(2016) 2(1) CJCCL

301

Panacea or Pandemic: Comparing
“Equitable Waiver of Tort” to
“Aggregate Liability” in Cases of
Mass Torts with Indeterminate
Causation
Craig Jones*
The equitable doctrine of “waiver of tort”, in which a plaintiff surrenders the right to
tort damages and seeks instead a disgorgement of the defendant’s wrongful profits, has
received a mixed reception in Canadian courts. In this article, the author explains the
doctrine and its difficult history, and proposes that the problem against which waiver
of tort is usually being applied — indeterminate causation in mass tort claims — is
very real. However, the author concludes that the use of the doctrine of waiver is a
partial solution at best, and advocates instead for a more fundamental rethinking of our
approach to causation in class actions.

*

Professor of Law, Thompson Rivers University. The author is grateful for
the research assistance of Kendra Morris, J.D. Candidate, Thompson
Rivers University.

302

Jones, Panacea or Pandemic

I.

Introduction

II.

Waiver of Tort in Canadian Law
A.
B.

III.

IV.

What is Wrong With a Waiver of Tort Claim?
A.

What is the Problem, Exactly?

B.

Wrongs “In the Air”

C.
Under-Compensation and Under-Deterrence
D.
The Divergence of Interests of Class Counsel
A Better Way
A.
B.

V.

I.

I

The Difficult Doctrine
The Jurisprudence

The Problem, Reiterated
Assessment of Liability, Harm, and Damages on an Aggregate Basis

Conclusion

Introduction

s the equitable doctrine of waiver of tort, as they say nowadays, “a
thing”? If so, should it be the “next big thing” in mass tort class actions?
My answers to these questions are a qualified “yes” and an emphatic
“no”. Waiver of tort should be recognized as a cause of action, I will
argue, and may at times be useful, but it should not be the doctrine of
first, or even second, resort in mass tort class actions, as it is generally
inferior to the available alternative: the evolution of tort law to permit the
aggregate determination of causation in large-scale claims.
What exactly is waiver of tort? After 20 tumultuous years during
which it has been pleaded and occasionally argued before the courts in
Canada, we still do not know much about the rather obscure doctrine. In
its most robust formulation, waiver of tort allows a plaintiff who is able
to prove all the constituent elements of a traditional wrong except that
they have suffered a loss as a consequence of the defendant’s breach to

(2016) 2(1) CJCCL

303

“waive” tort compensation and claim only “disgorgement”1 of the profits
that the defendant earned as a result of the wrongful behaviour. As the
Supreme Court of Canada noted in its decision in Pro-Sys Consultants
Ltd v Microsoft Corporation2 (“Microsoft”): “[a]n action in waiver of tort
is considered by some to offer the plaintiff an advantage in that it may
relieve them of the need to prove loss in tort, or in fact at all”.3
The doctrine is based on the intuitively appealing notion, deeply
rooted in equity, that a defendant ought not be able to profit from
wrongdoing. It also bears on deterrence: equitable waiver allows a plaintiff
who has not suffered from harm to perform a corrective role in depriving
a wrongdoer of profits, disincentivizing antisocial behaviour. But where a
class of persons have suffered harm as a result of a mass wrongdoing, one
might ask, why would they give up what may well be the overwhelming
bulk of their claim?
One answer is straightforward: a waiver of tort claim “may be the
easiest cause of action to prove”,4 because disgorgement flows from
the wrongdoing of the defendant, rather than the harm caused to the
plaintiff. Under its theory, it is enough for a plaintiff to show that the
defendant behaved in a way that was wrong — usually that it breached a
duty somewhat “at large” or generic — but the plaintiff need not establish
a wrong — that is, there is no requirement that all the elements of a
complete tort be present. So, if a manufacturer produced a dangerously
defective product, or failed to provide a necessary warning to its customers,
or if an issuer of shares deliberately or negligently misrepresented facts in
a prospectus, or if a factory breached pollution standards and exposed its
neighbours to risk, the defendant could lose whether or not the plaintiff
1.

2.
3.
4.

The cases and literature on waiver of tort often use the language of
a number of equitable remedies — “accounting”, “disgorgement” or
“constructive trust”. But no matter how it is cast, the effect is the same:
some amount equivalent to the defendant’s ill-gotten profits will be
calculated and surrendered to the plaintiff or class.
2013 SCC 57 [Microsoft SCC].
Ibid at para 93.
Paul Perell, “Field Notes on Products Liability Claims in Class Actions”
(2011) 38:2 Advocates’ Quarterly 149 at 163 (describing possible
advantages in the context of products liability claims).

304

Jones, Panacea or Pandemic

could prove a direct connection between the wrongdoing and his or
her loss. This is the attraction of the resort to equity: it is often a very
difficult task to prove the necessary causal link, particularly where the
loss is indirect, or issues of reliance or scientific uncertainty are in play in
an individual case.
When it first appeared on the class action scene in the 2004 decision
of Serhan Estate v Johnson & Johnson5 (“Serhan”), waiver of tort was
welcomed as a panacea for class claimants, allowing them to vault over
many of the traditional rules of tort that barred recovery even where
defendants had miserably failed in their duties of care or had been shown
to have been fraudulent or malevolent. Waiver of tort claims, generally
pleaded in the alternative, became a routine feature of class action
pleadings,6 and as they began to trickle before the courts, the class action
bar held its collective breath.
It was a long wait. In the intervening two decades, judicial
skepticism, or at least ambivalence, had calcified into what appears to be
a trend of qualified rejection, and it appeared to some observers that the
doctrine was a dead letter.7 However, in the recent certification decision
of Microsoft, a unanimous Supreme Court of Canada refused to strike a
waiver of tort claim and permitted it to proceed to trial, holding that it
was not “plain and obvious” that it could not succeed.
Judicial reluctance, however, remains in the wake of Microsoft, as
exemplified by the subsequent decision of O’Brien v Bard Canada Inc8
(“O’Brien”), where Justice Perell, while allowing that disgorgement
through waiver of tort may be a suitable remedy in some cases, found
it wildly inappropriate for a mass tort premised on personal injury with
5.
6.
7.

8.

(2004), 72 OR (3d) 296 (Sup Ct J) [Serhan Sup Ct], aff’d (2006), 85 OR
(3d) 665 (Div Ct) [Serhan Div Ct].
See Perell, supra note 4 at 160-63.
Charles Murray, “An Old Snail in a New Bottle? Waiver of Tort as an
Independent Cause of Action” (2010) 6:1 Canadian Class Action Review
5; H Michael Rosenberg, “Waiving Goodbye: The Rise and Imminent Fall
of Waiver of Tort in Class Proceedings” (2010) 6:1 Canadian Class Action
Review 37; and Kit Scotchmer, “Waiver of Tort: A Potential Sea Change
in Class Action Law” (2011) 7:1 Canadian Class Action Review 159.
2015 ONSC 2470 [O’Brien].

(2016) 2(1) CJCCL

305

substantial potential damages.
In this article I will argue that the reluctance of Perell J to permit
plaintiffs to rely on waiver of tort was well placed, even if the doctrine
might have some residual utility. However, the problem that it sought
to address still looms unbearably large in the legal landscape: what to
do when a mass wrong has clearly been committed, where people have
been harmed, but where it is difficult or impossible to identify the actual
victims, or the degree of their individual loss, with any precision, if at all?
I suggest here that large-scale claims premised on widespread
defendant wrongdoing can be usefully broken into two categories, only
one of which is truly amenable to the application of waiver of tort. I
conclude that, in cases where a defendant has committed a wrong of a
type that is so serious that deterrence is called for even when it does not
produce legally-cognizable harm (and particularly when the wrong is, by
its nature, elusive of ordinary damages claims), waiver of tort can provide
an important behaviour modification device where no other is available.
This category might include criminal acts, some regulatory offences, and
intentional common law or equitable wrongs such as fraud, deliberate
misrepresentation or a wanton or otherwise egregious flouting of legal
duties.
However, with respect to most true mass torts with individual
causation issues, where it is known that the defendant was negligent and
that this had caused harm but it is not knowable with certainty which
victim’s injuries could be attributed to the wrong as opposed to other
causes, waiver of tort is not the preferable analysis when compared to
other devices that are (or should be) available as a matter of tort law. In
particular, I propose, the most appropriate solution in most cases is an
aggregate or global treatment of causation issues: treating causation of
harm as something that occurs in a population of persons, rather than a
collection of individuals.
My argument here proceeds as follows. In Part II, I describe the
doctrine and its origins, and trace its recent somewhat lurching progress
through 20 years of Canadian class action jurisprudence. In Part III, I
identify the built-in limitations of the doctrine and other problems that
would be inherent in its widespread adoption. In Part IV, I describe the

306

Jones, Panacea or Pandemic

principal alternative: the assessment of causation in mass tort cases on an
aggregate, rather than individual basis, and suggest why, in such cases, it
is preferable to waiver of tort. In my conclusion I outline and defend the
classification of large-scale claims into the two categories I have earlier
mentioned, only one of which should be dealt with through recourse to
waiver of tort.
I argue here that resort to equitable waiver is not the preferable
solution to the problem of indeterminate causation or a lack of nexus
otherwise between wrongdoer and victim, for three reasons.
First, divorcing damage from wrongdoing altogether risks distorting
the historical role of tort by removing an important (and occasionally
maligned, including by me) limiting principle: restricting recovery
through a nexus of harm-causation prevents the emergence of a purelyregulatory civil litigation regime where busybody plaintiffs and their
lawyers are incentivised to ferret out even harmless wrongdoing, leading
to a costly, inefficient and wasteful court “policing” of the economy that
does little to advance the role of the civil courts as fora for the vindication
of aggrieved victims. Having identified this problem, though, I suggest
that it is not insurmountable and can be addressed through a principled
application of the rules of standing.
A second and more serious concern regarding waiver of tort in class
actions is not based on a fear that they could accomplish too much,
but rather that they do too little, too easily. I will argue that where
the wrongdoing has actually caused harm, disgorgement of profits is
inferior to tort damages as a device of either compensation or deterrence,
especially when measured against other possible innovations available to
the courts.
This leads to my third and final argument for disfavouring the
waiver approach in most cases: although it may be argued that “some
compensation and deterrence is better than nothing”, reliance on waiver
of tort risks creating a schism between the interests of the class and the
public on one hand (who would want the defendants to pay the full cost
of harm, not just of profits) and class counsel on the other. This is because
a waiver of tort claim would provide “low hanging fruit” for lawyers,
who would be incentivised to pursue a high volume of “quick and dirty”

(2016) 2(1) CJCCL

307

settlements or judgments rather than seeking to maximise recovery for
their clients.
I argue here that the problem of under-deterrence and undercompensation in mass claims where the defendant’s wrongdoing is plainly
established is very real, and it is something that the substantive law of
tort should, and can, accommodate through rules of causation specific to
mass torts, and in particular class actions. This requires that courts treat
mass torts as fundamentally distinct from individual claims, as harm in
populations, rather than individuals. This is an idea that underlay (or at
least should have underlain) the attempts to utilize waiver of tort, but to
put it simply, tort law can do it better. In this respect, I will suggest that
very recent decisions of the Quebec Superior Court of Justice and the
Ontario Superior Court provide us an intriguing glimpse of the future.

II.

Waiver of Tort in Canadian Law

A.

The Difficult Doctrine

In introducing the concept of waiver of tort it is difficult to do better
than quote the description by Justice Epstein of the Ontario Divisional
Court in Serhan:
[i]ts origin lies in the expression “waiver of tort and suit in assumpsit”, the
latter being the historical antecedent of many modern common law “quasicontract” restitutionary claims. In invoking waiver of tort, the plaintiff gives
up the right to sue in tort and elects to base the claim in restitution, thereby
seeking to recoup the benefits the defendant has derived from his wrongful
conduct. The practical purpose behind it is that in certain situations, where
a wrong has been committed, it may be to the plaintiff’s advantage to seek
recovery of an unjust enrichment accruing to the defendant rather than normal
tort damages.9

Justice Epstein went on to quote Peter D Maddaugh and John D
McCamus in The Law of Restitution:
[t]he doctrine known as “waiver of tort” is perhaps one of the lesser appreciated
areas within the scope of the law of restitution. From the outset, it seems to
have engendered an undue amount of confusion and needless complexity.
The almost mystical quality that surrounds the doctrine is attested to by the
following famous couplet penned by a pleader of old [[J.L. Adolphus, “the
9.

Serhan Div Ct, supra note 5 at para 50.

308

Jones, Panacea or Pandemic

Circuiteers– An Eclogue” (1885) 1 L.Q Rev. 232, at p. 233]:]: “[t]houghts
much too deep for tears subdue the Court when I assumpsit bring, and godlike waive a tort”. One source of this confusion stems from the doctrine’s very
name. As one writer has pointed out, not entirely facetiously, it has “nothing
whatever to do with waiver and really very little to with tort”.10

Historically, waiver of tort was restricted to a narrow and discrete class
of “predicate wrongs”, cases involving conversion, detinue, trespass to
chattels and deceit.11 However, as Serhan (a products liability case) and
its progeny suggest, and the Supreme Court of Canada’s decision in
Microsoft (a competition class action) confirms, its possible application
may be almost limitlessly broad.
Waiver of tort is not compensatory, and in this sense it is distinct
from an equitable claim of unjust enrichment, which requires not only
an unlawful profit by the defendant but also a corresponding loss by
the plaintiff,12 a transactional relationship which has historically been
required to be quite direct.13 The waiver doctrine emerges from the basic
equitable idea that holds (as more or less a standalone principle) that a
wrongdoer should not be permitted to retain ill-gotten gains.
Much of the skirmishing around waiver of tort in class claims has so
far centred on whether the doctrine’s availability in a particular claim or
generally should be decided at the preliminary certification stage, where
10.
11.

12.

13.

Ibid at para 51, citing Peter D Maddaugh & John D McCamus, The Law
of Restitution (Aurora: Canada Law Book, 2005)(loose-leaf ) at 24-1.
Paul Perell, “Serhan Estate v. Johnson & Johnson and Soulos v.
Korkontzilas: Something Is Happening Here and We Don’t Know What
It Is”, Case Comment, (2007) 33:1 Advocates’ Quarterly 375 at 375-76.
The well-known elements required to establish an unjust enrichment are:
(i) an enrichment of the defendant; (ii) a corresponding deprivation of the
plaintiff; and (iii) an absence of juristic reason (such as a contract) for the
enrichment: see Alberta v Elder Advocates of Alberta Society, 2011 SCC 24
at para 82 [Alberta Elders]; Garland v Consumers’ Gas Co, 2004 SCC 25 at
para 30; Rathwell v Rathwell [1978] 2 SCR 436 at 455; Pettkus v Becker
[1980] 2 SCR 834 at 848.
But see Microsoft SCC, supra note 2 at para 87 (where the Supreme
Court of Canada permitted an unjust enrichment pleading to survive
the certification stage, notwithstanding that it was brought by indirect
purchasers).

(2016) 2(1) CJCCL

309

the test is the low threshold of “plain and obvious” and where novelty,
the Supreme Court of Canada has held, cannot be a necessary bar to
a legitimate claim lest it become a barrier to the progress of tort law
through the discovery of new or evolved causes of action.14 The Supreme
Court of Canada seemed to settle this aspect of the controversy when it
upheld the certification of the waiver claim advanced in Microsoft.15 But
the Microsoft decision, and those other “demurrer” cases that came before
and since, put off the central questions to another day: if it exists as an
independent cause of action, what would it look like? Clearly it must
be premised on wrongdoing by the defendant. But could any kind of
wrongdoing found recovery? And recovery by whom?

B.

The Jurisprudence

It was the 2004 decision of Justice Cullity in Serhan, upheld by a majority
of the Divisional Court two years later, that firmly fixed waiver of tort
in Canadian legal consciousness. Serhan was a products liability class
action brought by users of a device used to monitor blood glucose levels.
14.

15.

Alberta Elders, supra note 12 at para 20 (noting that the first criterion for
certification of a class action is that the plaintiff’s pleading must disclose
a cause of action); and Hollick v Toronto (City), 2001 SCC 68 at para 25
(where the Supreme Court of Canada confirmed that this requirement
is assessed on the same basis as a motion to dismiss, as set out in Hunt v
Carey Canada Inc [1990] 2 SCR 959 at 980: the question is, assuming all
facts in the Statement of Claim are true, whether it is “plain and obvious”
that the plaintiff’s claim cannot succeed).
However, see O’Brien, supra note 8 (where Perell J suggested that
Cromwell J’s dicta should not apply in a products liability case, and he
rejected waiver of tort as a legitimate cause of action in the facts before
him: “Pro-Sys Consultants Ltd. was a competition law action. The case
at bar is a products liability tort case. For decades, going at least as far
back as Donoghue v. Stevenson, [1932] AC 562 (HL), and continuing to
this day, courts have determined matters of policy in tort claims at the
pleadings stage and if it were necessary to do so I would decide whether
waiver of tort is a cause of action and, if it is a cause of action I would
decide whether it is a viable cause of action for a products liability
proposed class action so as to satisfy the cause of action criterion of
certification” at para 158).

310

Jones, Panacea or Pandemic

The device was admittedly defective, and, possibly as a consequence,
harmful. But it appeared likely that none of its thousands of users in
Canada could show any harm that could be attributed to the defect, and
they had all received it for free. Class members seemed to have no real
claim, in other words, under traditional tort or contract law, or under
any statutory regime. Nevertheless, Cullity J found that the plaintiffs had
pleaded material facts sufficient to support a claim under waiver of tort,
notwithstanding that it had not been specifically invoked. On appeal,
Epstein J of the Divisional Court allowed certification to proceed on
the equitable claim, and the case settled in 2011 with the application of
waiver of tort still deeply in doubt.16
Following Serhan, a number of Ontario decisions permitted waiver
of tort claims to proceed through the certification, based primarily on
the idea that the operation of the doctrine was not so settled so as to
make such claims certain to fail on the “plain and obvious” standard.17 A
number of other decisions disposed of waiver of tort claims on the basis

16.

17.

Serhan v Johnson & Johnson, 2011 ONSC 128, (The settlement provides
replacement devices for diabetics, and educational programs and other
methods of cy-près distribution, totalling $4.5 million. In approving
the settlement, the Court recognized that the case likely turned on the
waiver of tort issue. In discussing the “likelihood of success” criterion for
determining the reasonableness of the settlement, Horkins J wrote:
“[m]ost importantly, does waiver of tort exist as an independent cause
of action or is it only a remedy applicable to another tort? This difficult
question is at the heart of this case. While Ontario Class Counsel were
confident that a court would find that it was an independent cause of
action, there was a considerable risk that it would not” at para 69).
Griffin v Dell Canada Inc, 2009 CanLII 3557 (Ont Sup Ct J); Haddad v
Kaitlin Group Ltd, 2008 CanLII 66627 (Ont Sup Ct J); Heward v Eli Lilly
& Co (2007), 39 CPC (6th) 153 (Ont Sup Ct J), aff’d (2008), 295 DLR
(4th) 175 (Ont Div Ct); Pollack v Advanced Medical Optics Inc, 2011
ONSC 1966; Peter v Medtronic Inc (2009), 83 CPC (6th) 379 (Ont Sup
Ct J), aff’d 2010 ONSC 3777; Tiboni v Merck Frosst Canada Ltd (2008),
295 DLR (4th) 32 (Ont Sup Ct J), leave to appeal refused (2008), 304
DLR (4th ) 220 (Ont Sup Ct J).

(2016) 2(1) CJCCL

311

that the operation of the doctrine, if applicable, was moot.18 In each of
these latter cases the claim was defeated either at a preliminary stage or
after trial because no predicate wrong to support waiver of tort had been
established (that is to say, not only that there was no completed tort, but
that there was also no breach of duty or other illegality on which to found
the waiver).
There was a somewhat different, but no more conclusive, outcome
in Koubi v Mazda Canada Inc19 (“Koubi”), where the British Columbia
Court of Appeal decertified a waiver of tort class action on the basis that
the predicate wrong alleged was breach of statutory standards for which
the legislation itself provided comprehensive and exhaustive remedies —
essentially finding that, at least with respect to those Acts, the statutory
remedies had displaced all other private modes of redress, including
waiver of tort.
Koubi was indicative of a trend whereby British Columbia’s courts
took, overall, a somewhat more restrictive view of waiver of tort than
their Ontario counterparts, rejecting certification of claims in Reid v Ford
Motor Company20 (“Reid”), and Strata Plan LMS 3851 v Homer Street
Development Limited Partnership21 on the basis that “anti-harm” wrongs
such as negligence could not provide the predicate breaches to found the
restitutionary remedies available under the waiver of tort doctrine. In
Reid, Justice Gerow cited Networth Industries Ltd v Cape Flattery,22 for the
proposition that unjust enrichment could not be founded on negligence,
and therefore it was plain and obvious the claim for waiver of tort was
similarly bound to fail. She held:
[r]estitutionary claims are not made in negligence and nuisance because they
are in the main “anti-harm wrongs” in relation to which it is impossible, even
if they lead to an enrichment of the wrongdoer, to elevate the prevention of
18.

19.
20.
21.
22.

Aronowicz v Emtwo Properties Inc, 2010 ONCA 96 (shareholder dispute);
Arora v Whirlpool Canada LP, 2012 ONSC 4642 at para 300 (allegedly
defective laundry machines); Andersen v St Jude Medical Inc, 2012 ONSC
3660 (allegedly defective heart valves).
2012 BCCA 310.
2006 BCSC 712 [Reid].
2011 BCSC 569.
[1997] BCJ No 3174 (SC) [Networth].

312

Jones, Panacea or Pandemic

enrichment to the level of a primary purpose.23

The British Columbia Court of Appeal was somewhat more generous
with anti-competition claims, permitting waiver of tort pleadings to
survive certification in Pro-Sys Consultants Ltd v Infineon Technologies
AG.24 Justice Smith, writing for the Court, allowed that the plaintiffs
might not need to show damage if the doctrine were applied, and that
an aggregate monetary award could be certified as a common issue.
Similarly, in Steele v Toyota Canada Inc,25 Justice Hinkson (as he then was)
permitted a waiver of tort claim premised on breach of the provincial
competition legislation to proceed, also acknowledging that if waiver of
tort were an independent cause of action, proof of caused damage may
not be necessary in order for a global remedy to be available.
In Microsoft, the plaintiffs were indirect purchasers of Microsoft’s
operating systems who claimed that the software giant had conspired to fix
prices. Two judges at first instance had certified a claim including waiver
of tort;26 the majority of the Court of Appeal overturned certification
without expressly addressing that cause of action, simply holding that
indirect purchasers of Microsoft’s product had no competition claim
(only the dissenting judge, Justice Donald dealt with it, finding that it
did disclose a cause of action).27 On appeal to the Supreme Court of
Canada, the certification was reinstated, including waiver of tort as a
possible cause of action.28
There is one decision since Microsoft that deserves more than
mention. O’Brien was a products liability and failure to warn class action
involving the manufacturers of pelvic mesh implants. The claim involved
19 different products and thousands of class members, who were alleged
to have suffered one or more of a host of complications and injuries
23.
24.

Reid, supra note 21 at para 29, citing Networth, ibid at paras 24-26.
2009 BCCA 503, leave to appeal refused, [2010] SCCA No 32.

25.
26.

2011 BCCA 98.
Pro-Sys Consultants Ltd v Microsoft Corporation, 2010 BCSC 285; Pro-Sys
Consultants Ltd v Microsoft Corporation, 2006 BCSC 1047; and Pro-Sys
Consultants Ltd v Microsoft Corporation, 2006 BCSC 1738.
Pro-Sys Consultants Ltd v Microsoft Corporation, 2011 BCCA 186.
Microsoft SCC, supra note 2.

27.
28.

(2016) 2(1) CJCCL

313

as a result of the implants. Justice Perell, while allowing that waiver of
tort might be viable as a cause of action in some cases, said that where
mass tort claimants were pursuing possibly billions of dollars in personal
injury claims it was not a viable alternative pleading. I will have more to
say about Perell J’s concerns below.
So what is the state of the law in Canada today? Well, waiver of tort,
like Professor Schrödinger’s famous cat, presently appears to be both alive
and dead (or perhaps it is better to say neither alive nor dead) pending
an examination.29 In order to make out a claim it would be necessary for
the plaintiff to show a wrongdoing of the defendant (predicate wrong),
and a profit that has accrued to the defendant from the activity that was
unlawfully conducted. These conditions may in turn be subject to the
overriding objectives of equity itself.
At virtually every step of the way, lawyers representing defendants in
class actions have loudly — and as it turned out somewhat prematurely
— declared that waiver of tort, as an independent cause of action, was
dead or at least dying.30 But it has proven stubbornly resilient, and I
suggest that this is because the problem that it is being used to remedy is
real: what happens when the plaintiffs can establish (i) that the defendant
29.

30.

Physicist Erwin Schrödinger proposed a thought experiment in 1935
in which decaying radioactive material would either kill a hidden cat or
not. According to quantum theory of superposition, the cat would be
both alive and dead until it was observed to be one or the other, a result
Schrödinger regarded as absurd.
See e.g. Brandon Kain, “Waiver of Tort Gets a Reality Check at the
B.C. Court of Appeal” (23 July 2012), Canadian Appeals Monitor,
online: <www.canadianappeals.com/2012/07/23/waiver-of-tort-getsa-reality-check-at-the-b-c-court-of-appeal/>; Christopher Naudie
et al, “Class Actions Development: B.C. Court of Appeal Slams the
Door on Waiver of Tort in Statutory Cases” (20 July 2012), Osler,
Hoskin & Harcourt LLP, online: < https://www.osler.com/en/resources/
critical-situations/2012/class-actions-development-b-c-court-of-appealsl>; Peter Kryworuk & Rebecca Case, “Waiver of Tort — So Long,
It’s Been Nice to Know You?” (2 September 2015), Lerners (blog),
online: <http://lernersclassactionplaintiff.ca/blog/post/waiver-of-tortso-long-it-s-been-nice-to-know-you?utm_source=Mondaq&utm_
medium=syndication&utm_campaign=View-Original>.

314

Jones, Panacea or Pandemic

committed a breach of some statutory or common law duty and; (ii) that
the breach caused a certain amount of harm in persons who have been
exposed through purchase or use of a product; but (iii) it is difficult or
impossible to link, on a balance of probabilities, any particular victim
with the wrong?
If waiver of tort exists, what are the elements in modern Canadian
law? What constitutes a predicate wrong? Does negligence qualify? Some
breaches of statute (competition laws, for instance) seem to be covered,
where others that contain exhaustive recovery regimes (consumer
protection legislation) may not be. Who can claim under the doctrine, if
this is not to be determined by who has been harmed by the wrong? These
questions might eventually be answered with reference to the equitable
principles underlying the cause of action. So we might ask, first, whether
the wrong is such “that in equity and good conscience [the] defendant
should not be permitted to retain that by which it has been enriched”;31
and second, whether there is some connection or nexus between the
plaintiff (or class) and defendant such that the former may equitably
pursue and receive the benefit of the disgorgement. But at this point all
of these matters remain unresolved.
I should conclude my discussion of the Canadian cases by disposing
of one issue that has caused some angst among both academics and
jurists: is waiver of tort a stand-alone cause of action, or is it simply
a remedy?32 If the latter, the doctrine would only permit disgorgement
where a complete wrong (usually including proof of caused loss) is
independently established.
As the Supreme Court of Canada noted in Microsoft: “[t]he U.S.
and U.K. jurisprudence as well as the academic texts on the subject
have largely rejected the requirement that the underlying tort must be

31.
32.

Federal Sugar Refining Co v US Sugar Equalisation Board, 268 F 575
(SDNY 1920 (US)) at 582 [Federal Sugar].
The controversy is well described in Serhan Div Ct, supra note 5 at paras
45-76, and I need not detail it here.

(2016) 2(1) CJCCL

315

established in order for a claim in waiver of tort to succeed”.33 The Court
appeared less persuaded by other cases restricting the doctrine to cases
where the full tort, including proof of loss, was required34 and permitted
the waiver claim to proceed, apparently on the assumption that it might
provide an independent basis for recovery, notwithstanding Microsoft’s
assertion that it had been pleaded only as remedy.
But in any event, the Supreme Court’s analysis accords with a practical
reality: if providing a remedy (disgorgement of profits) is all waiver of
tort does, then it is a weak doctrine in most cases. Indeed it would be
largely redundant with punitive damages, which courts recognize can,
and perhaps often should, be measured by the profits of the defendant,
with the equitable and deterrence objectives of depriving the defendant
the fruits of his wrongdoing.35
It is difficult to see how equitable disgorgement, as a remedy, could
improve on this, and indeed it appears far less efficacious than punitive
damages which can be increased or decreased according to other relevant
factors, including ensuring proper deterrence.
So if waiver of tort is to mean anything, it must be an independent
cause of action, one that can succeed where ordinary tort or contract
claims cannot, and the remainder of my analysis of its efficacy and
desirability is premised on this, more robust, view of the doctrine.

33.

34.

35.

Microsoft SCC, supra note 2 at para 96, citing inter alia, National Trust
Co v Gleason, 77 NY 400 (App Ct 1879 (US)); Federal Sugar, supra note
31; Mahesan v Malaysia Government Officers’ Co-operative Housing Society
Ltd, [1979] AC 374 (PC (Malaysia)); Universe Tankships Inc of Monrovia v
International Transport Workers Federation, [1983] 1 AC 366 (HL).
Microsoft SCC, supra note 2 at para 96, citing United Australia Ltd v
Barclays Bank Ltd, [1941] AC 1 (HL) at 18; Zidaric v Toshiba of Canada
Ltd (2000), 5 CCLT (3d) 61 (Ont Sup Ct J) at para 14; Reid, supra note
20.
Whiten v Pilot Insurance Co, 2002 SCC 18 (where the Supreme Court of
Canada held “it is rational to use punitive damages to relieve a wrongdoer
of its profit where compensatory damages would amount to nothing more
than a licence fee to earn greater profits through outrageous disregard of
the legal or equitable rights of others” at para 72).

316

Jones, Panacea or Pandemic

III.

What is Wrong With a Waiver of Tort Claim

A.

What is the Problem, Exactly?

I propose here that there are two categories of cases where courts might
apply the doctrine. They suggest two different problems, and the fact that
waiver of tort is good at addressing one and poor at the other leads to my
conclusion that it should not be a preferred doctrine in most mass-wrong
cases.
The first category consists of cases where a defendant has profited
from a wrong, but there has been no loss to the plaintiff giving rise to
an ordinary cause of action. In such circumstances, waiver of tort can
provide a measure of deterrence of activity that is potentially harmful
and antisocial but did not cause harm “this time around”. Whether
any particular wrong should give rise to disgorgement is a question rife
with policy considerations. Presumably, the intervention of the courts
is warranted to deter wrongs of a type which society has an interest
in absolutely prohibiting (such as a crime), instead of just regulating
(such as, at least on one view, negligence). But even after this question
is answered, the court might further ask: is the plaintiff or class the
appropriate “prosecutor” of such an action, given that it is essentially
regulatory in ambition?
The second category of cases to which waiver might apply is where we
know the defendant has committed complete torts — that is, we know
some people have been harmed by the wrong but we simply cannot tell
which ones. Typically, this will arise in the context of mass claims that can
be prosecuted through the device of the class proceeding. In such cases,
I will argue, resort to waiver of tort is inappropriate and ineffective, and
a much more straightforward solution — the determination of causation
class-wide, in a population of persons exposed to the risk of harm — is
available and preferable.
In order to justify my arguments in this regard, I must review the
objections that are taken to waiver of tort as a cause of action. I hope
to show that the problems with the doctrine are most acute in the
second category of cases (i.e. where harm has in fact occurred but is
indeterminate), and are insignificant, or at least manageable, in the first

(2016) 2(1) CJCCL

317

category.

B.

Wrongs “In the Air”

We know that crimes, fraud and other malfeasance can form the basis
of a waiver of tort claim. But what of negligence-based cases, such as
products liability or toxic tort claims? Should different principles apply?
Some courts have expressed discomfort, at least in obiter, that waiver of
tort claims could relate to “anti-harm” wrongs like negligence, which are
overtly premised on compensatory rather than regulatory principles.36
Causation — not just factual but legal causation — has proven to be an
important limiting device in negligence law, and not just from the point
of view of compensation objectives. Tort scholars might disagree on the
reason why the right to pursue negligence claims has been limited to
those whom the defendant has harmed through its fault, but it certainly
has. And it may be that disconnecting the causation link altogether
will unnecessarily depress socially or economically useful activity as
“busybody plaintiffs” (or more likely lawyers) set themselves to ferreting
out “wrongdoing” and launching a wave of litigation with no good
purpose.
So there are good arguments why waiver of tort claims have never
been, and should not be, prosecuted entirely “at large”: that is, a person
who is a complete stranger to the defendant and the wrong ought not be
able to sue to obtain the defendant’s profits. On this view, there should
be some relationship between the wrong and the plaintiff sufficient to
permit the plaintiff to obtain standing on the basis that the plaintiff is in
an equitable position to pursue the claim.
I would suggest this nexus could be established on a couple of
different bases. First, if the plaintiff (or the class) was “within the ambit
of the risk” of harm created by the defendant’s wrong,37 they could
36.
37.

See for instance Reid, supra note 20 at paras 15, 29; Serhan Div Ct, supra
note 5 at paras 66-67.
Individual plaintiffs who cannot prove causation but who fall within
“the ambit of the risk” may be able to take advantage of exceptions to the
traditional “but for” test in individual causation cases: Resurfice Corp v
Hanke, 2007 SCC 7 at para 25 [Resurfice].

318

Jones, Panacea or Pandemic

claim. So, for instance, consumers of a product, purchasers of shares, or
victims of diseases epidemiologically linked with a pollution source or
toxic substance might qualify as appropriate plaintiffs in a waiver of tort
claim, where mere bystanders or busybodies might not. Thus in Serhan,
for instance, it made sense to permit the purchasers of the medical device
to bring the (eventually settled) waiver of tort claim. Another way of
assessing the proposed plaintiff might be whether there will be some
remedy (or settlement term) aside from simple disgorgement that will
benefit the plaintiff. So, to again use Serhan as an example, each user of
the diabetes testing device was, as part of the settlement, entitled to a
replacement device.
In many class action cases, the “busybody” problem will not arise
because waiver is almost always pleaded in the alternative to tort, fraud
or unjust enrichment claims, and so the class is defined with regard to
persons who have suffered harm as the result of the wrongs. If those
pleadings survive certification, which would require that the tort claims
have, at least “some basis in fact”,38 then this might provide a sufficiently
restrictive class who should have standing to pursue waiver claims in the
alternative.
Nevertheless, the separation of wrongdoer from victim that waiver of
tort entails must be counted among the difficulties to a widespread use
of the doctrine.

C.

Under-Compensation and Under-Deterrence

Another obvious but more serious problem with resort to waiver of tort
emerges in cases where we can know that the defendant’s wrongdoing
did in fact cause some harm. In some cases, application of the doctrine
might represent a windfall for plaintiffs, who might have suffered little
or no loss from a wrong associated with a highly profitable product or
activity of the defendant. In many other cases, the harm will be of such
a magnitude that any disgorgement of profits will be little more than
38.

Brandon Kain, “The Supreme Court of Canada and the Still-Curious
Requirement of ‘Some Basis in Fact’” (2015) 68 Supreme Court Law
Review (2d) 77.

(2016) 2(1) CJCCL

319

token.
The main objectives of tort law are usually expressed in terms of
compensation and deterrence. Though scholars will disagree about
which has primacy, the question is usually moot: tort law operates by
making the defendant pay the cost of the harm it has caused to the
person it has injured. This achieves compensation for the victim and
forces the defendant to internalize the cost of the harm, thus providing
the economically-optimal level of deterrence, or regulation, of the risky
activity.39
Disgorgement, as a sole remedy, upsets this balance. Certainly, in
cases where the plaintiff has suffered little or no harm, the disgorgement
of profits provides some disincentive to defendants to engage in risky
or other antisocial behaviour. So we see cases like Strand Electric and
Engineering Co Ltd v Brisford Entertainments Ltd,40 where Lord Denning
observed that in “[the] cases where the defendant has obtained a benefit
from his wrongdoing he is often made liable to account for it, even
though the plaintiff had lost nothing and suffered no damage”.41
But reliance on disgorgement as a deterrent is inferior to a system
based on damages, particularly where “habitual defendants” may be
systemically incentivised to take risks in the pursuit of profit. If all a
defendant risks are its profits, it might, overall, take more risks to obtain
more profits, on the expectation that there is no real “downside” to
doing so. If a wrongdoer gets caught, it gains nothing but loses nothing
either. If it does not get caught, it retains the benefit of the wrong. If it
internalizes the true costs of the harm it has caused, on the other hand,
the equilibrium provided by tort law is restored.
So in cases where the defendant’s wrong has not actually caused harm
39.

40.
41.

See e.g. Guido Calabresi, The Costs of Accidents: A Legal and Economic
Analysis (New Haven: Yale University Press, 1970)(the still-resilient “law
& economics” analysis of tort law); see also Richard A Posner, “A Theory
of Negligence” (1972) 1:1 Journal of Legal Studies 29; Richard A Posner,
Economic Analysis of Law, 6d (New York: Aspen Publishers, 2003);
Richard A Posner, “Utilitarianism, Economics, and Legal Theory” (1979)
8:1 Journal of Legal Studies 103.
[1952] 1 All ER 796 (CA).
Ibid at 800.

320

Jones, Panacea or Pandemic

“this time around”, waiver of tort and disgorgement provides a rational
means of providing some deterrence for activity that is inherently
antisocial or risky, but did not cause harm in the particular event (such
as in Serhan itself ).42 The difficulty is that, in the majority of cases,
some harm has actually been caused; the problem is that the connection
between the defendant’s wrong and each member of a plaintiff class
cannot be confidently established.

D.

The Divergence of Interests of Class Counsel

There is one further problem with reliance on waiver of tort that may
lead to systemic under-deterrence and under-compensation. That is
that recovery under the doctrine is so easy that it may actually prevent
legitimate damages claims from being brought, or at least from being
aggressively pursued.
The success of class actions depends on the interests of class counsel
being aligned with the class members’ own. The fact that this will not
always be so has influenced many aspects of the class procedure, such as the
requirements that class members be given notice of proposed settlements
and an opportunity to object, and that settlements be approved by the
court. The temptation is always that plaintiffs’ counsel can, implicitly
at least, collude with defendants to produce a quick settlement for a
small amount that provides substantial payment for the lawyers but little
42.

It is tempting to say, as a consequence of the distinction I propose here
(between cases where there is no damage versus true “indeterminate
causation” cases where damage is known but particular victims cannot
be certainly identified), that waiver of tort is never required in class
actions, because in the cases where I propose that it should be available
could be as easily pursued as individual actions. However this is probably
too simplistic a view for three reasons: (i) because waiver of tort may
legitimately be pleaded in the alternative in cases where it is not certain
into which category the claim should properly fall; (ii) because it
may be more fair or just to distribute the disgorgement more broadly
than to a single individual; and (iii) because a class action will dispose
simultaneously of the claims of all persons who would have standing to
pursue them individually, therefore avoiding an inefficient multiplicity of
competing claims for the same remedy.

(2016) 2(1) CJCCL

321

benefit for the members.43
In O’Brien, referred to earlier, Perell J found that even if it were a
valid cause of action conceptually, waiver of tort should not be permitted
to proceed to trial on the facts before him, not because it did too much
but because it did too little:
[i]n Serhan v. Johnson & Johnson, which is the case that started the debate
about the nature of waiver of tort, there were zero monetary damages for the
tort claim and waiver of tort was the route to access to justice and behaviour
modification. In the case at bar, assuming Bard were negligent, a waiver of
tort cause of action would not provide access to justice to class members
or any meaningful behaviour modification. It would be reprehensible for
Class Counsel to take a contingent fee based on an award calculated on
the disgorgement of profits. A judgment or a settlement based on waiver of
tort would create enormous conflicts between Class Members as to how the
disgorged funds should be distributed. It would be a waste of the court’s and
the parties’ litigation resources to expend discovery and trial time calculating
what profits, if any, Bard made from its Pelvic Mesh Products, when assuming
liability, everybody should be spending their litigation resources calculating
compensatory damages. In my opinion, in these circumstances, regardless
of whether waiver of tort is a reasonable cause of action, it would not be
reasonable to prosecute it as a class action. Even if the pleading of waiver of
tort satisfied the cause of action criteria, the class definition, and the common
issues criteria, in my opinion, the waiver of tort claim in the circumstances of
the case at bar would not satisfy the preferable procedure and the representative
plaintiff criteria. In these circumstances, I conclude that the waiver of tort
claim in the case at bar does not satisfy the cause of action criterion for a class
action. I would not certify the waiver of tort claim.44

One might quibble with Perell J’s mixing of the “cause of action” criterion
with the other threshold requirements of class certification, particularly
“preferability”. And his concern over the problems of distribution of a
class-wide award, which he saw as reflecting on the question of whether

43.

44.

See generally Bruce Hay & David Rosenberg, “‘Sweetheart’ and
‘Blackmail’ Settlements in Class Actions: Reality and Remedy” (2000)
75:4 Notre Dame Law Review 1377.
O’Brien, supra note 8 at paras 162-65.

322

Jones, Panacea or Pandemic

the representative plaintiff was appropriate, may have been overstated.45
But at least this one of his concerns over the use of the doctrine should
cause serious reflection: he worried that the class in O’Brien would be
“sold out” if their recovery were limited only to recovery of the profits
when the harm alleged was of a much higher magnitude.
When we come to appreciate that plaintiffs’ counsel in the class
action bar approach their work from the point of view of investment and
return, we can apprehend that there will come a point where a reliance
on waiver of tort actually undermines the compensatory and behaviourmodification objectives of both tort law generally, and class actions in
particular.
In the popular book Freakonomics, the authors identified a structural
conflict of interest between realtors and their clients. The difficulty arises
from the commission structures adopted by the industry: a realtor paid
on commission has a comparatively small interest in maximising his
client’s selling price: that is, a realtor being paid 3% commission on the
sale of a house worth $500,000 “loses” only $1500 if she quickly sells
the house for a $50,000 discount. It will generally be easier to invest the
realtor’s efforts in selling two such houses quickly and cheaply (earning
$27,000 commission) rather than taking the same time to sell one house
for its full value (earning only $15,000 commission), and the authors
suggested that a study showed that is exactly what realtors did.46
45.

46.

Ibid at paras 125-26 (the distribution problem Perell J identified would
apply to any lump-sum award, including punitive and exemplary damages
in tort. The deeper problem was the plaintiff’s attempt to certify a single
class for 19 distinct products, each with its own history and each with a
different group of alleged victims).
Stephen J Dubner & Steven D Levitt, Freakonomics: A Rogue Economist
Explores the Hidden Side of Everything (New York: Harper Collins, 2005)
at 5-8, (The methodology was ingenious: the researchers compared
realtors’ behaviour when listing and selling their own homes versus their
clients’. They found that realtors took longer to sell their own properties,
and realized higher sale prices. It might be observed as an aside that
commission structures that are non-linear, such as those providing 7% on
the first $100,000 and 2% thereafter (usually split between the buyer’s
and seller’s agents), are even more perverse in their incentives).

(2016) 2(1) CJCCL

323

Justice Perell’s caution in O’Brien seems well placed for the same
systemic reasons. If the plaintiffs’ counsel can get an easy but small
recovery from a waiver of tort claim, why should they pursue lengthy,
convoluted tort actions, at comparatively great investment risk? Judge
Henry Friendly recognized the inclination in the context of an individual
wrong in Alleghany Corp v Kirby,47 when he wrote that a lawyer has “every
incentive to accept a settlement that runs into high six figures or more
regardless of how strong the claims for much larger amounts may be …
[A] juicy bird in the hand is worth more than the vision of a much larger
one in the bush”.48
Yet if lawyers do not pursue the full measure of the harm, the tort
“market” is distorted: defendants internalize only a fraction of the harm
they have caused, and the class members receive only a fraction of their
true losses.
So either of the present approaches — insistence on individual
attribution of harm (with recovery denied in each case, even if the
claims are economically viable one-by-one), or replacing that with
limited recovery based on the profits of the defendant — are chronically
unsatisfactory and will demonstrably lead to under-compensation and
under-deterrence. Is there a better way?

IV.

A Better Way

A.

The Problem, Reiterated

As I hope is now apparent, the downsides of resorting to waiver of tort
in large-scale claims are most acute in mass torts where it is possible to
determine that the defendant’s wrong has caused harm, but each plaintiff,
or member of a class exposed to the risk of harm, cannot establish a causal
link between the wrong and the damage they have suffered.
In a class action, it is possible to assess the harm on a collective basis,
as harm caused within a population of persons, without the need to prove
that any particular class member’s harm was the result of a particular
47.
48.

333 F (2d) 327 (2nd Cir 1964 (US)).
Ibid at 347.

324

Jones, Panacea or Pandemic

defendant’s misconduct. Damages can also be assessed in the aggregate,
with only the problem of distribution remaining. As I will mention a bit
later, this is often the basis of court-approved class action settlements.
The difficulty is that class actions are generally regarded as only a
procedural device.49 Class proceedings statutes generally permit the
calculation of quantum of damages on an aggregate basis, but only once
liability has been established.50
This permits defendants to argue that, even if there can be collective
determination of “general causation” (i.e. that the defendant’s wrong can
cause the type of harm alleged, or even that, viewed in the aggregate,
it did cause harm),51 the claim cannot be legally made out until it is
known which class member actually suffered the harm from the wrong.
The Supreme Court of Canada has, from time to time, drank of this
water, as when it said that the judge “must still be satisfied on a balance of
probabilities that each element is present for each member”.52 In cases of
indeterminate causation, it simply cannot be. As such, in mass tort cases
where causation in individuals is indeterminate, tortious harm that can
be plainly seen in the aggregate may go unaddressed by the tort system,
which provides only, as some have called it, a “phantom remedy”.53
49.

50.

51.

52.
53.

Dell Computer Corp v Union des consommateurs, 2007 SCC 34 (a class
action is “only a legal procedure” at para 106); Bisaillon v Concordia
University, 2006 SCC 19 (a class action “neither modifies nor creates
substantive rights” at para 17).
See Ontario’s Class Proceedings Act, 1992, SO 1992, c 6, s 24(1) [Ont
Class Proceedings Act]; or identically BC’s Class Proceedings Act, RSBC
1996, c 50, s 29(1).
The same evidence that is used to establish general causation - that is that
the wrong creates a risk of harm - is evidence for the proposition that
it did in fact cause harm in a population, though the particular victims
might not ever be identified.
Bou Malhab v Diffusion Metromedia CMR Inc, 2011 SCC 9 at para 53
[emphasis added].
William R Ginsberg & Lois Weiss, “Common Law Liability for Toxic
Torts: A Phantom Remedy” (1981) 9:3 Hofstra Law Review 859
(examining how the economic realities of individual litigation, combined
with causation and other challenges, effectively preclude individual claims
for compensation in toxic tort cases).

(2016) 2(1) CJCCL

325

A decade ago, Professor Jamie Cassels and I proposed that, in mass
toxic claims, aggregate assessment of causation may prove not only
necessary, but superior to assessments done case-by-case (in the sense that
the former is both more accurate and more fair).54 In 2005, we wrote:
[c]ausation of harm in the aggregate becomes clearer even as the individual
identity of the victims, and their individual connection with each wrongdoer,
is lost… in our opinion, viewing inherently probabilistic causation in the
aggregate — as a definite harm in a percentage of the population rather than
a probabilistic harm in an individual — provides several advantages in the
resolution of mass tort claims.55

In 2011, while the individual causation case of Clements v Clements,56
was before the Supreme Court of Canada but before it was decided,
I reiterated my concern that individual causation rules should not be
crafted so as to frustrate mass tort claims:
[i]n such instances, we know that the defendant has, in fact, caused a certain
number of the injuries suffered in the population. We simply do not know
which of the afflicted were harmed by the defendant, and which would have
been injured in any event. Why should this be an insurmountable obstacle to

54.

55.
56.

See for instance, Craig Jones, Theory of Class Actions (Toronto: Irwin Law
Book, 2003); Jamie Cassels & Craig Jones, The Law of Large Scale Claims:
Products Liability, Toxic Torts and Complex Litigation in Canada (Toronto:
Irwin Law Book, 2005) [Cassels & Jones, Large Scales]; Craig Jones,
“Reasoning Through Probabilistic Causation in Individual and Aggregate
Claims: The Struggle Continues” (2011) 39:1 Advocates’ Quarterly 18
[Jones, “Reasoning Through Probabilistic Causation”]; Jamie Cassels &
Craig Jones, “Rethinking Ends and Means in Mass Tort: Probabilistic
Causation and Risk-Based Mass Tort Claims after Fairchild v. Glenhaven
Funeral Services” (2003) 82 Canadian Bar Review 597 [Cassels & Jones,
“Rethinking Ends and Means in Mass Tort”]. My and Professor Cassels’
work in this area was heavily informed by that of David Rosenberg,
particularly two foundational articles: David Rosenberg, “Class Actions
for Mass Torts: Doing Individual Justice by Collective Means” (1987)
62:3 Indiana Law Journal 561; and David Rosenberg, “Individual Justice
and Collectivizing Risk-Based Claims in Mass-Exposure Cases” (1996)
71:1&2 New York University Law Review 210.
Cassels & Jones, Large Scales, ibid at 208.
2012 SCC 32 [Clements].

326

Jones, Panacea or Pandemic

tort law?57

Professor Cassels and I had written these things as a series of cases,
especially Fairchild v Glenhaven Funeral Services Ltd58 and Resurfice Corp
v Hanke,59 suggested a relaxation of the rules of causation permitting
persons “within the ambit of the risk” to bring personal injury claims
notwithstanding that they could not demonstrate causation. We posited
that these techniques, applied in the context of a class action, could
considerably ease the problem of indeterminate causation in mass torts.
In this sense, it was disappointing that the Court seemed determined to
rein in the idea of “probabilistic causation” in the individual personal
injury case of Clements.
But it appeared that the Supreme Court was alive to the problem,
and might be prepared to relax the rules of causation in mass tort class
actions. In Clements, the Chief Justice (writing for a unanimous Court on
this point), upheld and reiterated the individualistic “but for” test of tort
causation in single cases, but then said this:
[t]his is not to say that new situations will not raise new considerations. I leave
for another day, for example, the scenario that might arise in mass toxic tort
litigation with multiple plaintiffs, where it is established statistically that the
defendant’s acts induced an injury on some members of the group, but it is
impossible to know which ones.60

B.

Assessment of Liability, Harm, and Damages on an
Aggregate Basis

Let us take the example of a mass tort involving the exposure to a toxic
substance or the use of a defective product, where evidence could establish
a probability of harm over and above “background risk”, and that this
harm could be attributable to the defendant’s wrong (sometimes called
“general causation”). Let us further suppose that we could identify a class
of persons who were exposed to the risk (consumers, or people living in
57.
58.
59.
60.

Jones, “Reasoning Through Probabilistic Causation”, supra note 54 at 22
[emphasis in original].
[2002] UKHL 22.
Resurfice, supra note 37.
Clements, supra note 56 at para 44.

(2016) 2(1) CJCCL

327

a geographic area) and whom had suffered harm of a type that would be
expected from the wrong. But even if we know that some of that group
was actually harmed by the defendant, we also know that others would
have suffered the harm without the occurrence of the wrong. We just do
not know who is who.
The ordinary tort system would provide no remedy, because
attribution of any individual’s harm to the defendant’s wrong would be
impossible.61 However if we calculated the harm in the aggregate, we could
impose liability on the defendant and recover damages in the amount of
the harm they had caused. We would still, of course, have a problem of
distributing the proceeds among the sufferers of harm. But the regulatory
function of tort would be preserved through proper deterrence, and the
longstanding tradition that negligence should be regulated by persons
who had been harmed and only to the extent that it has caused harm,
would be respected. Yes, it is only part fulfillment of tort’s compensation/
deterrence objectives, but it does provide the potential for some relief
of victims, and more importantly, through deterrence, it helps avoid
all harm to future victims who would be created if the tort system did
nothing.
Tobacco litigation is paradigmatic of the problem, and has also been
fertile ground for innovative solutions. The diseases caused by tobacco
are, in the main, elusive of individual attribution: it is very difficult for
an individual smoker who suffers from emphysema, or cancer, to prove
with any certainty that he would not have contracted the disease “but
for” smoking. All we can say for sure is that smoking increased the risk of
the disease. But increased risk in an individual means increased prevalence
61.

This is a problem for plaintiffs only, of course, if the background risk
was higher than the probability of the harm resulting from the wrong. I
suggest that this is the case in most toxic torts and many products liability
claims, but I allow that in some cases the “balance of probabilities” in the
individualistic system could result in every claimant succeeding even where
we know that some of the harm was not defendant-caused. This does not,
in my view, weaken the case for aggregate assessment of causation that I
make in this article, and in fact the opposite: it is fairer to both defendants
and plaintiffs because it neither over- or under-deters, and exacerbates
neither the “sweetheart” or “blackmail” settlement problem.

328

Jones, Panacea or Pandemic

in a population. We can know with scientific certainty that some smokers
with cancer would not have got the disease “but for” smoking, and we
may even be able to know, with some certainty, how many. To move
to a further level of abstraction, if the smoker’s claim is based on a
failure to warn, the question of whether an adequate warning would
have prevented the smoking is elusive in an individual case, while we
can at the same time know that warnings do reduce the prevalence of
smoking in populations. Thus, it should be possible to determine how
much tobacco-related disease could have been prevented by an adequate
warning, even if each individual’s claim must, under the principles of tort
law, fail. We can then place a dollar figure on the global loss.
In British Columbia a statute, the Tobacco Damages and Health Care
Costs Recovery Act62 (“BC Tobacco Act”), swept aside the particularistic
rules of tort in favour of an “aggregate action” by the government to
recover damages caused by tobacco-related wrongs, regardless of whether
any particular smoker could prove a complete tort. Rules were introduced
permitting liability, harm and damages to be assessed collectively. The
government filed its suit, and the defendant manufacturers challenged
the BC Tobacco Act, inter alia, on the basis that such rules were unfair.
The government argued that aggregate assessment of damages in cases of
causal indeterminacy was superior to individual adjudication, because
harm in populations can be more, rather than less, accurately measured
as a whole rather than as a sum of parts. In British Columbia v Imperial
Tobacco,63 Justice Major, writing for the unanimous Supreme Court of
Canada, wrote:
[t]he rules in the Act with which the appellants take issue are not as unfair
or illogical as the appellants submit. They appear to reflect legitimate policy
concerns of the British Columbia legislature regarding the systemic advantages
tobacco manufacturers enjoy when claims for tobacco-related harm are
litigated through individualistic common law tort actions.64

This echoes the view of the trial judge, who had found in 2000, when
reviewing an earlier iteration of the BC Tobacco Act that:
62.
63.
64.

SBC 2000, c 30.
2005 SCC 49.
Ibid at para 49.

(2016) 2(1) CJCCL

329

[t]he basic tenet that causation within a population may be more accurately
identified statistically than by means of attribution of individual causation in
a multiplicity of conventional tort-based actions appears sound. The use of
statistical and epidemiological evidence is an essential aspect of an aggregate
action. The question in issue becomes causation in the group rather than of
any individual member.65

Justice Holmes reiterated this endorsement when a revised version of the
BC Tobacco Act came before him again in 2003:
[t]here is nothing inherently unfair about an aggregate action. In fact, it
may often balance unfairness of proceeding either by individual actions or
by other forms of collective proceeding. Neither is the use of statistical or
epidemiological evidence itself evidence of an unfair trial. They are aids to the
resolution of issues in a unique but appropriate form of action.66

The breathtaking possibilities of the true aggregate approach have been
further demonstrated in the recent decision of the Quebec Superior
Court in a pair of tobacco-related class actions, Blais v JTI-Macdonald
Corp (“Blais”) and Létourneau v JTI-Macdonald Corp67 (“Létourneau”).
In those cases the plaintiffs had the benefit of some statutory provisions
that eased their problems of proof and permitted the trial judge to assess
liability, harm and damages in the aggregate.68
The claim in Blais was by a class of smokers who had contracted
65.
66.
67.
68.

JTI-Macdonald Corp v British Columbia (Attorney General), 2000 BCSC
312 at paras 74-75.
R v Imperial Tobacco Canada Ltd, 2003 BCSC 877 at para 156.
2015 QCCS 2382 (sub nom Blais v JTI-Macdonald Corp 2012 QCCS
469) [Létourneau].
Quebec’s Tobacco-Related Damages and Health Care Costs Recovery Act,
RSQ c R-2.2.0.0.1 (“[i]n an action brought on a collective basis, proof
of causation between alleged facts, in particular between the defendant’s
wrong or failure and the health care costs whose recovery is being sought,
or between exposure to a tobacco product and the disease suffered by,
or the general deterioration of health of, the recipients of that health
care, may be established on the sole basis of statistical information or
information derived from epidemiological, sociological or any other
relevant studies, including information derived from a sampling”. This
provision is made applicable to class actions by the last paragraph of
section 25, which states that the rules in, inter alia, section 15 “also apply
to any class action based on the recovery of damages for the (tobaccorelated) injury”: ss 15, 25).

330

Jones, Panacea or Pandemic

cancer and emphysema, allegedly from smoking. In Létourneau, the class
members were smokers who were claiming for the harm of addiction. In
both classes, individual causation was, at least, uncertain: given the nature
of disease processes and other causal issues such as causation-reliance, no
one could conclusively attribute any particular disease to smoking, nor
disease or addiction to the wrong of the defendant.
The fundamental finding of the court was that the three Canadian
manufacturers of cigarettes were at fault because they did not warn of the
dangers inherent in their product, adequately or at all. But with the breach
demonstrated, what then? Did the plaintiffs still have to show that every
individual for whom recovery was sought had suffered a harm causally
linked to the defendant’s wrong? Justice Riordon did not think so. He
documented a 50-year campaign of deception and misinformation, and
then went on to impose staggering damages ($15 billion, after interest)
despite the fact that there had been no evidence that any particular class
member had suffered harm.
The specific provisions of the Quebec legislation provided only that
statistical and epidemiological evidence, including sampling, could be
used to establish liability as well as damages. The defendants argued that
this did nothing more than permit questions of individual harm to be
decided with resort to such evidence. It did not, they suggested, extend to
permitting proof of causation in populations and assessment of causation
on an aggregate basis.
Justice Riordon rejected this interpretation. He wrote:
[t]he objective of the TRDA is to make the task of a class action plaintiff
easier, inter alia, when it comes to proving causation among the class members.
When the legislator chose to favour the use of statistics and epidemiology,
he was not acting in a vacuum but, rather, in full knowledge of the previous
jurisprudence to the effect that each member of the class must suffer the same
or similar prejudice. It thus appears that the specific objective of the act is to
move tobacco litigation outside of that rule.69

This, the trial judge observed, effectively overrode the “previous
jurisprudence calling for proof that each member suffered a similar

69.

Létourneau, supra note 67 at para 692.

(2016) 2(1) CJCCL

331

prejudice”.70 In Riordon J’s analysis, this led to the conclusion that
individual proof of causation was unnecessary altogether.
The Blais and Létourneau decision represents the first clear Canadian
manifestation of what is, if class actions are to fulfill their promise as an
effective compensatory and regulatory device,71 inevitable: the adaptation
of substantive law of causation in tort to accommodate the scale and
difficulties associated with truly massive wrongs.
Justice Riordon in Blais and Létourneau seemed sufficiently pleased
with the aggregate approach that he mused openly about whether the
techniques should be available in all class actions. He said:
[i]t will be interesting to see if the National Assembly eventually chooses
to broaden the scope of this approach to have it apply in all class actions.
Although such a move would inevitably be challenged constitutionally, its
implementation would go a long way towards removing the tethers currently
binding class actions in personal injury matters.72

Justice Riordon’s decision, while singular in the Canadian jurisprudence,
was not entirely without precedent. Judges faced with massive claims
spanning large periods of time have before been driven to techniques of
“wholesale justice” in order to “fit the forum to the fuss”. I have described
and discussed the resulting innovations, such as “market share”73 or

70.
71.

72.
73.

Ibid at para 693.
There are three commonly-accepted objectives of class actions in Canada:
compensation; behaviour modification (that is to say, the deterrence of
wrongdoing); and “access to justice”. The third objective may be seen as
valuable principally to the extent that it facilitates the first two, although
it can be argued also to be an independent social good.
Létourneau, supra note 67 at para 693, n 319.
Market share liability operates in individual cases as well as class actions,
and holds defendants liable on the basis of their risk contribution, where a
causal nexus between victim and wrongdoer can’t be established: Sindell v
Abbott Laboratories, 26 Cal (3d) 588 (Sup Ct 1980 (US)). In Canada, the
theory has been permitted to proceed through certification in Gariepy v
Shell Oil Co [2000] 52 OR (3d) 181 (Sup Ct J) at para 11, and referred to
as a potential claim in Ragoonanan Estate v Imperial Tobacco Canada Ltd
[2000] 51 OR (3d0 603 (Sup Ct J) at para 27.

332

Jones, Panacea or Pandemic

“sampled” liability74 and other rules facilitating proof of causation in
populations, extensively elsewhere.75
Recent Ontario decisions seem also to hint at the willingness
of courts to entertain questions of liability on an aggregate basis. In
Ramdath v George Brown College76 (“Ramdath”), the defendant college
had negligently misrepresented that completion of its courses would lead
to three professional designations, something that was not true. At the
trial of the common issues, Justice Belobaba found that the defendants
had breached their duty to the plaintiffs both under the Consumer
Protection Act, 200277 and under negligence law, but noted that this alone
would not entitle them to recovery for the tort without proof that each
had relied on the misrepresentation. He wrote:
[f]urther evidence may still be needed to establish legal liability for negligent
misrepresentation, namely, evidence of individual reliance. This question will no
doubt be addressed in the next phase of this litigation. However, legal liability has
been established under the CPA because, as already noted, under this statute,
evidence of actual reliance is not required … The common issues trial has
now been concluded. The next step in this class proceeding is to schedule a case
conference to discuss the “damages” phase of this lawsuit. Counsel should contact
my office to arrange a convenient date for the case conference.78

74.

75.
76.
77.
78.

In Hilao v Estate of Marcos, 103 F (3d) 767 (9th Cir 1996 (US)) the
Court directed that the class of claimants against the late Philippine
dictator for human rights abuses could be sampled to determine the
number of valid claims, to permit a global assessment of damages. Similar
approaches adopted by federal trial judges in the US, however, were
subsequently disfavoured by appeal courts: see e.g. Cimino v Raymark
Industries, Inc, 151 F (3d) 297 (5th Cir 1998 (US)) and McLaughlin
v Philip Morris USA Inc, 522 F (3d) 215 (DC Cir 2009 (US))(these
decisions, which found that abandoning individual proof of causation was
a violation of the 5th and 7th Amendments to the US Constitution or
the “predominance” requirements of US Federal Rule 23 (establishing the
class action procedure), have little utility in the Canadian analysis as these
provisions have no equivalent here).
See e.g. Cassels & Jones, “Rethinking Ends and Means in Mass Tort”,
supra note 54.
2012 ONSC 6173 [Ramdath 2012].
SO 2002, c 30, Schedule A.
Ramdath 2012, supra note 76 at paras 94-95 [emphasis added].

(2016) 2(1) CJCCL

333

On appeal, the College challenged the finding of the trial judge that
the college owed a duty of care to its students, because the tort required
“reasonable reliance”. The British Columbia Court of Appeal appeared to
treat the question, central to liability, as an aggregate, generic issue, rather
than an individualistic one, holding:
[t]he appellant’s concerns are primarily directed at whether each of the class
members reasonably relied on the representations and can prove damages. The issue
of damages was not certified as a common issue and will be determined, with
evidence, at the individual issues phase of the trial.79

So, without confronting the matter directly, the Court of Appeal appeared
to believe that the question of individual reliance, an element of liability,
could be dealt with simultaneously with the question of financial loss,
thereby blending the concept of liability and damages.
This “blending” appeared complete when the question was returned
to the trial judge for the damages phase of the hearing. By that time,
counsel had agreed to proceed solely on the Consumer Protection Act
remedy for “damages”. The main controversy was whether the legislative
remedy still required, like negligence, a causal link between the wrong
and loss. The trial judge had concluded in the prior hearing that it did
not, but the matter was reargued before him again:
GBC, however, argues that even if reliance is not required to establish an unfair
trade practice under the Act, or to rescind the consumer agreement and get
a refund of monies paid, some measure of causation must still be shown if
the consumer is claiming “damages”. The entitlement to claim damages, says
GBC, does not vitiate the need to prove causation. There has to be at the very
least some evidence of a causal connection or nexus between the unfair practice
and the damages being claimed. And this nexus can only be determined, argues
GBC, on an individual, i.e. not aggregate, basis. This issue - whether or not
the s. 18(2) damages remedy requires proof of a causal connection - dominated
both the written and oral submissions. The issue has not been addressed in
the case law and is not self-evident. Fortunately, I do not have to decide the
matter. I am satisfied on the uncontroverted findings that have already been
made in this litigation that a sufficient causal connection (for the purposes of
the s. 18(2) damages remedy) has been established. I refer in particular to the
following findings in the Common Issues and Appeal Decisions: (i) It was
the opportunity to complete the three industry designations that attracted
the plaintiffs to GBC in the first place, not the GBC certificate. None of
79.

Ramdath v George Brown College of Applied Arts and Technology, 2013
ONCA 468 at para 8 [emphasis added].

334

Jones, Panacea or Pandemic

them wanted or needed another college graduate certificate. (ii) The plaintiffs
claim they would not have enrolled in the Program but for the representation
about the industry designations. For each of them, and for the students they
represent, the value of the Program was the promised opportunity to complete
the requirements for the CITP, the CCS and the CIFF “in addition to” the
GBC graduate certificate. (iii) The promise of these industry designations made
the program very attractive to prospective students. I also rely on the common
sense observation that students applying for an eight-month college program
(especially those that are coming from foreign lands) will most likely review
the Program description before applying and paying a substantial tuition. In
short, I have no difficulty concluding that if the s. 18(2) damages remedy requires
some nexus or causal connection with the unfair practice, this has been sufficiently
established.80

So let us be clear on what is happening here. The trial judge found
that the requirement of individual reliance in all members of the class
(if indeed it was required), had been satisfied, not through individual
evidence, but rather on (i) the representative plaintiffs’ own pleadings
and evidence; and (ii) judge’s class-wide inference, based on “common
sense” and the circumstances of the misrepresentation, that all the class
members had probably relied. He then moved quickly to an assessment
of aggregate damages.
Ramdath was followed by Justice McEwen in Trillium Motor World
Ltd v General Motors of Canada Limited 81 (“Trillium”), a case turning
on “loss of chance”. There, the court decided it could award class wide
damages to car dealers who had lost an opportunity to negotiate due to
the default of the defendant. But would each in fact have negotiated?
Even though the evidence had been specific to the representative plaintiff
alone, the court felt comfortable extrapolating this causation question to
the entire class, and disposing of it simultaneously with the question of
class-wide, aggregate damages:
[t]he third precondition, s. 24(1)(c), is the most critical. In this regard, I
must determine the aggregate or a part of the defendant’s liability to the Class
Members and give judgment accordingly where the aggregate or part of the
defendant’s liability to some or all of the Class Members can reasonably be
determined without proof by individual Class Members. Justice Belobaba
80.
81.

Ramdath v George Brown College, 2014 ONSC 3066 at paras 17-19
[Ramdath 2014] [emphasis added, italics in original].
2015 ONSC 3824.

(2016) 2(1) CJCCL

335

recently analyzed this precondition in Ramdath v George Brown College, 2014
ONSC 3066 (CanLII), 375 D.L.R (4th) 488 at para 47, identifying three
requirements: (a) the reliability of the non-individualized evidence that is being
presented; (b) whether the use of this evidence will result in any unfairness or
injustice to the defendant (for example, by overstating the defendant’s liability);
and, (c) whether the denial of an aggregate approach will result in “a wrong
eluding an effective remedy” and thus a denial of access to justice. In my view,
Trillium has satisfied all three requirements. The basis for Trillium’s claim in
aggregate damages is loss of chance. This chance relates to the affected dealers
as a group, and the likelihood that negotiations of the terms of the WDA
would have taken place between the group as a whole and GMCL. The nonindividualized evidence is reliable, the use of the evidence does not result in any
unfairness to Cassels, and to deny the Class Members the aggregate approach
would amount to the denial of a remedy. Acting collectively in negotiations
with GMCL is a critical component of the Class Members’ claim against
Cassels. An individualized approach to damages would not only be unfair
to the individuals who would have banded together, it would be misguided
given the nature of their action. Determining how much more money would
have been available from GMCL for the Class Members had they had an
opportunity to negotiate for it does not cause any injustice to the defendant
Cassels by overstating its liability; rather, it simply quantifies that liability.82

In other words, where it can be inferred that the wrong has had generic
consequences across the class, and where there is no injustice to either
plaintiff or defendant, the requirements of section 24(1) of the Class
Proceedings Act of Ontario have been met, and group wide liability in
the sense of causation can be assessed simultaneously with “monetary
liability”.83 Section 24(1) of the Ontario Act, therefore, appears to be
operating not as simply a procedural device, but at least as a framework

82.
83.

Ibid at paras 540-41 [emphasis in original].
Ont Class Proceedings Act, supra note 50 provides that aggregate damages
may be awarded where: “(a) monetary relief is claimed on behalf of some
or all class members; (b) no questions of fact or law other than those
relating to the assessment of monetary relief remain to be determined in
order to establish the amount of the defendant’s monetary liability; and
(c) the aggregate or a part of the defendant’s liability to some or all class
members can reasonably be determined without proof by individual class
members”: s 24(1).

336

Jones, Panacea or Pandemic

within which courts may modify or extend the substantive law of tort.84
It is a very short step from these cases finding, without individual
evidence, causation in the entire class, to a court finding, on a similar
basis, that causation was made out in a portion of the population. Once
the decision is made to consider the question on an aggregate basis, then
argument can be introduced on the extent of the harm throughout the
class. So in Trillium, for instance, if the defendant had produced evidence
that some class members would not have negotiated and therefore suffered
no loss, that need not send the matter for individual adjudication of each
class members claim. The judge could still fairly assess aggregate liability
if the harm could be equitably assessed proportionately through expert
evidence, sampling or other devices (assuming also distribution concerns
could be assessed in the aggregate). This would still satisfy the three-part
test in Ramdath, and further the access to justice goals identified in that
case. Justice Belobaba in Ramdath had introduced his damages judgment
with the following paragraph:
[a]ggregate damages are essential to the continuing viability of the class action.
If all or part of the defendant’s monetary liability to class members can be fairly
and reasonably determined without proof by individual class members, then
class action judges should do so routinely and without hesitation. Aggregate
damage awards should be more the norm, than the exception. Otherwise, the
potential of the class action for enhancing access to justice will not be realized.85

The same points can, and should, be made with respect to liability. If
class-wide causation can be fairly and reasonably determined without
individual proof, then class action judges should do so routinely and
without hesitation. And we appear to be cautiously starting down that
road. This is how it should be: recognizing that class proceedings Acts
did not, in themselves, modify the substantive law does not mean that
they froze the substantive law at the time of their enactment. The law
of causation can adjust to the procedural context of a class action, and
should do so when it so obviously improves the efficacy of the tort system,
84.

85.

That these cases were decided recently is significant, as they appear to be,
at least to some extent, at odds with Rothstein J’s comments in Microsoft
SCC, supra note 3 at para 131 to the effect that class-wide liability must
be established before the aggregate damages provisions could be applied.
Ramdath 2014, supra note 80 at para 1.

(2016) 2(1) CJCCL

337

improves access to justice, secures compensation and effects appropriate
deterrence.
Finally, I cannot resist pointing out that aggregate assessment of
liability already underpins the judicial resolution of mass wrongs in
Canada. Most certified cases settle before trial, and the settlement terms
are subject to judicial oversight and approval. These settlements routinely
rely on estimates of the global liability of defendants to “groups of persons”
without proof of individual loss.86 If one is “substantively, procedurally,
institutionally, or circumstantially fair”, as class action settlements must
be,87 then how can the other (the use of the same devices to determine a
fair outcome at trial) not be?
Aggregate assessment of causation is a solution to the same problem
that has led to reliance on waiver of tort. But because the aggregate award
is actually premised on the true harm caused by the defendant’s wrong
(rather than just by the extent to which it has profited from it), the
assessment of liability and quantification of the damages on an aggregate
basis is far preferable whether viewing the question from a perspective
of adequate compensation or optimal deterrence. This was the central
insight of Riordon J in Blais and Létourneau, and the logic underpinning
Ramdath and Trillium.

V.

Conclusion

I have endeavoured to demonstrate, throughout this article, that in most
class actions with problems of indeterminate causation, tort law is a
better avenue of address than equitable waiver of tort.
Courts are beginning to recognize the efficiency of adapting the
86.

87.

See for instance the decision approving the settlement in Pro-Sys
Consultants Ltd v Infineon Technologies AG, 2014 BCSC 1936 (where
Masuhara J approved a settlement which incorporated principles of
aggregate causation and market share liability). I would argue that many,
if not most, class actions that settle before trial are “rough and ready”
estimates of class-wide liability that are premised on disconnecting the
causal link between wrongdoer and the individual victim.
See John C Kleefeld, “Facets of Fairness: Kidd v Canada Life Assurance
Company and the Approval of Class Action Settlements” (2015) 10:1&2
Canadian Class Action Review 33 at 64.

338

Jones, Panacea or Pandemic

substantive rules of tort to further the objectives of compensation and
deterrence, so that class action procedures can fulfill their promise as one
of the principal avenues for access to justice in large-scale claims. The
courts’ willingness to entertain waiver of tort in collective litigation is
a step on the way, and indicates that judges are alive to the problems of
indeterminate causation in class claims.
But the recognition of waiver of tort would represent only a partial
solution, one that might yield poor results and risk backfiring on the
victims of tort, undercompensating them and, through under-deterrence,
ensuring that more victims will be created in the future.
Conversely, a narrow, particularistic application of tort causation rules
is little better. It is possible for courts to use presently recognized devices,
such as “robust inferences of causation” or reversed onuses, to overcome
individual attribution issues even in class claims. On one reading, this
is what the Ontario courts were doing in Ramdath and Trillium, and
it also could be viewed as operating in Blais and Létourneau. But these
devices, which are artificial in individual claims, are even more plainly
so in a class claim, where defendants are inferred or presumed to have
caused a magnitude of harm that we know they did not cause. This could
be heavy-handed, even absurd, in many toxic torts and many products
liability cases: where a defendant’s wrong has caused a measurable, but
still incremental, increase in the incidence of a particular disease or injury,
the “robust inference” or reversed onus might mean it would be on the
hook for all such injuries.
The “better way” is to continue down the road blazed by the Quebec
and Ontario courts, and embraced by legislatures in provincial tobacco
legislation: viewing classable claims as cases of harm in populations,
rather than in individuals. True aggregate estimation of the harm, with
the plaintiff retaining the burden of proof, can be both more accurate
and fair, and, where it is, it should be employed.