multi-
Stakeholder Consultation on Plant Molecular Fanning: October 31-November 2, 2001 (Interim
report), online: Canadian Food Inspection Agency
Patent No. 6,476,290 “Transgenic tilapia comprising humanized insulin gene”. A less exotic example
of a life form that is unusually valuable in a particular market is naturally decaffeinated coffee: see
Shinjiro Ogita et al., “Producing Decaffeinated Coffee Plants” (2003) 423 Nature 823.
358
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
transplant), for the same purposes and in the same market as the patentee? Here the
farmer would have benefited directly and substantially from the patentee’s efforts. In
these circumstances the view that the farmer should be relieved from liability, though
not indefensible, is not nearly as compelling as if the farmer had sold the pig for meat.
This example is not a conclusive argument, nor is it intended to be. The point is
simply that the case for relieving even an innocent farmer from liability is much
stronger when the farmer does not benefit than when the farmer does benefit. It would
be wrong to allow the appeal of relieving an innocent non-benefiting farmer from
liability to lead us into a broad rule that would also relieve an innocent benefiting
farmer from liability. This point deserves emphasis; when calls are made to relieve the
innocent bystander from liability, the implied context is often that of a non-benefiting
innocent bystander. Consider the example given by the Federal Court of Appeal in
Schmeiser:
[Ilt seems to me arguable that the patented Monsanto gene falls into a novel
category. It is a patented invention found within a living plant that may, without
human intervention, produce progeny containing the same invention. It is
undisputed that a plant containing the Monsanto gene may come fortuitously
onto the property of a person who has no reason to be aware of the presence of
the characteristic created by the patented gene. It is also reasonable to suppose
that the person could become aware that the plant has that characteristic but
may tolerate the continued presence of the plant without doing anything to
cause or promote the propagation of the plant or its progeny (by saving and
planting the seeds, for example). In my view, it is an open question whether
Monsanto could, in such circumstances, obtain a remedy for infringement on
the basis that the intention of the alleged infringer is irrelevant.2′
The Court of Appeal’s suggestion that in this case the user should not be liable is
intuitively appealing. But note that no express mention is made of whether the person
in question benefited from the fortuitous presence of the patented gene. Since the
particular invention in question in Schmeiser was herbicide tolerant canola, so that an
innocent farmer would not derive any benefit from the invention, the implicit context
of the Court of Appeal’s remarks was that of an innocent non-benefiting fanner. We
should not infer that the Court would have made the same suggestion in respect of a
benefiting farmer.
On its facts, Schmeiser apparently falls in quadrant
the intentional non-
benefiting farmer.2 This is another difficult case. The general rule of benefit-based
I,
2, Supra note 1 at para. 57. With respect, it is not an open question as to whether Monsanto could
obtain a remedy for infringement against a farmer in such circumstances. As the cases cited by the
Court of Appeal make abundantly clear, intent is not an element of a patent infringement action: ibid.
at para. 56. The Court’s remark is better understood as a suggestion for patent reform than as a
statement of existing law.
22 The defendant in Schmeiser was clearly an intentional infringer, but it is not clear from the facts
whether he benefited. Since the issue of benefit was held to be irrelevant both at trial (Monsanto
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
359
liability
in such a case, yet there is
liability would not result in substantial
considerable appeal to the view that intentional infringement of a patent should not
escape unpunished. A farmer who intentionally grew the patented seed might intend
to benefit from it by selling it illicitly to other farmers, instead of taking advantage of
its unique properties himself. Since the propagation is much easier to detect than the
illicit sale, and there is no obvious legitimate reason for intentional infringing use, it
might be sensible to impose a fine or other punitive sanction for intentional
propagation, rather than requiring that the patentee detect the actual sale. On the other
hand, it may be that injunctive relief, which in effect imposes punitive sanctions for
repeated intentional use, is adequate.
Whether punitive sanctions for intentional use are desirable is beyond the scope
of this article. At this point we should recognize that the issue of the innocent non-
benefiting infringer and the intentional non-benefiting infringer are closely connected
in the Court of Appeal’s decision in Schmeiser. The Court of Appeal in Schmeiser
wanted to impose liability on the intentional non-benefiting farmer, while relieving
the innocent non-benefiting farmer from liability. Both these positions are defensible,
but the Court’s approach was problematic. In depriving Schmeiser of all his profit,
regardless of the proportion that was causally related to the use of the patent, the
Court asserted a punitive general rule of liability. Given this general rule, the Court
then had to suggest grafting an intent-based exception onto it in order to deal with the
innocent bystander. The better approach is to begin with a benefit-based general rule
and, if desired, graft a punitive rule onto it to deal with the intentional user.
This illustrates the importance of assessing the intent-based exception to liability
in the context of the correct general rule. The Court of Appeal’s suggestion that an
exception to the background liability rule might be needed to deal with the innocent
(non-benefiting) infringer was driven by the fact that it had set out a punitive general
rule. Since punitive damages in the absence of intent are contrary to general legal
principles, it is no wonder the Court felt that an exception from liability might be
needed in the case of an innocent bystander. But if a compensatory general rule is
applied rather than a punitive one, the need for an exception is much less compelling.
In summary, while there is a strong case for relieving the innocent non-benefiting
farmer from liability, there is no need for an intent-based exception to do so. It is
important that the case of the innocent non-benefiting user not be invoked in favour of
the intent-based exception from liability, as doing so makes the argument for such an
exception seem more compelling than it really is. The analysis in this article is
premised on a compensatory benefit-based background liability rule and asks whether
intent-based deviations from this rule should be implemented in the case of escaped
higher life forms. There is no need for such an exception in the easy cases-the
Canada Inc. v. Schmeiser (2001), 12 C.P.R. (4th) 204, 202 FT.R. 78 (T.D.) [Schmeiser (T.D.)]) and in
the Court of Appeal (supra note 1 at paras. 79-80), it is clear that the result would have been the same
had there been an express finding that the defendant had not benefited.
360
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
intentional benefiting farmer and the innocent non-benefiting farmer-as the general
rule gives substantially the same result as would an innocent bystander defence. The
innocent benefiting farmer of quadrant Id represents a much harder case, which is
analyzed in the balance of this article.
II. Potential Sources of Insight
A. CBAC Report
The most obvious starting point in our policy analysis of a possible innocent
bystander defence is the CBAC Report, which recommended that such a defence be
introduced
is
fundamentally flawed, and in consequence it is of no assistance in analyzing the
problem.
law.23 Unfortunately,
into Canadian patent
the report’s analysis
The entirety of the CBAC Report’s argument for an innocent bystander defence is
found in the following paragraph:
Currently, patent law does not require a patent holder to prove that an alleged
infringer knew or even ought to have known about the reproduction of a
patented invention. This situation places individuals without knowledge of the
reproduction of a patented plant, seed, or animal on their property or in their
care in a difficult situation. That individual (the “innocent bystander”) may face
a patent infringement suit–one of the most difficult and expensive legal
actions against which to defend-and damages for infringement without a
countervailing remedy against the patent holder. While in theory such an
individual may be able to sue for negligence for the adventitious spread of the
plant or seed or the reproduction of the animal, the practical difficulties of
doing so-proving a duty of care and a breach of that duty-may make this
remedy illusory.24
This paragraph has a number of problems, but the most serious is that the
argument is a non sequitur. Suppose we were concerned, not with intellectual property
rights in patented plants or animals, but with physical property rights in ordinary
animals. Asking whether a patentee should be entitled to sue for the benefit conferred
by the escape of patented plants is analogous to asking whether a pig farmer should
be entitled to sue for the return of pigs that escaped onto her neighbour’s land. If the
pigs did no harm at all, the neighbour would have no remedy against the pigs’ owner.
We would never suggest, though, that because the neighbour had no remedy against
owner, he should be entitled to keep the pigs. Nonetheless, this is exactly the structure
of the argument in the CBAC Report.
23Supra note 6.
24 Ibid. at 14. This passage was echoed by the majority of the Supreme Court in Harvard Mouse,
supra note 7 at para. 172.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
Now suppose that the escaped pigs did some harm, perhaps by destroying some
crops. In that case, it is true that the neighbour should have a remedy against the pigs’
owner, but again we would not suggest that the remedy should be that the neighbour
can keep the pigs. This is because the harm to the neighbour and the proposed
“remedy” are entirely unrelated. Perhaps the value of the pigs is just compensation for
the loss of the crops, but it is more likely that it is far too much or far too little.
The point here is not that the neighbour should not have a remedy, but that the
remedy should be tailored to the harm suffered. Even if the neighbour did not have a
remedy in the case where the pigs destroyed his crops, the obvious law reform
response is to provide a remedy that compensates for the harm done, not to allow him
to keep the pigs.
The analogy between the pigs and volunteer crops is a close one. In the vast
majority of cases a farmer will not suffer any harm by the adventitious entry of
patented plants. Herbicide tolerant canola grows as well as ordinary canola, so having
some mixed into the crop will not harm the farmer. In some cases a farmer may be
injured by contamination from the patented crop. The most obvious case would be
that of an organic farmer who might not be able to market contaminated crop as
organic. It is true that the organic fanner in this case will find it difficult to recover in
negligence because of the difficulty in establishing a breach of duty. But this is hardly
fatal to the farmer’s claim, since the obvious cause of action is nuisance.” Admittedly
a nuisance action might also fail, since success in nuisance depends on the
reasonableness of the claim.6 But it is not clear that there is anything wrong with this
result. It is far from obvious that a farmer who decides to set up an organic farm in the
midst of neighbours who uniformly use genetically modified crops should be able to
sue his neighbours or the seed supplier for the contamination of his field. In any event,
the fairness of recovery in nuisance is directly the province of the law of nuisance. If,
as a policy matter, we were to decide that the existing law of nuisance is
unsatisfactory in this respect, surely the appropriate response is to modify nuisance
law rather than to introduce a radically new doctrine into patent law.
B. Remedial Innocent User Provisions in Other Jurisdictions
The law in other jurisdictions is another potential source of insight. While intent
is irrelevant in Canadian patent law, this is not universally the case. The UK,
Australia, New Zealand, and the US all have provisions that prevent a patentee from
recovering damages and/or an accounting of profits against an infringer who did not
Trespass would probably fail on the fault requirement: see John G. Fleming, The Law of Torts, 9th
ed. (Sidney: The Law Book Company, 1998) c. 3 at 46ff. (“Intentional Invasion of Land: Basis of
Liability”).
26 Ibid. at 466.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
have notice of the patent.” It might be thought that these provisions could shed useful
light on the innocent bystander problem. Unfortunately, that is not the case.
The remedial innocent user provisions are of no direct aid to the innocent
bystander for two reasons. First, the defendant must establish lack of knowledge of
the existence of the patent. 8 Since the farmer typically knows of the patent, even if he
does not know that he is using the invention, the farmer would not be able to take
advantage of the provision as it stands. More substantively, an injunction remains
available
the innocent bystander has no control over
adventitious entry, if the patentee did obtain an injunction, the farmer might be unable
to refrain from breaching it, and so would suffer punitive sanctions for something
beyond her control.’
to the patentee. 9 Since
While these provisions would not be helpful to the innocent bystander, they at
least suggest that the notion of incorporating intent into patent law is not entirely
without precedent. The requirement that the farmer be ignorant of the patent might be
only an artifact of the way in which innocence was expressed. And it might seem a
short step from relieving the innocent farmer from monetary liability to eliminating
liability altogether. Thus we might hope that these provisions and the rationale behind
them shed some light on the innocent bystander problem. It turns out that this hope is
not realized, as the US provision is quite irrelevant to our problem, while the policy
rationale of the other Commonwealth provisions is hopelessly obscure.
the US provision
The policy underpinning
is quite clear. It bears some
resemblance to an innocent user defence, as in some cases it requires the patentee to
give the infringer actual notice of the patent before damages are available. This
appearance is misleading. It is a marking requirement, intended to encourage copying
of unpatented aspects of a new product by allowing the copier to identify exactly
which aspects are patented and which are not. If the product is marked as required,
the user’s knowledge or lack thereof is irrelevant, so that damages are available even if
the user had never seen the product and had no knowledge of the patent.’ The section
27 Patents Act 1977 (U.K.), 1977, c. 37, s. 62; Patents Act 1990 (Cth.), s. 123 (Austl.); Patents Act
1953 (N.Z.), 1953/64, 37 RS 509, s. 68; U.S.C. tit. 35 287.
28 For example, the New Zealand act provides that “damages or account of profits shall not be
awarded against a defendant who proves that at the date of the infringement he was not aware, and
had no reasonable ground for supposing, that the patent existed … ” (ibid., s. 68(1)).
29 This is stated explicitly in subsection 123(3) of the Australian Patents Act 1990 (supra note 27)
and in subsection 68(4) of New Zealand’s Patents Act 1953 (supra note 27). It is implicit in the UK
act (supra note 27, s. 62), which refers only to damages and an accounting of profits.
3 Eliminating the injunction, on the other hand, amounts to eliminating liability entirely, and so
reduces this remedial provision to the complete innocent bystander defence, which will be analyzed
below.
“, This is clear on the face of the provision and is affirmed by Wine Railway Appliance v. Enterprise
Railway Equipment, 297 U.S. 387 at 395 (6th Cir. 1936) [Wine Railway].
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
363
is also inapplicable as a defence to a claim for damages if the invention is such that it
is not capable of being marked, as in the case of a process or method, 2 or if the
product is never sold at all.” The innocence of the user is incidental; the goal of the
provision is to encourage the patentee to mark the product, and the user is relieved
from liability only to provide an incentive to this end.
The UK and New Zealand provisions were also plausibly intended by the
legislature as marking provisions.’
In the UK, at least, the provision has been more
generously interpreted, so that the issue is the knowledge of the defendant, and
innocence is only one factor to be considered.” The Australian provision is clear on its
face that it is open to the defendant to establish that it was not aware of the patent,
even if the product is properly marked and openly marketed.” Thus the UK and
Australian provisions are truly innocent infringer provisions, as marking is not
determinative but is only a factor to be taken into account in determining whether the
infringer was in fact innocent.
Unfortunately, cases applying the provisions in the UK, Australia, and New
Zealand are few, and there is no judicial discussion of the rationale for the defence.
Cases that do apply the provision interpret it in a textual fashion, with no reference to
the underlying policy.7 No criticism of the courts is intended, but the result is that we
32 See e.g. American Medical Sys. v. Medical Eng’g Corp., 6 E3d 1523 at 1538 (Fed. Cir. 1993).
33 See Wine Railway, supra note 31.
3 The Patents Act 1953 (N.Z.) (supra note 27, s. 68(1)) provides that “a person shall not be deemed
to … have had reasonable grounds for supposing [the patent existed] by reason only of the application
to an article of the word ‘patent’ [or] ‘patented’ … unless the word or words are accompanied by the
words ‘New Zealand’ or the letters ‘N.Z.’ and by the number of the patent.” The text of the UK
Patents Act 1977 (supra note 27, s. 62(1)) is substantially the same. The “unless” in each of these
provisions suggests that when the requisite marking is present the user shall be deemed to have had
reasonable grounds for supposing the patent existed, regardless of the user’s actual state of
knowledge. This is characteristic of a marking provision.
” See Lancer Boss Ltd. v. Henley For klifi Co. Ltd. and H. & M. Sideloaders Ltd. (1973), [1974]
FS.R. 14 at 27, [19751 R.PC. 307 (Ch.D) [Lancer Boss cited to ES.R.] (though the infringer in that
case was a copier and consequently was not permitted to take advantage of the defence). There is
some suggestion in New Zealand law to the same effect, though the point is by no means clear. See
e.g. Ashmont Holdings Ltd. v. Jurox Pty Ltd., [2001] 2 N.Z.L.R. 130 at paras. 25-26 (N.Z.H.C.)
(Lexis).
” In subsection 123(2) the Patents Act 1990 (supra note 27) states that “the defendant is to be taken
to have been aware of the existence of the patent unless the contrary is established.’
” There is some suggestion in the case law that the provision would not be available to someone
who directly copies the patentee’s invention. This would support the view that the rationale may have
been to protect those who did not benefit from the patentee’s efforts. These are, however, mere hints
in the cases and are not sufficient to establish the rationale for the provision. See e.g. Lancer Boss,
supra note 35 at 27; Texas Iron Works Inc.’s Patent (1997), [2000] R.P.C. 207 at para. 8.4 (Pat. Ct.)
(suggesting that copying weighs against the availability of the defence). In TK Valves Ltd. v. Hindle
Cockburn Ltd. (13 January 1989) Ch. (Lexis)), the defence was held to be available on the facts and it
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
have no guidance as to the substantive policy considerations that gave rise to the
provisions. At most, we can say that these provisions indicate that there may be some
room for intent in patent law, but we have already accepted as much, and these
provisions provide no additional insight.
Ill. The Problems
Introduction
A.
Intent is not, at present, an element of patent infringement. To decide whether this
should be changed by introducing an innocent bystander defence, we must define
both the problem the defence would be intended to address and the problems it may
create. That is the aim of this Part.
This Part will first address the basic argument against introducing an innocent
bystander defence. The difficulty arises from the patent incentive problem. To provide
socially appropriate incentives for invention the patentee should be entitled to retain
the entire benefit from the use of the patent, regardless of the user’s intent. But some
departure from this strict principle is acceptable and indeed inevitable, and the
distortion of incentives that would result if truly innocent bystanders were entitled to
retain the benefit of the use of the patent would not be serious in practice. Thus the
fact than an innocent bystander defence would relieve innocent bystanders from
liability for infringement is not objectionable from a patent incentive perspective. But
this does not mean that an innocent bystander defence is unobjectionable. The real
problem is more subtle. The main difficulty with an innocent bystander defence, as
we will see, is that it may create evidentiary problems which make it difficult for the
appears that the defendant’s product was developed independently, though this is not entirely clear
from the facts. In Wilderman v. F W Berk & Co. Ltd. ((1924), [1925] Ch. 116, 42 R.EC. 19) it was
held that the defence was available on the facts, but the case was decided on the basis that
infringement was not made out in any event. In Benmax v. Austin Motor Co. ((1953), 70 R.P.C. 143
(Ch.D.)), the defence was held not to be available on the facts, as the defendant should have been
aware of the possible existence of a patent. In Lux Traffc Ltd. v. Pike Signals and Faronwise Ltd.
([1993] R.P.C. 107 (Pat. Ct.)), the defendant successfully argued innocent infringement, even though
it appears that the defendant may have copied the invention from the patentee (though this is by no
means clear). This aspect, however, of the defence was not contested, and the trial judge noted, “I
therefore do not need to resolve any dispute as to fact or law on the application of s. 62 of the Act”
(ibid. at 144). There is almost no relevant caselaw construing the equivalent Australian and New
Zealand provisions. In the unreported Australian decision in Woodbridge Foam Corporation v. AFCO
Automotive Foam Components Pty Ltd. ([2002] FCA 883), the only question was whether innocent
infringement was sufficiently pleaded. The respondent pleaded innocence on the basis that the
infringing product had been supplied to it by a third party and it had no notice of any complaint. The
court held that sufficient material facts had been pleaded in support of the defence though the court
remarked, “To be frank, I doubt whether the respondent could make out the statutory defence on those
‘facts’ even if they are true” (ibid. at para. 11).
20041
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
365
patentee to enforce its rights against intentional users. The difficulty is not in the
direct intended effect, but in the unintended consequences.
The discussion then turns to the other half of the equation. The argument in
favour of the innocent bystander defence is founded on a concern for farmer
autonomy. The notion of farmer autonomy will be unpacked and recast in terms of
respect for the farmer’s property rights, which will allow us to assess the problem in
terms of our overarching goal of improving the overall social welfare. We will see that
the real threat to farmer autonomy is not the obvious one that farmers may be liable
for the benefit gained from innocent use, as liability for a benefit that was never
earned or anticipated does not trench on one’s autonomy. The problem is more subtle;
it is errors in the assessment of the benefit that can potentially have an impact on
farmer autonomy. But while such errors could pose a real threat to farmer autonomy,
in practice they are unlikely to be serious and similar errors are tolerated in other areas
of law.
B. The Patent Incentive Problem
The justification for holding a user liable to the extent of any benefit obtained
from the use of a patent, whether the use is innocent or otherwise, is straightforward:
to do otherwise will undermine the incentives for innovation established by the Patent
Act. The Patent Act provides incentives to develop new inventions by allowing the
patentee to reap the benefit of the invention for a limited period of time.” It is
significant that the Patent Act does not grant a fixed reward to the patentee. Instead,
the inventor is given a property right, which allows the inventor to reap the benefit of
the use of the patent by exacting licence fees. This makes the reward from the
invention proportional to the social benefit conferred by the invention; the more useful
the invention, the more people are willing to pay for it, and the greater the incentive to
invent. To the extent that the patentee cannot capture all of the benefit flowing from
the invention during the term of the patent, the incentives will be distorted, as
resources will be directed to inventions with the highest privately appropriable benefit
rather than the greatest social benefit.
Conversely, the same rationale implies that liability should not be imposed on any
non-benefiting user, whether innocent or intentional. Patent incentives that are too
great are just as socially harmful as inadequate patent incentives. If the patentee is
entitled to damages that exceed the benefit causally related to the invention, inventors
will have an inefficiently high incentive to invent and patent new plants and animals
as the private benefit will be greater than the social benefit. Excessive investment in
inventive activity means that resources will be devoted to marginal inventions that are
not worthwhile as compared with alternative uses for those resources. At the same
time, if non-benefiting users are liable, then costs in the users’ enterprise will be
3 The term of a patent in Canada is twenty years from the filing date: Patent Act, supra note 5, s. 44.
366
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
greater than the true social cost of the activity, and incentives to engage in that
enterprise will be inefficiently low. Just as the inventors will have inadequate
incentives to invent if they cannot capture the full social benefit of their labour, so
farmers will have inadequate incentives to farm if some of the fruits of their labour are
transferred to patentees. In short, making a non-benefiting farmer liable will result in
too many worthless inventions and not enough worthwhile farming.
This is why the causation link that is fundamental to liability generally is
particularly appropriate in the patent context. By ensuring a causal link between the
use of the invention and the remedy, incentives to invent are made to correspond to the
social benefit conferred by the invention. Equally, this is why the remedial calculation
used by the Court of Appeal in Schmeiser is wrong from a policy perspective;
awarding the patentee an amount greater than the benefit that flows causally from the
invention-perhaps many times greater-distorts patent incentives just as seriously as
would a denial of liability.
To summarize, the “but for” benefit-based approach provides appropriate patent
incentives, which are neither too little nor too great. For this reason this article takes
the benefit-based approach to be the appropriate general rule that would apply in the
absence of an innocent bystander defence.
With that said, the need to preserve patent incentives
is not an absolute
imperative. For a number of reasons the benefit that can be captured by a patentee is
always less than the full benefit to society. Most obviously, the limited term of a
patent means that the social benefit accruing after the expiry of the patent is not
appropriable by the patentee. And limitations in the ability of the patentee to price
discriminate mean that many users will value the patent at more than its cost. This
“consumer surplus” means that the patentee cannot capture the full social benefit of
the invention even during its term. 9 This is not to say that because some distortion of
incentives is inevitable, any distortion is acceptable. But it is clear that some departure
from full capture of all social benefit is acceptable. The seriousness of the patent
incentive problem depends, therefore, on the significance of these incentive effects. If,
in practice, the innocent bystander defence results in the loss of only a small
proportion of the patentee’s potential revenues, incentives will not be greatly affected.
” The “consumer surplus” is the difference between the price a consumer would be willing to pay
for the product and the price charged by the vendor. “Price discrimination” occurs when the vendor
charges different prices to different groups of consumers (for example, airline discounts for advance
bookings, or senior discounts for movie tickets) in order to capture some of the consumer surplus.
“Perfect price discrimination” occurs when the vendor is able to charge each consumer the maximum
they are willing to pay. Perfect price discrimination is never administratively possible. See generally
Alfred E. Kahn, The Economics of Regulation: Principles and Institutions (Cambridge, Mass.: The
MIT Press, 1988) vol. 1 at 131-33.
2004]
N. SIEBRASsE – THE INNOCENT BYSTANDER PROBLEM
367
How significantly would patent incentives be affected if innocent bystanders were
not liable to disgorge any benefit they gained? There are a number of ways in which
patented higher life forms might enter adventitiously onto the property of a bystander,
depending on the precise nature of the life form. Seed might simply fall out of a torn
bag and fly off a passing truck that is carrying patented seed. Gene flow is possible,
with the incidence depending on the species.’ In the case of canola, a relatively weedy
species that depends in part on wind-borne and insect pollination, gene flow by cross-
pollination is possible.4 ‘ All of the most likely sources of adventitious entry will result
in relatively low levels of contamination. Even gene flow from outcrossing of plants
with wind or insect-borne pollen will normally be restricted to the immediate physical
area of the patented plants. So, in the case of canola, most cross-pollination occurs
within 50 to 100 meters of the plants in question, with cross-pollination reduced to
negligible levels beyond 200 meters. 2 Thus only the portion of the fields immediately
bounded by a field planted with patented canola would be affected.
The adventitious entry that does occur will not often deliver a significant benefit.
Many kinds of engineered crops require the farmer to adopt special farming methods
if any benefit is to be realized, and adventitious entry of these varieties will provide no
advantage. Differences between the volunteers and the dominant variety plant may
mean that an engineered plant will provide no benefit even if it would be capable of
delivering a benefit under normal conditions if it were the dominant variety-
volunteer canola in a field of barley is a weed no matter how desirable the particular
variety of canola might be when planted on its own. Even a difference between
varieties of the same crop might rob the volunteer of any benefit. Crops have long
been bred for specific traits such as time to maturity, or disease resistance, which may
vary significantly between different varieties of the same crop. Patented genes for
traits such as higher yield or herbicide tolerance are introduced into a number of
different varieties, to allow farmers a similar choice of varieties as would be available
for ordinary plants. Volunteers engineered, for example, for an exceptionally high
yield, would provide no benefit if they reached maturity at a different time from the
planted crop, since they would be under or over-ripe at harvest. 3 Further, if the entry
4 Adventitious entry is probably less likely with animals than with plants, since more secure
containment methods are used for animals. For example, in the case of salmon aquaculture, while
escapes certainly occur, they are generally the result.of some gross mechanical failure in a particular
net pen. Escaped salmon will generally not be able to enter neighbouring pens that remain secure.
4″ See Canadian Food Inspection Agency, Plant Products Directorate, Plant Biosafety Office,
Decision Document DD95-02: Determination of Environmental Safety of Monsanto Canada Inc.’s
Roundup@ Herbicide-Tolerant Brassica napus Canola Line G773, esp. Part IV.2 “Potential for Gene
Flow to Wild Relatives Whose Hybrid Offspring May Become More Weedy or Invasive”.
42 Telephone Interview with Philip MacDonald, National Manager, Environmental Release
Assessment Plant Biosafety Office, Canadian Food Inspection Agency (26 January 2004).
“3 It is true that a variety planted in one field is more likely to be compatible with a variety planted in
a neighbouring field than with another variety planted at random. The point remains, however, that
368
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
of the gene occurred by cross-pollination, rather than by entry of the patented seed
itself, the progeny will not generally have desirable traits. Half of the offspring will
carry the patented gene, but this is not enough to deliver benefits since the general
characteristics of the plant in which the gene appears must also be desirable; a high-
yield gene in a disease-susceptible plant may be less advantageous than a lower
yielding plant with disease resistance. A cross between two different varieties will not
normally have a desirable combination of characteristics. It may sometimes occur that
the two cross-pollinating varieties are the same, except for the patented gene, since
some varieties are more suited to specific geographic regions and there is therefore
some likelihood that neighbouring farmers will be growing exactly the same variety.
But even then, many desirable varieties into which an engineered gene is inserted are
first generation (F1) hybrids.” First generation hybrids do not breed true, so the
offspring will not have the same desirable characteristics as the parents.
In summary, most instances of truly adventitious entry will provide little or no
benefit to the farmer, and it follows that the loss of this benefit is unlikely to have any
significant effect on patent incentives. However, it would be wrong to conclude from
this that an innocent bystander defence is unobjectionable from the patent incentive
perspective. All that we can conclude is that a perfectly tailored and completely
enforceable innocent bystander defence would be unobjectionable. No legal rule is
perfectly defined or perfectly enforceable. The problem with an innocent bystander
exception is not that it exempts truly innocent bystanders from liability, but that it
opens up legal space that would allow room for illicit intentional propagation.
Intentional use, where a farmer systematically uses and profits from patented seeds,
could undoubtedly have a serious effect on patent incentives. In other words, the key
question in considering the innocent bystander defence does not involve innocent
bystanders at all. The question is whether an innocent bystander defence can be
crafted in a way that does not unduly facilitate intentional propagation. It is no
coincidence that the only litigation to date, Monsanto v. Schmeiser, has been in
relation to intentional propagation. This is consistent with the view that it will rarely
be worthwhile for a patentee to sue for benefits from truly adventitious entry and that
the main goal of the patentee is to prevent intentional propagation, which poses a
much greater threat to patent incentives.
C. The Problem of Farmer Autonomy
The argument for relieving the innocent benefiting farmer from liability is based
on a concern for farmer autonomy. If liability can arise simply because pollen blows
into the farmer’s field, without any volitional act on the part of the farmer, liability
many chance variables must be aligned in order for volunteer plants to provide any significant benefit
to the farmer in whose field they enter adventitiously.
” This depends on the crop. For example, most corn varieties are F1 hybrids, while this is true of
only some varieties of canola.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
in
the
that raw form we will be faced with
may be unavoidable. But while concern for autonomy undoubtedly identifies a key
aspect of the problem, it is an analytically unsatisfactory resting place. If we leave the
argument
task of balancing
incommensurable factors, namely the need to provide incentives for new inventions
on the one hand and respect for autonomy and property on the other. In this section
we will try to unpack this concern for property and autonomy and recast it into terms
that are more amenable to comparison with the patent incentive problem.
Autonomy is generally thought to be concerned with freedom of action-it is
essentially prospective, concerned with the ability to plan and carry out one’s plans.”
More specifically, the autonomy problem is manifested in interference with the
farmer’s property rights. If liability attaches to unintentional use, farmers will have
partially lost the right to determine the use of their own land.
How then does liability for innocent benefiting use constrain the farmer’s ability
to plan? Under the general “but for” liability rule the farmer will be liable only for the
benefit that he gained over and above that profit he would have made without the use
of the patented crop. Assume for the moment that the benefit to the farmer from the
escaped plant can be perfectly quantified. On this assumption the autonomy problem
disappears. If adventitious entry occurs and the farmer is made to disgorge the benefit
gained from the use of the invention, he will be placed in exactly the same position he
would have been in had the entry not occurred at all. This means that the farmer
should make exactly the same plans whether or not he considers the possibility of
adventitious entry. Similarly, there is no hardship to the farmer in requiring him to
give back the excess attributable to the patent as he is made no worse off than he
would have been had he not used the patent.
This is not to dismiss the autonomy argument, as the above argument rests on the
assumption that the benefit gained by the farmer can be perfectly quantified. This
assumption is false. The benefit gained will depend on the farmer’s particular
circumstances. 6 While a court using the benefit-based approach to liability will
attempt to quantify the benefit as accurately as possible, a court can never quantify the
benefit as accurately as can the farmer herself, as the farmer has better information
regarding her own circumstances. Not all relevant information can be proven at a
reasonable cost. If a farmer purchased a licence, we know that the farmer believed, on
the basis of his own intimate knowledge of his fanning operations, that the benefit
was greater than the licence fee.
” See e.g. Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986) at 373 where he
remarks that for a person to enjoy an autonomous life, in addition to the mental capacity to plan
actions, he must have “adequate options available for him to choose from [and] his choice must be
free from coercion and manipulation by others …”
6 In the case of pest-resistant crops, for example, a farmer whose fields are pest free will benefit less
than a farmer whose fields are infested. Further, there are many varieties of any given crop, which are
suitable for different conditions and which may even require different harvesting techniques.
370
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
In other words, the autonomy problem arises, not because of the prospect of
liability for innocent use per se, but because of the prospect of errors in quantification
of liability. This argument is essentially a reprise of Calabresi and Melamed’s famous
analysis of property and liability rules in “One View of the Cathedral”. 7 In their
analysis, because a property right is protected by injunctive relief, it gives an
individual the right to keep an entitlement unless he chooses to part with it voluntarily
at a price that is satisfactory to him. A liability rule, on the other hand, allows another
individual to take the entitlement on payment of its fair value–damages-as
determined by the court.4″ Property rules are to be preferred to liability rules when
bargaining is possible, because an individual is better able to value an entitlement for
herself than is a third party such as a judge. In short, we know voluntary exchanges
are beneficial to both parties, and we should be suspicious of charging someone for a
benefit that they did not request.
This is not the end of the story. As Kaplow and Shavell have pointed out in an
important refinement, this argument for property rights is not persuasive if judicial
assessment of value, even though less accurate than those of the parties, is correct on
average.’ In that case there will be no adverse ex ante incentive effects on the
behaviour of defendants. Farmers as a group will not be discouraged; even though
some will be faced with excessive damage awards, others will receive net benefits
when the damage award is less than the benefit received. Similarly, an individual
farmer might be harmed in one season but benefit the next. In planning, the farmer
will have no reason to expect that any award against her will be too high or too low, so
she will have no reason to change her planting decisions. Prospectively, a potential
award that is expected to be accurate has the same effect as an award that is in fact
accurate.’
Conversely, though, if errors in judicial assessment of damages tended to favour
the patentee, the adverse incentive effects would be the same as if punitive sanctions
were imposed explicitly. In other words, the main threat to autonomy arises not from
the prospect of liability for innocent use per se, nor even from the prospect of judicial
error in quantification of liability, but from the prospect of systematic judicial error.
” Guido Calabresi & A. Douglas Melamed, “Property Rules, Liability Rules, and Inalienability:
One View of the Cathedral” (1972) 85 Harv. L. Rev. 1089. This article has spawned an enormous
body of commentary: see e.g. Symposium, Property Rules, Liability Rules, and Inalienability: A
Twenty-Five Year Retrospective (1997) 106 Yale L.J. 2081. A particularly clear analysis is provided
by Richard A. Epstein, “A Clear View of The Cathedral: The Dominance of Property Rules” (1997)
106 Yale L.J. 2091.
48Calabresi & Melamed, ibid. at 1092.
,’ Louis Kaplow & Steven Shavell, “Property Rights Versus Liability Rules: An Economic
Analysis” (1996) 109 Harv. L. Rev. 713.
0 If the awards were grossly inaccurate, this would increase the risk to the farmer, and so might
affect decision-making, even if the awards were accurate on average. However, there is no particular
reason to believe that judicial decisions will generally be grossly inaccurate.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
Kaplow and Shavell develop this argument to show that the case for a property rule is
strongest in the case of things that have idiosyncratic or subjective value. This is for
two main reasons. First, the difficulty of accurately assessing subjective value leads
courts to tend to limit damages to objective value rather than risk wildly inflated
subjective value or absorb large amounts of court resources in attempting to quantify
the unquantifiable.’ Secondly, it is likely that the current owner of the thing values it
more than a potential taker. The reason the owner acquired the property in the first
place is that she especially liked it. 2
This theoretical argument is consistent with our ordinary intuition that property
rights are most important in respect of personal effects; as Richard Epstein asked
rhetorically, “Do we allow one person to take the wedding ring of another simply by
paying its market value?””. The likelihood of systematic judicial error is much less in
the case of fungible goods like crops. Canola is not like a wedding ring. A farmer
values canola for its cash sale value, and not out of special emotional attachment.
While judicial assessment of the value will usually be less accurate than assessment
by the farmer in any given instance, since the court does not have the same
information regarding the specific farmer’s costs and methods, it is likely to be
reasonably accurate on average because the value depends entirely on objective
factors.
Thus the case for applying a true property rule to protect the farmer’s right to
place a value on the patented crop is relatively weak. There is one important condition
required to support this conclusion. Kaplow and Shavell’s argument that property
rights should be favoured if the current owner of a thing is likely to value it more than
the average person has a parallel in the innocent bystander context. Farmers who
licence the technology are likely to derive more benefit than non-contracting farmers,
since one reason that farmers may choose not to licence is that they do not expect to
benefit, for whatever idiosyncratic reason. 1f the benefit received by the defendant
farmer were simply measured by reference to the royalty payments made by licencing
farmers this would introduce a major source of systematic bias into the assessment of
damages. Fortunately, it is relatively easy to avoid this problem. The standard
licencing payment is the prima facie measure of damages due to lost licencing,’ but it
is clear law that this is only presumptive and special factors that may reduce an
individual farmer’s benefit must be taken into account. Most obviously, if a farmer
had only one-quarter of her field planted to a higher-yielding patented variety, she
should only be liable for one-quarter of the standard royalty. More generally, other
See Kaplow & ShaveU, supra note 49 at 730-31.
52See ibid. at 760-62.
5 Epstein, supra note 47 at 2093.
5′ See Meters Ltd. v. Metropolitan Gas Meters Ltd. (1911), 28 R.PC. 157 at 164-65 (F.C.A.).
” See General ire & Rubber v. Firestone Tyre & Rubber Ltd., [1975] ES.R. 273 at 279, 1 W.L.R.
819 (H.L.) and the discussion in Part 3 “Damages” in Siebrasse, supra note 15.
372
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
individual factors that may also reduce the benefit of a patented variety-for example,
insect resistant crops are less valuable if a field is pest free because of its favoured
location-must also be taken into account. The more fundamental point here is that in
the examples discussed by Kaplow and Shavell, the additional value placed on the
thing by the owner is the result of subjective value, which is difficult to measure in
any individual case even if it is possible to determine a correct average value. In
contrast, the reduced benefit gained by an innocent bystander is due to objective
factors that can be assessed by the court. There will be errors, but there is no reason to
think that these will be systematically biased, so long as the court does not
mechanically apply a rule that damages are equal to the standard royalty.
These conclusions from the economic theory of property are consistent with
general legal principles. Most obviously, the problem that judicially awarded damages
can never perfectly reflect the value of the right in question is hardly unique to patent
law, and it is a well-established rule that the difficulty of assessing damages is no
reason they should be denied. 6 Even in the most difficult cases, where the harm is
subjective in nature, the law now grants damages, albeit with some restrictions. 7 This
emphasizes the strength of the general rule in cases, such as patent cases, where
damages are assessed by completely objective evidence.
The law of unjust enrichment provides an even more directly relevant comparison
with the economic analysis. It developed to deal with situations in which a defendant
benefited at the plaintiff’s expense when alternative claims were unavailable or
inadequate. This is closely analogous to our problem of the innocent benefiting
bystander; a classic example of unjust enrichment is mistaken payment of money,
which might as well be termed “adventitious entry of money” into the defendant’s
pockets. Unjust enrichment is particularly pertinent not only because the patentee’s
claim for return of a benefit conferred on the defendant echoes the typical claim in
unjust enrichment, but also because the law is couched in terms of broad principles
that highlight the basic considerations of justice at issue.
We have seen that the central tension in the innocent bystander problem is that
between the rights of the person conferring the benefit and the autonomy of the
defendant. Exactly the same problem has been recognized by the Supreme Court in
the unjust enrichment context:
516 See e.g. Wood v. Grand Valley Railway (1915), 51 S.C.R. 283 at 289, 22 D.L.R. 614; McRae v.
Commonwealth Disposals Commission (1951), 84 C.L.R. 377 at 411 (H.C.A.).
“‘ At one time, the common law refused compensation for emotional distress because its subjective
nature made it difficult to quantify, but recovery is now allowed in the case of intentional infliction of
emotional distress resulting in physical symptoms or an identifiable mental illness: see Allen M.
Linden, Canadian Tort Law, 7th ed. (Markham, Ont.: Butterworths, 2001) at 53-54; Wilkinson v.
Downton, [1897] 2 Q.B. 57. Conversely, awards for non-pecuniary loss were capped by the Supreme
Court in Andrews v. Grand & Toy Alberta Ltd. ([1978] 2 S.C.R. 229, 83 D.L.R. (3d) 452) because of
the subjective nature of the loss (at 261), but were not denied altogether.
2004]
N. SiEBRASSE – THE INNOCENT BYSTANDER PROBLEM
373
The third set of tensions lies on the philosophical-policy level. The traditional
reluctance of the law to permit recovery to a plaintiff who had provided non-
contractual benefits to another was founded on a philosophy of robust
individualism which expected every person to look out after his or her own
interests and which placed premium on the right to choose how to spend one’s
money…. The new approach of general principle, on the other hand, questions
the merits of this view and the quality of justice which it entails. It shrinks from
the harsh consequences of individualism and seeks to effect justice where
fairness requires restoration of the benefit conferred.58
This tension in the law of unjust enrichment is faced most clearly in the first
element of the claim, which is the need to show a benefit to the defendant. 9 This
factor is prima facie satisfied in our context as we are concerned only with cases in
which the farmer benefits from the presence of the patented seed. But the law of
unjust enrichment also reflects the economist’s concern with preserving the autonomy
of the defendant in assessing whether he has benefited from the use of the invention.
The law of unjust enrichment parallels the economic analysis in that the defendant’s
intention is not taken into account directly. Subject to any defences, “liability is strict.
… [E]ven the totally innocent recipient has no ground for suggesting that he should
keep the enrichment.”‘ Nor is it relevant that the plaintiff could have taken measures
to prevent the enrichment, even if he was careless.’
The defendant can nonetheless argue that the “benefit” received, though perhaps a
benefit to most people, is not a benefit to the defendant:
The central problem [in the measurement of enrichment] is easy to understand.
It turns on subjectivity of value. While money is the very measure of
58 Peel (Regional Municipality) v. Canada, [1992] 3 S.C.R. 762 at 785-86, 98 D.L.R. (4th) 140 [Peel
cited to S.C.R.].
9 The other elements of an unjust enrichment claim are a corresponding detriment to the plaintiff
and the absence of any juristic reason for the defendant’s retention of the benefit: see Peel, ibid. at
784, citing Pettkus v. Becker, [1980] 2 S.C.R. 834, 117 D.L.R. (3d) 257. These elements are arguably
satisfied by the loss of licencing revenue and the need to maintain patent incentives, respectively. We
need not go into these elements in detail, as the point here is not to establish that the patentee would
be entitled to relief under unjust enrichment law, but simply to examine the way that it deals with the
tension between the plaintiff’s property rights and the defendant’s autonomy.
60 Peter Birks & Charles Mitchell, “Unjust Enrichment” in Peter Birks, ed., English Private Law
(Oxford: Oxford University Press, 2000) vol. 2 at 15.59. See also Mitchell Mclnnes, “Enrichments
and Reasons for Restitution: Protecting Freedom of Choice” (2003) 48 McGill L.J. 419 at 457-58
(“Liability in unjust enrichment is prima facie strict … Assuming proof of an enrichment and a
corresponding deprivation, the grounds for relief normally proceed without reference to the recipient’s
participation, acquiescence, or knowledge.”) As discussed in the text, this is subject to any defences,
in particular the defence of change of position.
61 See e.g. Kelly v. Solari (1841), 152 E.R. 24 (Ex. Ct.); Lord Goff of Chieveley & Gareth Jones, The
Law of Restitution (London: Sweet & Maxwell, 1993) at 126-27. There is an exception if the benefit
was conferred with conscious disregard for the consequences, as when the plaintiff has entered into a
compromise of a claim: see Goff & Jones, ibid. at 50.
374
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
enrichments, benefits in kind have different values to different people. Some
people have
their poodles permed, other abhor permed poodles. If C
mistakenly pays D 100, D is unequivocally enriched. If C mistakenly perms
D’s poodle, D can object that it is impossible to affirm that his has been
enriched without infringing his right to value benefits in kind according to his
own priorities. There is a market, and a market price, but D must be free to
dissociate himself from the demand that sets that price.
62
the plea of subjective devaluation, namely by showing
This is now known as a plea of “subjective devaluation”, 3 and the tension
between the plaintiff’s right to disgorgement of the benefit and the defendant’s
autonomy is balanced by the responses to this plea. There are two ways of
overcoming
(i) an
incontrovertible benefit, or (ii) a request or free acceptance.’ Request or free
acceptance is not relevant in our context except in the easy case where the farmer took
active steps to bring the patented seed onto his land. 5 Incontrovertible benefit, on the
other hand, is directly relevant. When the benefit is incontrovertible the plea of
subjective devaluation is defeated and the plaintiff may recover the benefit, because as
it is said, the problem of subjective devaluation does not arise: “While the principle of
freedom of choice is ordinarily important, it loses its force if the benefit is an
incontrovertible benefit, because it only makes sense that the defendant would not
have realistically declined the enrichment.”‘ Money is the best example of an
incontrovertible benefit. 7
In unjust enrichment law the requirement of incontrovertible benefit raises
difficult problems in the context of services, since it is difficult to assess the value of
personal services. It also arises when the plaintiff paid money to a third party in
circumstances where it is questionable whether the defendant would otherwise have
made the payment,’ or where the defendant has received a realizable financial gain
that is not actually realized.’ In contrast, our case is easier than any of these. Crops
62 Birks & Mitchell, supra note 60 at 15.38.
63 The phrase was coined by Peter Birks, An Introduction to the Law of Restitution, rev. ed. (Oxford:
Clarendon Press, 1989) at 109. See the extensive discussion in Mclnnes, supra note 60 at 431ff.
See Mclnnes, ibid. at 433.
65 It might be suggested that the farmer might invoke free acceptance if it could be shown that she
had actually observed the patented seed entering onto her land. But in that case “acceptance” would
not necessarily be evidence of a subjective belief that the farmer will benefit, since it could equally be
evidence of a disinclination to take burdensome steps to eliminate a harmless nuisance.
66 The Honourable J.R. Maurice Gautreau, “When Are Enrichments Unjust?” (1989) 10 Advocates’
Q. 258 at 270-71, cited by McLachlin J. in Peel, supra note 58 at 795 [emphasis added by McLachlin
J.]. See Mclnnes, supra note 60 at 433-45.
67 “For example, choice is not a real issue if the benefit consists of money paid to the defendant or
paid to a third party to satisfy the debt of the defendant that was owing to the third party” (Gautreau,
ibid. at 270). See generally Mclnnes, ibid.
66 This was the main issue in Peel, supra note 58.
6 See Mclnnes, supra note 60 at 438-42.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
are a commodity, almost as fungible as money. Even if the farmer has not actually
realized the benefit at the time the action is commenced, she certainly will do so in the
near term. Thus our case falls into the category of an actually realized financial gain;
the benefit is essentially monetary, and the authorities are clear that recovery by the
plaintiff should be allowed in such circumstances even under the most stringent test.70
In summary, the autonomy problem can be characterized at three levels of
increasing severity: at the first level, the farmer may lose the ability to choose whether
to plant the patented crop; at the second level, the farmer may lose the ability to place
a value on the patented crop; and at the third level, the patented crop may be
systematically overvalued.
The first level of impairment is present but trivial under any measure. A farmer
never has complete control over the nature of his crop. Adventitious entry of various
plants is inevitable, whether it is volunteer barley from the neighbour’s property in a
wheat field, or simply weeds such as wild oats, which can be controlled but never
eradicated. No reasonable theory would demand autonomy at this level.” The second
level of impairment of autonomy is also present. While it is a more serious concern
than the first level of impairment, it is unlikely to have a significant economic impact
or affect autonomy in a strong sense as it will not change the farmer’s planning
decisions. The third level of impairment is most serious, as it would affect the
farmer’s ability to plan. Farming decisions would be different and less efficient. Given
the nature of the property in question, however, there is no particular reason to think
that the farmer’s autonomy will be impaired at this third level.
There are some caveats to these conclusions. First, even though accurately
assessed damages will not affect the farmer’s planning,
the transaction costs
associated with the
legal action or settlement negotiations are an undesirable
deadweight social loss. This will not be a problem if in practice the benefit conferred
on the farmer is so minimal that it is not worthwhile for the patentee to pursue, but it
might be a concern if adventitious entry resulted in large benefits. Secondly, even if
damages are accurately assessed on average, variation due to inaccuracy in the
70 Birks (supra note 60 at 116-17) states that the most stringent approach to rebutting the plea of
subjective devaluation is the “no reasonable man” test, which asks whether the benefit is such that “no
reasonable man would say that the defendant was not enriched.” He contrasts this with the less
stringent “objective” test, which merely asks whether there is an ordinary market value for the benefit
in question. Birks notes that “[i]t hardly needs to be added that the commonest example of a benefit
which is proved to be an enrichment by the ‘no reasonable man’ test is the receipt of money itself’
(ibid. at 117). Mclnnes (supra note 60 at 439) elaborates that in the case of a realized benefit, “having
turned the plaintiff’s services into money, it is as if the defendant received from her two thousand
dollars in cash. And, as always, money is immune to subjective devaluation.” Note that in the law of
unjust enrichment the question of subjective devaluation is normally raised in respect of services, as
the value of services is most likely to vary according to the taste of the recipient.
“, As Raz notes, “No one can control all aspects of his life” (supra note 45 at 373).
376
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
assessment of benefit from adventitious entry will increase uncertainty. To the extent
that farmers are risk averse this may lead to an undesirable change in farming
practices. Again, this effect will be minimal if the benefits from adventitious entry are
small in practice, as this source of uncertainty will be swamped by the multitude of
other sources of uncertainty in farming, most notably the weather. Finally, while there
is no particular reason to expect systematic judicial bias in assessing damages, it is
possible that such a bias might arise in practice for unforseen reasons. While the bias
might favour the farmer, it is also possible that it would favour the patentee, in which
case there would undesirable incentive effects. The larger the volume of damage
awards, the more likely it is that there will be some significant systematic bias in those
awards.
All of these secondary factors will be more important as the quantum of damages
claimed against innocent bystanders increases. We have seen that at present benefits
from adventitious entry are likely to be small, so that it is not worthwhile for the
patentee to sue. This reflects the current reality, where the only litigation, in Monsanto
v. Schmeiser, has been in respect of intentional propagation. The autonomy problem
would be more serious if adventitious entry gave rise to greater benefits. This might
potentially come out for two reasons: intentional release by the patentee, or new forms
of inadvertent gene spread.
Just as intentional propagation is the greatest threat to patent incentives, so also, it
might be suggested, intentional release by the patentee poses the greatest threat to
farmer autonomy. While the threat to autonomy from truly adventitious release is
small, in the absence of an innocent bystander defence, a patentee might intentionally
release patented seed and then sue farmers to collect the benefit conferred.”2 While this
would indeed be a serious impairment of farmer autonomy, it is not a realistic
concern. In the first place, a patentee who intentionally released its product in the
manner suggested would almost certainly be found to have granted an implied licence
to use the patent. It is clear law that a sale of a patented product is accompanied by an
implied licence to use the product.” It follows that giving a product away, as for
promotional purposes, is similarly accompanied by an implied licence to use the
product. Restrictions on the scope of the implied licence must be express at the time
72 Note that current environmental regulations related to plants with novel traits are also inadequate
in this respect, since these regulations are intended to prevent harm to the environment by spread of
undesirable traits, such as weediness. See e.g. Plant Biosafety Office, Plant Products Directorate,
Canadian Food Inspection Agency, “Regulatory Directive Dir94-08: Assessment Criteria for
Determining Environmental Safety of Plants with Novel Traits”, online: Canadian Food Inspection
Agency
” See Incandescent Gas Light v. Cantelo (1895), 12 R.PC. 262, 11 T.L.R. 381 (Q.B.); Badische
Anilin und Soda Fabrik v. Isler, [1906] 1 Ch. 605, 23 R.PC. 173 [Badische Amilin und Soda Fabrik
cited to Ch.].
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
377
of transfer; 4 the terms of the implied licence cannot be modified by the secret
intention of the patentee.
More importantly, a strategy of intentional release is unlikely to be profitable for
the patentee. If the invention is meritorious farmers will be willing to licence
voluntarily, and this is much more profitable for the patentee than lawsuits. If the
invention is not meritorious, “but for” damages will be minimal, and even residual
legal costs will swamp any potential profits.75 The position of the farmer is very
different from that of the patentee; while intentional unlicenced propagation by the
farmer will reduce the farmer’s costs by avoiding licence fees, intentional release by
the patentee’s costs by increasing transaction costs
the patentee will increase
associated with collecting royalties. This is why intentional propagation is a realistic
concern, while intentional release is not.
Similar reasoning indicates that new forms of inadvertent gene spread are
unlikely to seriously exacerbate the problem of adventitious entry. The concern is that
the characteristics of some future patented organism might be such as to increase the
likelihood of adventitious spread of the patented gene. Perhaps a future organism
would be susceptible to gene transfer to viruses, which would acquire the patented
gene and spread it by “infecting” other organisms. Because of the cost and difficulty
of recovering the benefits conferred by such uncontrolled spread, it would be in the
interest of the patentee to develop biotechnology to prevent this type of release.76 This
is not a conclusive argument. Biological organisms are complex and not entirely
predictable, and even though it is in the patentee’s interest to minimize adventitious
spread, technology may not always be able to control biology. This argument,
however, does indicate that there is reason to expect the patentee to make efforts to
prevent this type of undesired release, and given the difficulties of an innocent
bystander defence, discussed below, it would be wrong to shape policy based on this
perceived problem unless and until it actually materializes.
” The implied licence extends beyond mere use to, for example, resale, and a patentee might wish
to restrict the right of resale by express terms: see e.g. Betts v. Willmott (1871), L.R. 6 Ch. App. 239.
” Note that this point is valid only on a “but for” benefit-based approach to remedies. On the
approach adopted by the Court of Appeal in Schmeiser, in which any use of a patented plant-even
one conferring only a marginal benefit-would entitle the patentee to recover all of the fanner’s
profit, this problem would be a real concern.
76 One possibility would be a variation of the “terminator gene” technology: see text accompanying
note 94.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
IV. Innocent Bystander Defence
A. Introduction
We are now in a position to define the innocent bystander defence more precisely.
The best version of the innocent bystander defence will be as responsive as possible to
the problem of farmer autonomy while minimizing adverse effects on patent
incentives. In particular, we have seen that the primary concern from the patent
incentive problem stems from the possibility that the innocent bystander defence will
be used to shield intentional propagation. The key problem in designing the innocent
bystander defence is therefore to ensure that its bounds are sufficiently enforceable to
effectively control intentional propagation.
B. Availability of the Defence
The first question is who can take advantage of the defence? As we noted at the
outset, the proposed innocent bystander defence would not be a general intent-based
defence, available to all innocent users. It would apply only to a subset of innocent
users, in those cases where the patented invention had entered adventitiously onto the
user’s property. Precisely how is this limit to be defined?
The proposed innocent bystander defence is a limited exception to the general
rule that intent is not relevant to patent infringement. That general rule finds its most
important application in the case of independent inventors. An independent inventor
may well be unaware of a prior patent, so that her infringement by implementation of
her own invention may be entirely innocent.” Lack of intent cannot be a defence in
such circumstances or patents would be reduced to a form of copyright, as the key
substantive element distinguishing the two is that independent creation is a defence
under copyright law but not in patent law. The innocent bystander defence must
therefore be crafted to exclude independent inventors. Another category of potentially
innocent user is the ultimate customer for a product who purchases it from an
infringing producer. On existing law the end-user infringer will herself be an
” In competitive research and development intensive industries, competitors may monitor each
other’s patent portfolios closely, and so would be aware of relevant patents soon after they are laid
open. But in other cases, a workshop innovation that a company develops for its own purposes, such
as internal process improvement, may turn out to have been patented by another party. In any event,
patents are not laid open until 18 months after the application (supra note 5, ss. 10(2)-(3)), so that a
competitor may have priority despite the junior inventor’s best efforts to monitor the patent database.
And indeed, in patent systems like Canada’s, in which the patent is granted to the first applicant to
file, the independent inventor may actually have developed the invention first.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
379
infringer, notwithstanding that it never occurred to her that the product might not have
been manufactured under a valid licence.”
These categories of innocent user cannot be distinguished from the innocent
bystander on the basis of intent, as these users may be as innocent as any bystander.
The distinction between the innocent bystander and other innocent users must instead
be founded on the issue of adventitious entry, which we have already identified in
general terms as being central to the innocent bystander defence. The distinction is
that both the independent inventor and the unsuspecting end-user use the invention
through their own conscious act, whereas the innocent farmer uses the invention
without volition or knowledge. Put another way, an independent inventor knows what
she is doing, but does not know the patent exists. The innocent bystander, on the other
hand, normally knows the patent exists, but does not know what she is doing, in the
sense that she does not know the nature of the crop. This absence of a volitional act is
directly relevant to the principle issue of autonomy, and it also leads to a functional
difference in the steps that innocent users can take to protect themselves.79 An
independent inventor can in principle protect herself from infringing by searching the
patent register, and an end-user can also search the register, or more plausibly, can
obtain a warranty from the vendor that the product does not infringe.” The innocent
bystander could, at best, monitor her fields and would in any event be unable to
prevent the initial entry.
While the requirement of adventitious entry does place limits on the practical
scope of the defence, as discussed below, there is an important point of principle here
as well. Lack of intent is not in itself sufficient to relieve a user from liability in patent
law. Indeed, the independent inventor in particular has, in some ways, a stronger claim
to relief from liability than does an innocent bystander. Not only is the independent
inventor likely to be as innocent as the bystander, but the independent inventor did not
78 On its face, the Patent Act (supra note 5, s. 42) provides that an unauthorized user is liable for
infringement, regardless of the provenances of the infringing device. Reported cases are few,
presumably because it is rarely worthwhile to pursue an end-user, rather than a distributor or
manufacturer. Nonetheless, the principle is well established in the cases: “If a person innocently uses
a patented invention, not knowing that there is a patent, he is none the less an infringer, and if a person
innocently buys a patented invention from a licensee and uses it not knowing that there are limits on
the licence, I conceive that he is equally an infringer” (Badische Anilin und Soda Fabrik, supra note
73 at 611). See also Microbeads AG v. Wnhurst Road Markings Ltd. (1974), [1975] 1 All E.R. 529, 1
W.L.R. 218 (C.A.) [Microbeads] and Hanson v. Alpine Valley Ski Area Inc., 718 E2d 1075 (Fed. Cir.
1983) for cases in which the innocent end-user was found liable for patent infringement.
79 1 am indebted to Raman Balakrishnan, a student in my Advanced Intellectual Property course in
the spring of 2003, for emphasizing this argument.
‘ The warranty of quiet possession, and in most cases, the implied warranty of the right to sell,
which are implied by the various provincial sale of goods acts (see e.g. Sale of Goods Act, R.S.O.
1990, c. S-I, s.13(a)-(b)) would give protection in such a case, even in the absence of an express
warranty: see Microbeads, supra note 78.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
derive any benefit from the efforts of the patentee, while the innocent bystander’s
excess profits derive directly from the efforts of the patentee. The innocent bystander
defence, therefore, cannot be justified simply because the farmer, being innocent,
derives from his innocence a right superior to that of the patentee. If that were the
case, the independent inventor would have an equal claim to a defence. To the extent
that innocence justifies an exemption from liability, it is as a reflection of other policy
considerations. Even the ability to protect oneself is not the sole criterion for relieving
a user from liability. An independent inventor might spend large amounts on
development only to discover, when a rival’s patent is laid open, that her efforts have
been pre-empted, and she will be unable to recoup her costs. Ultimately, it is
evidentiary concerns, and not some broader matter of principle such as innocence or
ability to avoid the loss, that underpins the liability of an independent inventor and
distinguishes patent
and
enforceability, which will be raised below in respect of the innocent bystander
defence, are no weaker in principle than the similar considerations that lead to the
liability of an independent inventor.’
from copyright. The considerations of evidence
The more immediate point is that volitional use of the thing that is the subject of
the patent is the key functional difference between the innocent bystander and other
innocent users on which any justification for the innocent bystander defence must
rest. The defence should therefore be restricted to cases in which the patented
invention entered onto the defendant’s property without the assistance of the
defendant.
Consider how
this principle would operate
in some salient cases. The
paradigmatic case of adventitious entry is when a patented seed blows off a passing
truck into a farmer’s field. Another central case is when cross-pollination occurs
between the patented crop in a neighbouring field and the farmer’s unpatented variety.
Some seed will inevitably fall between the cracks and escape harvest and some of this
seed will carry the patented gene. In both of these core cases the defence would apply.
In contrast, suppose the farmer’s claim is that he bought common seed from
another farmer and, unbeknown to him, the .seed was contaminated with patented
seed. The innocent bystander defence should not be available in this type of case.
From the patent incentive perspective it raises very serious enforcement concerns,
since the claim that the farmer did not know the nature of the seeds is essentially
unverifiable. It is likely that the vendor will have informed the purchaser of the true
nature of the seed, in order to be able to charge a premium because of the presence of
s Note that the liability of the independent inventor is contrary to the principle, discussed above in
Part Ill(A), that the patentee should only be able to recover the fruits of its labour, and no more. For a
detailed discussion of the evidentiary concerns that override this principle, see Norman Siebrasse, “A
Property Rights Theory of the Limits of Copyright” (2001) 51 U.T.L.J. 1, especially Part I1(D) (“The
Patent Copyright Boundary”).
2004]
N. SiEBRASSE – THE INNOCENT BYSTANDER PROBLEM
the improved seed. Allowing the purchasing farmer to make this claim will force the
patentee to focus its enforcement efforts on the vendor. Nor is there a serious concern
from the autonomy perspective. We noted above that the innocent bystander is distinct
from the broader innocent user problem only because, in the case of adventitious
entry, that user has no control over his own use of the patented invention. In the case
of purchase of seed, in contrast, the farmer is voluntarily bringing the seed onto his
property. It is true that the farmer generally cannot tell from inspecting the seed
whether it is of a patented variety,, but the voluntary nature of the act of purchase puts
the farmer in the same position as a purchaser who buys patented goods from an
unlicenced manufacturer. Higher life forms do not raise any novel problems in this
respect, and if the innocent bystander exception were extended to purchasers of seed,
it would be necessary to extend a similar exception to innocent purchasers of goods
generally.
Another source of “volunteer” plants in ordinary farming practice is seed that is
inadvertently carried over on farm machinery. So, in a case where a farmer was using
a patented variety in one season and then discontinued its use in the next season, some
patented seed might remain on the machinery. Autonomy considerations do not
favour an exemption from liability in this case, as the farmer knew of the presence of
the use of the seed in the previous season and can protect himself from contamination
in the subsequent season by cleaning his machinery thoroughly. At the same time, an
exemption in this case would offer considerable scope for intentional withholding of
seed, which would be defended as having been retained on the farm equipment.
this manner would be on a modest scale, but the
Intentional retention
inconvenience of cleaning the equipment thoroughly is also modest, and given the
lack of autonomy concerns, the balance tips against allowing the defence in this case.
in
C. “Depth” of the Defence
The next question is the “depth” of the defence. What benefits is the farmer
entitled to retain when the seed enters adventitiously onto his property? At a
minimum, a farmer would have to be entitled to retain the excess profit from the sale
of the crop in the “ordinary” market, as for example, in the case of a high yield crop,
or the innocent bystander defence would disappear. Should the farmer be entitled to
retain seed for his own use the next year? Should he be entitled to sell the seed to
other farmers? Would the farmer be allowed to sell the crop in a specialized market,
available only to the patented crop if, for example, the crop produces vaccines? The
answers to these questions are crucially important to the patent incentive problem.
In suggesting the possibility of an innocent bystander defence, the Court of
Appeal in Schmeiser made it clear that the defence would not extend so far as to
intentionally replant patented seed, even if it entered
entitle
the farmer
to
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
adventitiously onto her property.” On its face this limitation accommodates the
problem of farmer autonomy. Negotiation prior to the initial adventitious entry is not
possible, but negotiation over the licencee fee prior to intentional replanting is
possible. The same argument applies equally if the farmer wishes to sell the saved
seed to other farmers.
There are nonetheless some difficulties with this position. Consider a farmer who
normally saves seed for replanting. If the innocent bystander defence does not permit
replanting, an innocent farmer whose crop had been contaminated with patented seed
would be required to sell the part of his crop that he would normally have retained for
seed, and then incur the additional expense of purchasing new seed. Similarly, a
farmer who normally sold common seed to other farmers would be prejudiced if his
crop were contaminated. These considerations suggest that in order to fully protect
farmer autonomy, the innocent bystander defence should extend to replanting and
even resale of patented seed if the initial entry was adventitious.”
However, extending this defence to this extent would create unacceptable patent
incentive problems. If replanting were permitted, selective replanting of only the
patented seed would also be permitted, since no practical distinction could be drawn.
This means that a farmer who had experienced adventitious entry could magnify his
yield of patented plants by selectively harvesting patented plants for seed. For
example, in the case of herbicide resistant seed this could be done simply by spraying
the affected area with herbicide to kill all the unpatented plants and then selectively
saving seed from the patented plants at the end of the season for sowing the next
season’s crop.’ Within a season or two, depending on
the degree of initial
contamination, the farmer could develop a crop consisting entirely of patented seed.
Allowing resale would then allow a farmer who had experienced adventitious entry to
concentrate the seed and set himself up as a competitor to the patentee in the sale of
the seed. This type of large scale sale by a competitor who is not required to pay
normal royalties would completely undermine the patent monopoly. The effect would
not be very different from denying patent protection entirely, except that the
competitor would have to wait for adventitious entry before going into competition.
Since some degree of adventitious entry is almost inevitable, this is not a significant
Supra note 1 at para. 57.
8′ The argument can be taken even further. Full protection of the autonomy of the farmer would
require that any farmer who purchased patented seed from the farmer for whom the seed had initially
entered adventitiously onto her property would also be permitted to save seed and sell the progeny,
since the latter farmer’s market would otherwise be reduced.
‘ This is what Schmeiser appears to have done: see Schmeiser (T.D.), supra note 22, especially at
paras. 38-40. Selective harvesting of other types of patented plants might require more effort, but it
would generally be possible if it was worthwhile. It is, of course, possible that in some cases,
depending on the crop or the patented features, technical considerations would prevent this type of
selective replanting. However, in such cases, the question of whether the defence extends to permit
replanting would be moot.
2004]
N. SlEBRASSE – THE INNOCENT BYSTANDER PROBLEM
383
distinction. Given our assumption that patent incentives are necessary for the
development of desirable new seeds (otherwise the optimal policy would be to deny
patents altogether) it is clear that this rule would be unacceptable.
Even if only replanting and not resale were allowed, there would still be a very
serious patent incentive problem. In the first place, farmers who did not normally save
seed might do so on discovering adventitious entry of a desirable patented variety.
Considerations of seed quality that lead most farmers to purchase seed rather than
save their own would provide some disincentive, but the appeal of avoiding the
licence fee would likely compensate for this in many cases. The problem would be
substantially exacerbated by intentional propagation. If replanting were permitted, a
farmer could gain a significant long term advantage by intentionally introducing an
amount of patented seed sufficiently small that the defence could plausibly be
claimed, and then concentrating it over a number of seasons. Indeed, even a farmer
who intentionally planted an entire field at the outset using black market seed might
claim that the seed was obtained by concentration over the course of a number of
seasons, and this might be difficult to disprove.
More importantly, extending the defence to permit replanting would create a
serious risk of a black market in the patented seed, even if resale were not allowed. A
farmer who experienced adventitious entry could concentrate the seed and sell it
illegally as common seed, while verbally informing the purchaser of its true nature.
While this resale would be illegal, it would be difficult to prevent as the patentee
would have to detect the actual sale transaction, as opposed to simply monitoring for
the presence of the patented seed in the growing crop. This is not to say that the
patentee’s rights would be completely illusory. Resale would certainly be reduced as
compared to the scenario in which it was permitted. But it would also certainly be
more prevalent than if replanting were prohibited. Further, requiring the patentee to
enforce by detecting sales rather than plantings would increase the patentee’s
enforcement costs, and these would be passed on to legitimate purchasers in the form
of higher prices.
Thus the line suggested by the Court of Appeal is sound, even though it impairs
farmer autonomy to a degree. While allowing replanting or resale would respect
farmer autonomy most fully, it would create unacceptable patent incentive problems.
It should be recognized that even the minimal innocent bystander defence, which
permits the farmer to retain profits from adventitious entry, creates significant
problems in preventing intentional propagation. The most obvious form of evidence
as to intention is the planting pattern. Adventitious entry will never result in an entire
field full of the patented crop, but if the intentional infringer is careful to plant in a
manner that is consistent with adventitious entry, for example, by mixing a relatively
small amount of the patented seed into the seed mix, it may be difficult to establish
intentional propagation by direct evidence. To prevent this kind of activity with a
purely intent-based defence, the patentee would have to show that the acquisition of
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
the seed was intentional. This presents significant enforcement problems, as the
farmer would presumably acquire the seed as ostensibly common seed with a verbal
assurance that it is patented seed. It is certainly not impossible to detect and restrict
this kind of transaction, but is much easier to detect growth of patented crops. The
patentee might also pursue the farmer who sold the seed, knowing it was mixed with
patented seed. But the seller would presumably grow the seed on relatively small
plots, since he would seek to make a profit from the sale of the seed for planting, and
not for grain, and these plots could be hidden far from the roads in ways that would
make gathering evidence difficult. Again, detecting this type of infringement is not
impossible, but it is significantly more difficult than simply detecting the growing of
patented crops.
Turning then to our final question, if the farmer is entitled to retain the benefit
from the sale of the crop in the ordinary market, should she also be entitled to retain
the benefit from the sale in a special market that is available only for the patented
crop? This is the case, discussed earlier, of a pig that is engineered to produce organs
that are immunologically compatible with humans and very valuable for that purpose,
or a plant engineered to produce human hormones. Permitting sale in the specialized
market is not defensible from the farmer autonomy perspective, since the farmer
would never have intended to sell an ordinary pig in this specialized market. But the
defence probably cannot be tailored to prohibit this type of sale, as there is no way of
distinguishing this case from those where the crop provides a benefit even when sold
in the ordinary market. The benefit is present in either case, and the farmer does not
need to take special steps to realize that benefit. The only distinguishing factor is that
in one case the product is sold in the same market as ordinary crops of that kind, and
in the other it is sold in a specialized market. It is not practical to draw a distinction on
this basis. Distinguishing between different kinds of markets would be very difficult,
and even if it could be done, it would only invite arbitrage. The farmer would sell in
the ordinary market, but let the buyer know the quality of the goods. The buyer would
then pay a premium, knowing that he could segregate that purchase and resell it in the
specialized market. The result would be the same, except that the benefit would be
split between the farmer and the first buyer. Attempting to control this arbitrage
process would almost certainly be prohibitively cumbersome.
D. Substantive Limits of the Defence
The next question concerns the substantive trigger for the defence. In proposing
an innocent bystander exception to patent liability, the CBAC Report recommended
that persons “without knowledge of the reproduction of a patented plant, seed, or
animal on their property” should not be liable for patent infringement.” This
statement is not sufficiently precise to define the rule, as knowledge falls along a
” Supra note 6 at 14. The term “innocent bystander” is nowhere expressly defined.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
385
continuum. The extremes are clear enough. A farmer who never discovered the true
nature of his crop would not be liable, while a farmer who knowingly took active
steps to bring the seed onto his property would be. But there is a range of possibilities
between these extremes. The farmer might have been aware of the nature of the crop
from the outset, even though he did not cause it to enter his property, if for example,
he happened to be in his field when the seed blew from a passing seed truck. Or he
might discover the nature of the crop only after it was sold; for example, if the crop
was wheat engineered to deliver a higher yield, the farmer might realize this only on
receiving a larger cheque than anticipated for the grain. In between these extremes, he
might discover
initial
contamination, but before sale, either on being informed by the patentee, or by his
own observation.’
the nature of the crop at almost any
time after the
We saw in the preceding section that making the innocent bystander defence
available as long as the patented seed entered adventitiously would create significant
patent enforcement problems, as the patentee would have to focus on showing
intentional acquisition rather than simply detecting the use of the patented crops. One
way to reduce that problem would be to provide that the farmer will be liable not only
for intentional propagation, but also if knowledge of propagation is gained before
some specified time. Liability would be triggered not only by intent, but also by
knowledge, which would normally be acquired by the patentee giving notice to the
farmer that the field contained the patented crop. The patentee would then be able to
focus its efforts not just at the point of sale, but also on detecting infringing crops in
the field.
There are two problems with this modification to the intent-based rule. The first
problem is that such a rule introduces significant arbitrariness, as a day’s difference in
the time of discovering the nature of the crop would make a difference of tens or
hundreds of thousands of dollars in liability. Secondly, the effectiveness of such a
change in improving enforceability depends on how late in the season knowledge can
be acquired and still trigger liability. If the cut-off point that triggers liability is the day
after planting, there is little advantage over a purely intent-based rule. At a minimum,
the cut-off point must be late enough that field samples can be tested. Enforceability is
improved as the cut-off time moves beyond this minimum towards harvest, but then
the rule itself becomes tantamount to the benefit-based rule. Under the intent-based
rule the patentee must discover the nature of the crop and inform the farmer in order
to trigger liability, while under the benefit-based rule, liability is not, in principle,
conditioned on the farmer’s knowledge. But as a practical matter, in order to enforce
its rights, even under the benefit-based rule, the patentee must discover the nature of
the crop, and once this is done, informing the farmer follows easily. Thus a rule under
6 As when Schmeiser apparently noticed the survival of some canola after he sprayed his ditches
with herbicide: see Schmeiser (C.A.), supra note 1 at para. 22. Some patented life forms might be
obviously different from their naturally occurring counterparts.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
which the farmer will be liable at any time up to the sale is almost equivalent to a
benefit-based rule, with the only difference coming in those cases in which the
patentee does discover the nature of the crop, but only after the sale. (If the rule is that
the farmer is liable if he ever discovers the true nature of the crop, even after sale, then
the knowledge-based rule reduces exactly to the benefit-based rule, since a farmer
who benefits will or should always know the nature of the crop after the sale brings in
more money than expected.) There is no obvious point between these extremes that is
more satisfactory. As the cut-off point moves from planting toward harvest the
enforcement problem decreases but the arbitrariness problem increases.87
In sum, if acquisition of knowledge at any time up to the sale will trigger liability,
an intent-based rule is almost indistinguishable from the benefit-based rule, except
that it introduces an extra element of arbitrariness. An intent-based rule in this form
clearly cannot be justified in comparison with the benefit-based rule. Thus in its most
attractive form the defence would relieve the farmer from liability so long as he did
not intentionally bring the patented seed onto his property. Knowledge acquired after
entry would not trigger liability.
One exception to this strict intent-based approach is desirable. Suppose the
invention requires special methods to take advantage of its properties and the farmer,
though she did not intentionally bring the invention onto her property, observes the
crop on her property at a time early enough to take the appropriate special steps to
benefit from the invention. The farmer should be liable for the benefit obtained in
such a case. The distinction should be easy to implement in practice, as it should be
entirely uncontroversial as to whether the crop in question requires special methods to
take advantage of its properties.
Such a limitation on the scope of the innocent infringer exception is also
substantively acceptable from the perspective of farmer autonomy. Recall that the
basic problem is that we want the farmer to be able to negotiate over the value of any
benefit. On this basis, crops that require special steps to provide a benefit are entirely
distinguishable from those that do not, since the farmer will necessarily have
knowledge of the nature of the crop before any special steps are taken and so will be
able to negotiate over the value of the benefit. If the farmer considers the licence fee
demanded by the patentee to be excessive, she can refrain from taking the special
steps and thereby avoid any substantial liability.
” The only plausible intermediate point at which to draw the line is where the farmer acquires
knowledge at a point at which it would be possible to remove the crop at minimal cost. But this would
be very early in the planting season, so the enforcement problem would remain severe, and this
approach has no advantages over a benefit-based approach, where an order to deliver up will not be
made after it is too late to replant. Further, the nebulous nature of such a rule would encourage
expensive litigation.
20041
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
387
E. Legal Scope: Tangible Property Claims
A final but important question concerns the legal scope of the defence. The Court
of Appeal and the CBAC Report proposals for an innocent bystander defence focused
on patent liability. In at least some cases, this will not be sufficient to protect the
innocent bystander. Patented life forms embody two forms of property at once: the
patentee’s intellectual property rights and the personal property rights of the owner of
the particular chattel. If Farmer A buys seed, and on his way home some of that seed
escapes through a tear in the bag and flies out of the back of the open pickup truck
onto the land of Farmer B, then it is perfectly clear that Farmer A is entitled to the
recovery of the seed or its value.88 It is no defence to the claim that Fanner B was not
at fault, nor is it a defence that Fanner A was careless. 9 In many cases the trifling
value involved will mean that this type of action will not be a practical concern.
However, there may be instances in which the value of the property itself will be
sufficient to make the action worthwhile. This may be because large numbers are
involved, perhaps if numerous escaped fish happen to end up in a nearby pen, or
because the life forms involved are individually valuable, such as the pig that is
immunologically compatible with humans.
The potential for a claim based on tangible property rights raises two questions.
Most obviously, should the innocent bystander defence be extended to provide
protection against these proprietary claims? Secondly, if not, why not? On what
principle can the patentee’s intellectual property right be distinguished from the
tangible property right of the owner of the escaped property?
This can be pursued by an action for conversion: see generally Fleming, supra note 25, c. 4 at
60ff. (“Intentional Interference with Chattels: Conversion”). This is apart from any issue as to the
triviality of the claim.
On intent of the defendant, Lord Diplock states:
At common law, one’s duty to one’s neighbour who is the owner, or entitled to
possession, of any goods is to refrain from doing any voluntary act in relation to his
goods which is a usurpation of his proprietary or possessory rights in them. Subject to
some exceptions which are irrelevant for the purposes of the present case, it matters not
that the doer of the act of usurpation did not know, and could not by the exercise of any
reasonable care have known, of his neighbour’s interest in the goods. The duty is
absolute; he acts at his peril (Marfani & Co. Ltd. v. Midland Bank Ltd., [1968] 2 All
E.R. 573 at 578, 1 W.L.R. 956 (C.A.) [Marfani cited to All E.R.]).
This statement was approved and applied in Simpson v. Gowers (1981), 32 O.R. (2d) 385, 121 D.L.R.
(3d) 709 (C.A.) and cited with approval in 384238 Ontario Ltd. v. Canada (1983), 8 D.L.R. (4th) 676
at 684 (EC.A.). On negligence of the plaintiff, see also Marfani, ibid. at 577-78, cited with approval
in Boma Manufacturing Ltd. v. Canadian Imperial Bank of Commerce, [1996] 3 S.C.R. 727 at para.
31, 140 D.L.R. (4th) 463: ” … the moral concept of fault in the sense of either knowledge by the doer
of an act that is likely to cause injury, loss or damage to another, or lack of reasonable care to avoid
causing injury, loss or damage to another, plays no part.”
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
From the perspective of the innocent bystander, there is no reason to distinguish
the claim of the owner of the specific property from the claim of the patentee. The
patentee’s claim is the same in principle as that of Farmer A: Fanner B has received
valuable property from the claimant, and the claimant seeks return of the property, or
if the property itself cannot be returned, return of the value. Indeed, the patentee’s
claim will be monetarily diminished, since the patentee can claim only for the value
of the patented seed that is over and above that of the same unpatented variety, while
Farmer A will claim for the entire value of the seed. Both claims interfere equally
with the farmer’s autonomy. In either case Farmer B will be asked to pay for a benefit
that she never requested. Thus in order to protect the farmer’s autonomy, the innocent
bystander defence must extend to every type of claim that might be made against the
bystander, whether it is a patent claim or a property claim, or an unjust enrichment
claim.
Are there any relevant differences between the patentee’s claim and the claim of
the owner of the tangible property (the seeds themselves)? It is not relevant that the
patentee seeks the value of the property rather than the thing itself, since the normal
order in an action for conversion is damages.’
The CBAC Report suggested that the reason for denying recovery to the patentee
was that Farmer B might be harmed by the presence of the patented seed.’ But this
simply reinforces the analogy with chattels; if the seeds themselves somehow harmed
Farmer B, perhaps by contaminating a wheat crop with canola, Farmer B would have
a potential action in nuisance against Farmer A just as a farmer might have a potential
action against a patentee. In ordinary property law it is quite clear that the fact that the
escape of Farmer A’s property has harmed Farmer B would not prevent Farmer A
from recovering that property, but would simply expose Farmer A to liability for the
harm caused. The obvious reason for this is that there is no necessary or, even likely
proportionality, between the harm caused and the value of the property, so that
allowing Farmer B to retain the property as “compensation” for the harm caused is
entirely arbitrary.
One possible difference is that in a contest between Farmer B and the patentee,
Farmer B has a property interest in the seeds themselves and the plants that grow from
them that is superior to that of the patentee, whereas in the contest between the two
farmers, Fanner A’s claim is clearly superior to that of Farmer B in all respects. But
the fact that the defendant has mixed his property with that of the plaintiff is not a
defence to a proprietary claim. It may affect the remedy, so that the owner may be
0 At common law the defendant in an action for detinue had the option of whether to return the
property or to give its value to the plaintiff: see Fleming, supra note 25, c. 4 at 81 (“Conversion:
Specific Restitution”). Specific restitution may now be ordered, but this is in the discretion of the
courts (ibid.).
“‘ Supra note 6 at 14.
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
389
restricted to recovery of the value of the property rather than recovery of the property
itself,2 but recovery of the value is all the patentee is asking for in any event.
Similarly, the claim that the patentee should not recover because it allowed the seed to
escape is no stronger than the claim that the neighbouring farmer allowed the physical
seed to escape, and this, as we have noted, is irrelevant to the claim for the return of
chattels.
We can only conclude that the claim for the return of the value of the tangible
property is indistinguishable in principle from the claim for the return of the value of
the intellectual property. Certainly this is true from the perspective of the farmer’s
autonomy. This implies that the innocent bystander defence must be extended to
apply equally to claims for the return of tangible personal property that embodies the
intellectual property.
F Summary: Fashioning the Innocent Bystander Defence
Though the above analysis incorporates a number of potentially contentious
matters of judgment, I have argued that the best form of the innocent bystander
defence would have the following features. It would be restricted to cases in which the
patented life form entered onto the user’s property without any volitional act on the
part of the user.” So long as that condition is satisfied, subsequent discovery of the
true nature of the crop would not affect the availability of the defence, except that the
user would not be entitled to take advantage of subsequent knowledge to gain a
benefit that she would not have gained had she remained ignorant. The user would be
entitled to retain only those benefits that flowed directly from the adventitious entry,
with no right to replant or sell the seed to others for planting. As well as relief from
liability for patent claims, the defence would extend to protect the bystander from
claims made by the owner of the physical chattel that embodied the patented
invention.
Conclusion: Desirability of an Innocent Bystander Defence
We are at last in a position to answer the central question: would the
implementation of an innocent bystander defence, in its best form, be a desirable reform
of patent law? While there is some merit to the proposal, overall, the answer is no.
92Fleming, supra note 25.
9 Note that while the innocent bystander defence is motivated by concerns relating to patented
higher life forms, the exception itself would not need to be expressly restricted to higher life forms.
Restricting, however, the applicability of the defence to cases of escape limits the problem largely to
cases of patented higher life forms, since it is this type of invention that is most likely to escape from
the user.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
It is certainly true that a farmer should not be liable to the patentee simply for the
adventitious entry of patented crops into her field. But it is essential to recognize that
adventitious entry will not generally give rise to liability under the general liability
rule. The general rule imposes liability only to the extent of excess profit caused by
the use of the patented crop, and adventitious entry of patented crops will rarely
benefit the farmer. This is because many types of patented crops require special
treatment to realize their advantages, and an innocent bystander will not know to use
these special methods. Even when special treatment is not required, patented crops
scattered in a field of a different variety will often not provide a benefit. This is
because farming methods must be tailored to the majority (non-patented) variety, and
differences in any of a range of characteristics, such as time to maturity, can negate
the benefits of the patented variety.
With this in mind, the unfairness in requiring the innocent farmer to disgorge the
excess profits he obtained from the fruits of the patentee’s labour is minimal, as the
farmer is made no worse off that he would otherwise have been. An innocent recipient
of ordinary tangible property in similar circumstances is required to return it, or its
value, to the owner, and there is no obvious reason why the innocent recipient of
intellectual property should be treated differently. Indeed, a complete innocent
bystander defence would have to extend to abrogate this rule of property law, in
addition to creating a defence to a patent claim. Nor is it clear that the innocent
bystander is in any stronger position to resist the patentee’s claim than is an
independent inventor; the innocent bystander profited from the patentee, while the
independent inventor did not, yet sound evidentiary reasons necessitate that the
independent inventor be considered an infringer.
With that said, farmer autonomy is impaired to some extent when a patentee
recovers the value of the benefit conferred by adventitious entry of patented crop, but
the impairment is not serious. A serious threat to autonomy would arise if judicial
assessments of damages were systematically biased against the farmer, but there is no
reason to expect this type of bias. The reasons that justify a strong property right for
personal goods such as a wedding ring do not apply in the case of fungible goods like
crops. There will be some impairment of autonomy as the farmer loses the right to
place a value on the benefit, but this type of impairment is relatively minor as it does
not affect the ability to plan. Neglecting this type of loss of autonomy is consistent
with general principles of law, including the law of unjust enrichment and the rule that
difficulty of assessing damages is no reason that they should be denied.
Thus, the substantive case in favour of the innocent bystander defence is relatively
weak. And on the other side of the coin, the innocent bystander defence, even in its
best form, would have a number of shortcomings. An innocent bystander defence
could not completely cure the autonomy problem. In order to avoid unacceptable
patent incentive problems the defence would have to be restricted to permitting the
farmer to keep only the benefits of the direct fruits of the adventitious entry. This is
both too narrow and too broad from the autonomy perspective. It is too narrow
2004]
N. SIEBRASSE – THE INNOCENT BYSTANDER PROBLEM
because the farmer would not be permitted to save or sell seed, as he might have done
in the absence of contamination. And it is too broad because the farmer would be
permitted to retain a benefit gained by selling the life form in a market in which it is
unusually valuable, even though this is not defensible from an autonomy perspective.
More importantly, it is not possible to craft an innocent bystander defence that
does not to some extent facilitate intentional infringement. Allowing a farmer to retain
the profit so long as the initial entry was adventitious means that prevention of trade in
unlicenced seed would have to focus on detecting sale transactions, rather than
monitoring fields for unlicenced crops. The discrete nature of these sales would make
enforcement of the patentee’s legitimate rights significantly more difficult. Without an
innocent bystander defence the patentee can still pursue the initial illicit sale of
patented seed, and it can also recover the benefit after sale from the purchasing
farmer, thus reducing the farmer’s incentive to buy illicit seed in the first place. This is
not to say that enforcing legitimate rights would be impossible if an innocent
bystander defence were implemented, but it would be more difficult and costly.
Finally, we should recall that legal and technical defences are complementary. If
an innocent bystander defence makes enforcement of the patentee’s rights more
difficult, the patentee will respond with technical measures to protect its property.
Most obviously, this might take the form of inserting a so-called “terminator gene”
into the patented seed. This technology allows plants to be engineered so that specific
genes are turned on only in the presence of a chemical inducer. If the gene in question
is essential to plant growth, the plant will grow only when sprayed with the required
inducer. In effect, this technology could be used to turn all patented crops into
“special methods” crops, which would require a special trigger to grow so that only
intentional users would ever derive a benefit.’ While this would eliminate the problem
of adventitious entry, it would also increase costs to farmers generally, who would be
required to buy the chemical inducer. This increase in farming costs would at least
partially offset the benefit obtained from the use of the patented crop. Allowing
control of illicit use by legal as well as technological means can reduce costs and
increase food supply.
The appeal of an
to patent infringement
innocent bystander defence
is
understandable, but the need is not pressing. The general liability rule addresses the
most pressing concerns regarding the escape of higher life forms, and the residual
impairment of farmer autonomy is minimal. An innocent bystander defence would
lead to increased legal and technological costs in order to control the consequent
enforcement problems, and these costs would be passed on to farmers and then to
consumers. The minimal benefits of the innocent bystander defence do not warrant
these costs. Maybe some day in the future, technological change will result in a more
” See U.S. Patent No. 5,723,765 “Control of Plant Gene Expression”.
392
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
pervasive problem of uncontrolled escape, in which case an innocent bystander
defence may be desirable. That day is not now, and it may never come.
