Canadian Class Action Law: A Flawed Model for European Class Actions

Activists urging European Union nations to adopt the class action device have recently begun citing Canada as a model. Like the United States, Canada has adopted formal class action rules that permit plaintiffs to bring class proceedings. And there is a perception that, to date, Canada has been spared the sort of rampant, U.S.-style class action litigation that has been widely criticized for imposing "huge, avoidable, and unnecessary cost[s]" on the economy.1 Some European class action advocates have therefore suggested that Canada provides a guide to creating a class action regime without opening a "Pandora's box" leading to the American experience.2

On closer inspection, however, Canadian class action law turns out not to be the ideal model it is sometimes claimed to be. European policy makers should think twice before importing it. Perhaps the most striking feature of Canadian class action law is its exceptionally permissive standard for class certification, which makes it markedly easier to certify a class in Canada than in the United States. This feature of the Canadian system stems not simply from permissiveness among Canadian judges or any other similarly contingent cause but rather from a core structural feature of Canadian class action law itself: the lack of a formal "predominance" requirement prohibiting class certification unless common issues predominate over individualized issues in a proposed class.

The permissiveness of Canadian class certification procedures makes Canada a dangerous model for Europe, because it opens the door to the kind of class action abuse found in the United States. As experience in the U.S. and Canada shows, class certification is the key battleground in class action litigation: although class action statutes typically contemplate that certification will only be the first step in a class proceeding, the reality is that in nearly all instances, it is the last step as well. This is so because once a class is certified the risk of trial loss to a company's financial stability can be so great, even as to extremely weak claims, that corporations have no choice but to settle. Permissive class certification rules thus create a powerful and dangerous incentive that encourages plaintiffs to file weak or even frivolous claims, simply to take advantage of their potential settlement value. This value is enhanced—and the incentive to file such suits heightened—because permissive class certification rules also lead to inherently unfair and abusive class procedures, such as "perfect plaintiff" classes, "common issues" trials, and "trials by statistics," that (as discussed below) tilt class proceedings sharply in plaintiffs' favor.

These kinds of abusive procedures have already appeared in Canada, apparently as a direct result of its permissive certification rules. However, to date, other features of the Canadian system, such as the "loser pays" rule and the implicit bar on contingency fee arrangements, appear to have kept the volume of class action litigation below that of the U.S. But class actions in Canada are a relatively recent innovation, and the full character of the Canadian system probably has not fully emerged—in the United States, for example, pervasive class action abuse took decades to develop fully. European policymakers should therefore not simply assume that procedural rules like fee-shifting authorizations or contingency fees prohibitions will hold the line against class action abuse in the long run.

Notably, U.S. plaintiffs' lawyers have recently begun to use Canadian class actions as an integral part of cross-border litigation strategies, filing parallel class actions in the U.S. and Canada, attempting to use U.S. discovery to support Canadian claims, and generally taking advantage of efficiencies of scale to decrease the cost of bringing class actions in foreign countries. The ultimate impact of these kinds of strategies on Canadian class actions remains to be seen. But given this uncertainty, it would be foolish to assume that Canada is somehow immune from the kind of class action abuse found in the United States, simply because such abuse has not yet fully blossomed.

Instead of relying on this kind of rudimentary (and misleading) empirical correlation, European policymakers should look more closely at the substance of Canadian class action law itself. That law is characterized by liberal certification rules that create a powerful incentive to file class action claims—regardless of merit. It may be that, in some instances, this incentive can be counterbalanced (at least temporarily) by other features of a legal system. But rather than attempting to adjust these conflicting incentives to achieve a balance that avoids U.S.-style abuse, European policymakers would be far better served by rejecting the Canadian model and declining to import this flawed structure in the first place.

I. History of Class Actions in Canada

The permissiveness of Canadian class certification procedures is no accident. The history of Canadian class actions evidences a clear effort to create a procedure that is more amenable to certification than the U.S. model. Class actions (or "class proceedings") are a relatively recent innovation in Canada. Although Quebec became the first Canadian province to adopt class action legislation in 1978,3 the move toward widespread class action legislation in Canada did not begin until 1982, when the Ontario Law Reform Commission issued a Report on Class Actions. That report set forth three objectives for future lawmakers to consider—(improving judicial efficiency, affording greater access to justice for individuals, and achieving behavior modification among manufacturers)—and recommended that, to meet these objectives, provinces adopt a class certification procedure modeled after U.S. Federal Rule of Civil Procedure 23.4 The Commission's recommendation, however, departed from U.S. law in one crucial respect: rather than a separate requirement that common questions predominate over those that are specific to individual members of the class, the report recommended that courts consider predominance as just "one of the factors employed to gauge whether the class action is superior."5

The report had little impact at first, but in the mid- 1990s two provinces passed significant legislation aimed at accomplishing the Reform Commission's missions: the Ontario Class Proceedings Act in 1992, and the British Columbia Class Proceeding Act of 1995.6 Both laws "draw heavily on the experience under Rule 23 of the U.S. Federal Rules of Civil Procedure," but, like the Commission's report, are in certain respects substantially more permissive of class certification than the U.S. rule.7 This was deliberate: the drafters of the Ontario Act "expressly adopted a regime thought to be more hospitable to the certification of class actions than had previously been exhibited for many years in the United States."8 Under both the Ontario and the British Columbia statutes, courts "shall" or "must" grant certification of a class if seven conditions are met:

1. The pleadings invoke a cause of action;9

2. there is an identifiable class of two or more persons;

3. the claims of the class members raise common issues;

4. the class proceeding is the "preferable procedure" for resolving the common issues;

5. the class representative fairly and adequately represents the interests of the class;

6. the class representative produces a plan for the proceeding; and

7. the class representative does not have a conflict with the interests of the class members.10

These laws have set the legal standard under which most class actions in Canada are litigated. Ontario and British Columbia are two of the most populated Canadian provinces; further, courts in those jurisdictions are willing to approve national class actions, certifying classes even when some (or even most) members of the class reside in other provinces.11 In addition, with the exception of Quebec—whose certification laws are even more liberal—other Canadian provinces and territories that have enacted comprehensive class action legislation have followed the lead of the Ontario and British Columbian legislatures and adopted substantially similar certification requirements.12

Today, Canadians in every province have the option of bringing class actions, even if the province in which they reside has not passed class action legislation. In 2001, the Canadian Supreme Court decided Western Canadian Shopping Centres, Inc. v. Dutton, which extended class actions to all Canadian jurisdictions, holding that "[a]bsent comprehensive legislation, the courts must fill the void [using] their inherent power."13 After surveying the class action legislation enacted throughout the country, the Court pointed to the Ontario and British Columbia statutes as models that should "provide guidance" to courts throughout Canada. It then distilled four minimum criteria for the certification of a class:

1. The class must be "capable of clear definition;"

2. there "must be issues of fact or law common to all class members;"

3. there cannot be conflicting interests among class members—victory for one must mean victory for all; and

4. the class representative must adequately represent the class.14

The Court also articulated several discretionary factors that could be considered in a class certification proceeding, including whether "[c]lass members... raise important issues not shared by all members of the class."15 Accordingly, in these provinces as well, predominance is at best a discretionary factor, rather than a required element for class certification.

II. Canada's Liberal Class Certification Test

The enactment of class proceedings legislation in Canada has, unsurprisingly, led Canadian courts to grant certification "in many cases that would not traditionally have qualified as representative actions" under the common law.16 More significantly—particularly for those who would look to Canada as a model of class-action restraint—the enactment of class action legislation in Canada has produced a system that is even more liberal in its certification policy than the U.S. class action regime.17

In the United States, Rule 23 of the Federal Rules of Civil Procedure provides that a class may not be certified unless "[common] questions of law or fact... predominate over any questions affecting only individual members."18 As a result, in U.S. class action proceedings, federal and state courts "devote a great deal of attention to whether there is sufficient identity of interest or cohesiveness among the class members" to warrant certification, "particularly when separate transactions, acts, or omissions are involved."19 For example, putative class actions in the U.S. are generally "doomed to fail" if part of the proposed lawsuit involves evaluating injuries and damages on an individual basis.20 The rationale behind the predominance requirement is, as the U.S. Supreme Court has explained, to ensure that "proposed classes are sufficiently cohesive to warrant adjudication by representation"21—i.e., to ensure that class action litigation does not devolve into multiple trials of issues and facts relevant to only some of the class members.22

In practical terms, the predominance requirement serves an important screening function: it is "an attempt to achieve a balance between the value of allowing separate actions to be instituted so that individuals can protect their own interests and the economy that can be achieved by allowing a multiple party dispute to be resolved on a class-action basis."23 Put differently, the predominance requirement exists to ensure that the class action device is not unmoored from its fundamental purpose—namely, "to facilitate the adjudication of disputes involving common questions and multiple parties in a single action."24

In Canada, "the threshold for class certification" is "generally considered to be lower than in the United States," mainly because there is no formal predominance requirement.25 Legislatures in Saskatchewan, British Columbia, Alberta, Newfoundland, and Manitoba have purposely and expressly disclaimed this requirement in the text of their laws, directing courts to certify classes if the class members raise common issues, regardless of "whether... those common issues predominate over issues affecting only individual members."26 Laws in other provinces (e.g., Ontario) are silent about the predominance requirement, but at the direction of the Supreme Court, courts there, as elsewhere, have consistently required only that class members share "a substantial common ingredient;" they have not found it "necessary that common issues predominate over non-common" ones.27 Laws in one province—Quebec—have an even lower commonality requirement than the rest of Canada,28 permitting certification if the class members' claims raise "identical, similar, or related questions of law or fact."29

The permissiveness of class certification procedures in Canada is (at least in some instances) exacerbated by other procedural and evidentiary rules that make it even easier for plaintiffs to certify classes. In Quebec, for example, a decision certifying a class cannot be appealed by the defendant, but if a class is not certified, the would-be plaintiffis allowed to appeal.30 Moreover, "recent changes to Quebec's procedural rules make it very difficult for defendants to challenge the veracity of the plaintiffs' factual submissions at the certification hearing."31 It is actually a matter of the court's discretion whether a defendant has a right to present evidence at the certification stage at all.32

These permissive standards have led Canadian courts to certify classes that would almost certainly have been rejected in the United States. One court in Quebec, for example, authorized a class proceeding for 2,400 members of a community who had suffered damage to their homes due to industrial air pollution.33 The court acknowledged vast differences among the class members: only some owned their homes (others were tenants), only some were long-term residents, and only some had houses constructed with material vulnerable to damage from the relevant form of pollution.34 In fact, the court observed that it was "doubtless" that "the damages and inconvenience caused by the air pollution varied from house to house, depending on the nature and value of the house and the distance from the port installations and its location with reference to the prevailing wind."35 But despite these differences among class members, the court certified the class anyway, concluding that the law did not require that all—or even the majority—of the questions in the case be related.36 It was enough, the court held, "if the claims of the members raise some questions of law or fact that are sufficiently similar or sufficiently related to justify a class action."37

This kind of liberal authorization of class actions has been widespread. In Boulanger v. Johnson & Johnson, in the aftermath of alleged cardiac reactions connected to Propulsid (a drug designed to treat stomach disorders), a Canadian court in Ontario certified a class of all persons in Canada outside of Quebec who ingested the drug.38 The court's opinion acknowledged that "not all class members stand to recover damages at the same level," and that "some class members may not be able to demonstrate that they have sustained injuries and losses" at all.39 Nevertheless, the court continued, Canadian class action law imposes only a "low bar, and the fact that many individual issues may remain after resolution of the common issue... is not a bar to certification."40

To be sure, Canadian courts do give limited consideration to whether a proposed class action turns on common issues as opposed to individual ones. They address this factor, however, only when evaluating the catch-all test in their certification process—whether the class action is the "preferable procedure" to resolve the dispute.41 Statutes in some provinces explicitly list predominance as one of the "preferable procedure" factors;42 courts in other jurisdictions have considered it without explicit legislative direction to do so. Consideration of predominance in this way, however, has generally been treated as a matter of discretion that lies solely with the judge: if individual issues are present in a lawsuit, these do not bar the certification of the class.43 Thus, to the extent courts apply a predominance restriction in Canada at all, it is a diluted version of its U.S. counterpart.

The upshot of these structural features of Canada's class action law is that a large proportion of proposed classes in Canada are certified. There is a general scholarly consensus that Canadian class certification is more liberal than that in the United States.44 For example, in the medical products liability field, there was "no successful opposition to certification... in Ontario [from] 1994" to at least 2003.45

III. Why it Matters:

The Consequences of A Liberal Certification Rule Canada's low bar for class certification, driven by the conspicuous absence of a formal predominance requirement, has important practical consequences that should alarm European policymakers contemplating Canada as a model for class action form. Any class action system has the potential to increase the pressure on defendants to settle cases, simply because a verdict in a class action will result in liability to numerous class members, rather than to a single plaintiff. This pressure is increased by liberal certification rules. That is true in part for the straightforward reason that liberal rules make it easier for class plaintiffs to subject defendants to this class-based pressure to settle. But it is also true because liberal certification rules can lead to a series of class procedures that are tilted severely in plaintiffs' favor. For example, absent a strict requirement that the named plaintiffs prove the elements of each individual class member's claim, courts may authorize "perfect plaintiff" classes, "common" issues trials, and "trial by statistics." These procedures prevent defendants from defending themselves fully. The use of such procedures—which skew the class action process in favor of plaintiffs—forces defendants to choose between two unpalatable options: either face an unfair trial they are likely to lose, or settle plaintiffs' claims regardless of their merit.

A. The Prejudicial Procedures Resulting From Liberal Certification Rules

The loose class certification requirements currently applied in Canada make it possible to certify a broad class that includes plaintiffs who are more different than they are alike, making a fair trial of all class members' claims virtually impossible. Because such classes involve dissimilar plaintiffs alleging highly individualized claims, it is not possible for a court to determine whether each class member can establish the elements of his or her claims based on common evidence alone. Thus, courts must instead determine a class action defendant's liability to the class "generally." This often occurs in one of three different ways: (1) the determination of liability based on a "perfect plaintiff's" claims; (2) the certification of "issues" classes; or (3) the use of "trial by statistics." These generalized methods of adjudicating class claims essentially stack the deck in plaintiffs' favor, allowing plaintiffs to prevail without ever having to demonstrate that individual class members can establish the elements of their claims through common proof. As a result, defendants are often forced to settle claims regardless of merit, due to the risk inherent in trying claims on a class-wide basis where plaintiffs are not required to prove key elements of their claims.

1. Defending Against a Fictional, Composite Plaintiff

In a class action, representative plaintiffs are appointed to litigate their claims on behalf of the other members of the proposed class: if the representative plaintiff's claims succeed, so do the claims of every class member. As a result, class actions function correctly only if the named plaintiffis so similarly situated to the class members that a finding in the representative's case applies fairly and equally to every other class member's case as well.

If, however, a court certifies a class in which the plaintiffs' claims vary, plaintiffs can "stitch together" the strongest portions of their various cases—for example, highlighting evidence that the defendant made serious misrepresentations to one plaintiff, that a second plaintiffrelied on defendants' statements, and that a third plaintiffsuffered damages. An action based on such "a fictional composite" is often "much stronger than any plaintiff's individual actions would be."46 Some courts have referred to this as the "perfect plaintiff" problem,47 because plaintiffs enjoy "the practical advantage of being able to litigate not on behalf of themselves but on behalf of a ‘perfect plaintiff' pieced together for litigation." 48

Allowing class members to prevail based on the fictional case of a "perfect plaintiff" is problematic for several reasons. First, it raises serious fairness concerns, as it allows the fact- finder to ignore the weaknesses in one plaintiff's case in light of the strength of another plaintiff's case. As such, it forces the defendant to defend against a much stronger collection of claims than it would have to face absent class certification. But, perhaps more importantly, it often leads to class recovery in cases where the individual members of the class would be unable to establish the elements of their claims standing alone. If two individual plaintiffs were alleging separate tort claims against a single defendant, each would have to prove that the defendant owed them a duty, breached that duty, and as a result caused them damages. If each individual plaintiffcould not prove both elements—if, for example, one plaintiffcould not prove that the defendant owed her a duty and the other plaintiffcould not prove that she suffered injury as a result of defendant's conduct—neither could prevail. However, in a "perfect plaintiff" scenario, those two plaintiffs can effectively join forces, with one plaintiffproviding evidence of a duty and the other proving causation and damages. This kind of patently unfair result is exactly what occurs in class actions involving plaintiffs alleging individualized claims, as class members with weak claims are allowed to combine to present an airtight case against the defendant.

2. "Common Issues" Trials

Some Canadian courts have attempted to avoid the problems inherent in class trials of dissimilar claims by certifying classes for the limited purpose of addressing one or more "common issues" that the court has determined apply to each class member's claims, and then leaving the individualized aspects of the claims to be resolved in subsequent individual trials.

For example, in Hewerd v. Eli Lilly & Co., the court certified an issues class brought by users of the anti-psychotic drug, Zyprexa, who claimed personal injury as a result of using the drug.49 In that case, the court determined that seven "common" questions would be answered as to the entire class based on a single trial, including whether "defendants knowingly, recklessly, or negligently breach[ed] a duty to warn or materially misrepresent[ed] any of the risks of harm from Zyprexa" and whether "class members [are] entitled to special damages for medical costs incurred in the screening, diagnosis and treatment of diseases related to Zyprexa."50 Similarly, in Boulanger v. Johnson & Johnson Corp., the court certified a class of users of the acid reflux drug Propulsid who claimed that the drug caused cardiac injury.51 There, the court determined that issues such as "whether [Propulsid causes] adverse cardiac events" and whether defendants engaged in the "negligent design, manufacture, marketing, and sale of [Propulsid]" were common to the entire class and therefore could be resolved in a single proceeding.52

The idea behind these types of issues classes is to streamline the resolution of common questions applicable to all plaintiffs and to try individual issues separately. In reality, however, it is impossible to resolve these issues in a vacuum, divorced from the facts of each plaintiff's case. For example, in the product liability suits described above, the Canadian courts certified questions such as whether the product at issue could cause the injury plaintiffs allege or whether the defendant generally failed to warn consumers about the product's risks. However, answering these questions in the abstract, without reference to the facts of any particular plaintiff's case, does nothing to advance the class members' cases. For instance, an issues trial on "general causation" asks the court (or jury) to answer the highly prejudicial question of whether plaintiffs have shown that the product is capable of causing harm. This is an undeniably easy standard for plaintiffs to meet since they do not have to prove that the product actually did cause injury to class members—a showing that would require evidence of each class member's medical history, susceptibility to injury, and risk of injury from other sources. Instead, plaintiffs must merely prove that such an injury is generally possible.

The same is true with regard to an issues trial on the question of whether a class action defendant generally "failed to warn" consumers about the risks associated with a product. In such a trial, plaintiffs would only be required to prove that the defendant's statements would likely have misled consumers, not that the statement did mislead each individual consumer—an inquiry which would require highly individualized evidence related to, inter alia, each consumer's exposure to the defendant's statements, the changing nature of the defendant's own knowledge of the risks of its product, and the availability of risk information from other sources.

The supposed purpose of an issues trial is to resolve common issues in one trial so that a second trial on individualized issues can be conducted quickly for each plaintiff. However, because "general" findings as to causation or failure to warn made in an issues trial do not relate to the facts of any individual class member's claim, they offer no real efficiency benefit. Instead, issues trials simply stack the deck against defendants because: (1) the "common issues" are so generalized that they are almost certain to result in a plaintiffverdict; and (2) the second phase of the trial—i.e., the actual trials of individual plaintiffs' claims—is inevitably infected by the verdict from the first phase. Plaintiffs in an issues trial are not required to present any case-specific evidence; they merely need to "establish" general liability—whether on the issue of general causation, failure to warn, or some other allegedly "common" issue. Thus, it is almost impossible for a defendant to prevail on the issues phase of such cases. And even though each class member would have to prove in a subsequent trial that the defendant actually caused his or her injury specifically, that second trial would inevitably be affected by the initial finding that the defendant's actions could have caused harm to the class. In essence, the defendant is forced to defend the individualized second-phase trials with a red mark next to its name.

Moreover, even though it does not result in even a cent of actual liability, an issues trial verdict against a class action defendant can have serious public relations and investment consequences. Even before a single class member has proved that he or she was actually injured as a result of the defendant's conduct, or legally entitled to any relief, plaintiffs are nevertheless able to claim publicly that they have "established" general liability. The resulting media coverage can have a serious negative impact on a class action defendant's business and financial well-being. As a result, defendants are effectively subject to economic punishment despite the fact that no plaintiffhas ever proved his or her claim.

3. Class Trial by "Statistics"

Permissive class certification requirements that allow certification of dissimilar, highly individualized claims can also lead to the use of generalized "statistics" as proof of plaintiffs' claims. Simply put, if class members' claims vary, it may not be possible to resolve all of their claims based on the facts of a representative plaintiff's case. As a result, class action plaintiffs often attempt to prove highly individualized elements of their claims—such as causation and damages—on a class-wide basis using over-generalized statistical evidence that fails to address the merits of any one class member's claims.

The problems inherent in such a "trial by statistics" are obvious. Like a "common" issues trial, a trial based on statistical evidence and expert testimony does not establish the defendant's liability as to individual class members. Instead, statistical evidence merely indicates that it is probable that the defendant's actions caused harm to the group as a whole. Thus, assessing liability based solely on this type of generalized proof essentially allows individual class members to recover without ever having to prove the basic elements of their claims. As a result, defendants are forced to compensate an entire group of plaintiffs without any one of those plaintiffs having to prove that he or she was actually injured or that his or her injury occurred as a result of the defendant's conduct.

In addition, assessing a defendant's liability as to an entire class based on statistical evidence robs the class action defendant of the ability to adequately defend itself against plaintiffs' claims. For example, in a product liability failure-towarn case, the defendant will likely have a variety of defenses against an individual plaintiff's claims based on the individual circumstances of that plaintiff's exposure to warnings, history of heeding warnings, and knowledge of the alleged risks from other sources. Some plaintiffs will have a history of ignoring warnings and using dangerous products. Other plaintiffs may have heard from another source about the risks that the defendant allegedly withheld. Thus, those plaintiffs will have a much weaker failure-to-warn case than a plaintiffwho had no knowledge of the alleged risks and is adamant that she would never have used the product if she had been aware of the risk. A trial by statistics makes no allowance for such distinctions—even though common sense suggests that a more or less random sample of consumers, with highly varied habits and histories, will have very different claims. As a result, even though some plaintiffs' cases may be fatally flawed (including plaintiffs who cannot establish all of the elements of their cause of action or whose claims are susceptible to individualized defenses, such as the statute of limitations), those flaws will never be uncovered during a "trial by statistics."

In short, a trial by statistics is more akin to a theoretical "trial by average" where it is determined that the defendant's actions would probably have harmed the "average consumer," but never established that the defendant actually harmed any real consumer. Like a trial of putatively "common" issues, it is an example of "rough justice" at its most stark: an acceptance of prejudice to defendants and unproven recovery to plaintiffs, simply to achieve the expediency of disposing of multiple claims at once.

B. The Pressure to Settle

The risks to a defendant from "perfect plaintiff" trials, "issues" trials and "trials by statistics" all have the same effect: they exacerbate the already-existing pressure on defendants to settle class actions regardless of the merits of the plaintiffs' claims. In any class action system, there is a risk that opportunistic plaintiffs, with an eye towards settlement, can seek to identify potential classes, knowing that once the potential liability for a claim is multiplied by the number of class members, the defendant will face tremendous pressure to settle, regardless of the merits of the underlying claims.53

As noted earlier, the mere fact that a defendant faces a class, rather than an individual plaintiff, dramatically increase the downside risk of proceeding to trial. The bulk of this pressure to settle, as one American court recently explained, is felt at the moment a class action is certified: "class certification may be the backbreaking decision that places ‘insurmountable pressure' on a defendant to settle, even where the defendant has a good chance of succeeding on the merits."54 Thus, in the U.S., the class certification proceeding is the "major battleground" of the litigation.55 Once a class action has been certified, in the eyes of many defendants, the game is over.

The weight of settlement pressure is magnified where class certification standards are relaxed. In Canada, for example, it is relatively rare for class action lawsuits to proceed all the way to a trial. By one commentator's count, only two class action cases in Canada were tried between 1993 and 2001, and, although some class actions were resolved by summary judgment, most concluded in settlement.56 This tendency toward settlement is likely linked not simply to the permissiveness of Canadian class certification requirements but also (and more specifically) to Canadian courts' general acceptance of "issues classes." Put simply, once a class is certified, class action defendants are faced with a lose-lose proposition: either spend substantial sums litigating a case that begins with a series of "common" issues trials they are destined to lose, or agree to pay out substantial settlement costs to class members whose claims have never been tested and are likely without merit.

For this reason, the "relatively undemanding" criteria for class certification in Canada are a cause for deep concern among potential class action defendants in that country—and should likewise concern policymakers tempted to use Canada as a model for class action legislation.57 If classes are even easier to certify in Canada than in the United States, plaintiffs' lawyers will have an even greater incentive to file unmeritorious claims, simply to exploit their potential settlement value. This can, in turn, cause a variety of disruptive effects common to American class action litigation, such as causing companies to minimize or avoid investment in particular jurisdictions, or causing them to divert resources from innovation and other productive activities to the defense of class action litigation.

IV. Avoiding Unintended Consequences: Learning Lessons from American Mistakes

Even if Canada has, to date, avoided some of the excesses (particularly the volume of litigation) that characterize the U.S. class action regime, the structure of the Canadian system, with its lack of a predominance requirement and permissive class certification requirements, render it even more inherently prone to abuse than the U.S. system. Indeed, as discussed, the kinds of inherently unfair procedures that generally accompany liberal certification rules (such as "common" issue trials) are already visible in Canada. The fact that Canada has avoided all forms of U.S.-style abuse to date does not mean it will remain immune from such problems.

In this regard, it is instructive to recall that the kind of class action abuse that permeates the U.S. system did not appear overnight. In 1966, when U.S. rule-makers amended Federal Rule of Civil Procedure 23 to alter certain features of U.S. class action law, they "apparently believed that they were simply making rule 23 a more effective procedural tool."58 According to its accompanying note, the 1966 amendment, which reformulated the text of the rule, was intended:

(1) to redefine the cases that could proceed under rule 23, by adopting more functional definitions of class actions, (2) to clarify the effect of a class action judgment on members of the class, (3) to codify some of the better class action practices that federal judges had developed, (4) to provide district court judges more guidance regarding their procedural powers and responsibilities, and (5) to deal explicitly with the notice that should be provided to absent class members.59

This seemingly benign adjustment to the law's structure was not driven by "revolutionary notions," but rather by far more modest and benign intentions: to create a more efficient legal "mechanism for securing private remedies."60 As U.S. civil procedure scholar Arthur Miller has explained, "the class action onslaught caught everyone, including the draftsmen [of Rule 23] by surprise."61

The development of rampant class action abuse resulted in large part from the unforeseen interaction between these apparently modest changes in class action law and other developments in U.S. law that were still under way when Rule 23 was revised. Specifically, in the second half of the twentieth century, Congress and the American federal courts began recognizing new substantive rights in the areas of (inter alia) civil rights, antitrust, securities litigation, and consumer protection laws.62 American product liability law, for instance, has undergone massive changes since the 1960s, evolving from "a rule that held that manufacturers were not liable unless they had been negligent in the manufacture of their products to a (mostly judge-imposed) rule that manufacturers were liable (even if there had been no negligence) if products left the factory in an ‘unreasonably dangerous' condition."63 Though these changes were not initially enacted with the class action device in mind, entrepreneurial American class action lawyers were quick to see the advantageous marriage of the class action device and these broad new substantive rights.

Meanwhile, as the potential uses for the American class action device began to grow, there were no safeguards in place to prevent abusive litigation. Lawyers who bring class actions in the United States provide their services in exchange for an (often large) percentage of the ultimate settlement or award. This contingency-fee arrangement, in combination with the lack of a "loser pays" rule in the U.S., makes it relatively easy and inexpensive for U.S. plaintiffs to institute class actions. Moreover, plaintiffs' lawyers have ample incentives to bring such suits almost regardless of their merits, due to the potentially enormous awards (which can include a large punitive damage component) and the fact that the percentage devoted to the lawyer's fee is typically quite high.64 Further adding to the potential for abuse is the fact that liberal American discovery rules impose extraordinarily high costs on defendants from the start of litigation, a fact that plaintiffs can leverage into an early settlement.65

In short, although Rule 23 was crafted with modest intentions, unforeseen interactions with evolving substantive law, together with the peculiarities of U.S. procedural rules, led to the widespread abuse that now characterizes the American class action regime.

CONCLUSION

The fact that Canada has not yet been overrun with the same volume of class actions as the United States is no reason to recommend it as a model. Indeed, any limitation on the zeal of class action lawyers in Canada comes not from its liberal class action rules but from other features of Canadian law. Canadian plaintiffs' lawyers are not allowed to take cases on pure contingency arrangements. Instead, courts determine how much compensation a Canadian class action lawyer is entitled to collect by considering the number of hours worked and the size of the ultimate award collected.66 Further, these fees are generally not as large as the ones collected in the U.S., due to caps on pain-and-suffering damages and restrictions on punitive damages.67 Moreover, under Canadian law, when the defendant prevails in a class action, the class representative himself is responsible for the defendant's legal fees. This, too, serves as a disincentive to mounting a frivolous suit.68

But European policymakers should not conclude from this Canadian experience that they can simply import the Canadian system into Europe with impunity, thereby avoiding some of the abuses that have plagued the United States. Canada's class action regime is less than thirty years old and is still developing. As demonstrated above, U.S. class action abuse problems did not appear overnight, and were not anticipated by the framers of the U.S. class action rules. It would be naive to assume that the same harms can be easily and casually avoided by others now. The fact that Canada has thus far managed to avoid some of the excesses of the U.S. system does not mean the structure of Canadian class action law is the secret of its success. To the contrary: as discussed above, the structure of Canadian class action law, and in particular its lack of a formal predominance requirement, has already led to inherently unfair class procedures, and leaves the Canadian regime even more inherently vulnerable to manipulation than the law in the United States.

Moreover, as class actions continue to grow in Canada, political pressure will likely mount to shed some of the safeguards (such as contingency fee restrictions and "loser pays" rules) that until now appear to have protected the country from pervasive class action abuse. Several Canadian provinces, for example, have already adopted strategies to alleviate the harsh consequences of the traditional loser pays rule for the named plaintiff: Quebec has made class representatives liable for only nominal costs; British Columbia excuses class representatives from paying defendants' legal costs unless the failed suit was "frivolous or vexatious;" and Ontario legislation provides that the named plaintiffhas to pay the defendant's costs only if the action was a "test case, raised a novel point of law or involved a matter of public interest."69

The character of Canadian class action practice is also under pressure from the recent increase in cross-border litigation. As other commentators have noted, the U.S. plaintiffs' bar is making aggressive moves into Canada, and has increasingly engaged in concerted cross-border strategies, such as attempting to use U.S. discovery and expert testimony in Canadian courts, or using parallel litigation in the United States and Canada to increase the overall pressure on defendants to engage in a settlement.70 These developments both reduce the costs of litigating class actions in Canada, and sweep Canada into a multi-national dynamic that includes, and is driven by, class action litigation in the United States.

The ultimate effect of these developments is still uncertain, but there is good reason to believe that class action abuses in Canada will continue to mount. Accordingly, European policymakers should look past the simple, unreflective empirical conclusion that Canada has fewer class actions than the United States and instead look more closely at how exactly Canadian class action law is structured. And on such examination, it becomes clear that Canada provides an extremely liberal certification regime, with a firmly embedded structural bias in favor of permissive certification. This bias, in turn, creates both inherently biased class procedures, and a powerful incentive toward bringing class actions—even class actions based on frivolous claims.

Endnotes

1 Christopher Hodges, Multi-Party Actions: A European Approach, 11 Duke J. Comp. & Int'l L. 321, 346-47 (2001). See also Edward Sherman, Group Litigation Under Foreign Legal Systems: Variations and Alternatives to American Class Actions, 52 DePaul L. Rev. 401, 431 (2002) ("Canadian class action settlements have not drawn the same level of criticism as have settlements in the United States."); id. at 402 (noting that Canada has not "experienced the same mushrooming of class action practice as in the United States").

2 Sherman, supra note 2, at 403 ("[M]ost other countries view American class actions as a Pandora's box that they want to avoid opening.").

3 See Quebec Civ. Code, R.S.Q. 1978, ch. C-25, §§ 999-1051; see also Edward Sherman, American Class Actions: Significant Features and Developing Alternatives in Foreign Legal Systems, 215 F.R.D. 130, 153-54 (2003).

4 W.A. Bogart, Questioning Litigation's Role -- Courts and Class Actions in Canada, 62 Ind. L. J. 665, 692 (1987).

5 Id. 6 Class Proceedings Act, S.O., ch. 6 (1992) (Ont.); Class Proceedings Act, R.S.B.C. 1996, ch. 50 (1995) (B.C.).

7 Michael McGowan, Certification of Class Actions in Ontario: A Comparison of Rule 23 of the U.S. Federal Rules of Civil Procedure, 16 Carswell Practice Cases 172 (1993).

8 Paul J. Martin et al., International Legal Development in Review: 2003 Comparative Law – Canadian Law, 38 Int'l Law. 553, 553 (2004).

9 This does not mean, however, that courts will inquire into the merits. Thus, the fact that a cause of action may be barred by a particular defense (a limitations period for example), does not bar certification of the class action. See McGowan, supra note 7, at 172.

10 See R.S.B.C. 1996, c. 50, s.4 (B.C.); S.O. 1992, c.6, s.5 (Ont.).

11 S. Gordon McKee, Why the Development of Mass Torts in Canada is Important to Corporate America, 71 Def. Couns. J. 32, 34 & n.2 (2004) ("Often there will be a national class certified in Ontario, comprised of either all Canadian residents or Canadian residents other than those in British Columbia and Quebec, with companion actions in those provinces. Trial level courts in Ontario and B.C. and an appellate court in B.C. have been willing to certify national classes.").

12 Steven Penney, Mass Torts, Mass Culture: Canadian Mass Tort Law and Hollywood Narrative Film, 30 Queen's L.J. 205, 217 (2004); Alex Kotkas & Casey Smith, Class Actions in Canada: Another Western Province (Alberta) Enacts Legislation, 28 No.1 Class Action Reports 4 (Feb. 2007) (noting that with Alberta's 2004 legislation, there are now six Canadian provinces who have enacted comprehensive class action legislation).

13 [2001] 2 S.C.R. 534 ¶ 34.

14 Id. ¶¶ 38-41.

15 Id. ¶ 42.

16 Penney, supra note 12, at 218.

17 Garry D. Watson, Class Actions: The Canadian Experience, 11 Duke J. Comp. & Int'l L. 269, 272 (2001) ("in certain respects the Canadian legislation is more liberal in facilitating class actions than its American counterpart"). See also Andrew Borrell, The Evolving Evidentiary Standard for Certification in Canada, 26 No. Class Action Reports 3 (2005); Caroline Davidson, Tort au Canadien: A Proposal for Canadian Tort Legislation on Gross Violations of International Human Rights and Humanitarian Law, 38 Vand. J. Transnat'l L. 1403, 1442 (2005) ("[i]n many respects, particularly certification, the Ontario regime is more liberal.") (quoting Michael McGowan, Certification of Class Actions in Ontario: A Comparison of Rule 23 of the U.S. Federal Rules of Civil Procedure, 16 Carswell Practice Cases 172 (1993)).

18 See Fed. R. Civ. P. 23(b)(3).

19 Edward Sherman, American Class Actions: Significant Features and Developing Alternatives in Foreign Legal Systems, 215 F.R.D. 130, 140 (2003).

20 Gary Will & Paul Miller, Association of Trial Lawyers of America, Participation of United States Counsel in Canadian Class Actions and Mass Torts, 1 Ann. 2006 ATLA-CLE 483 (2006).

21 Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 623 (1997).

22 Sherman, supra note 19, at 140-41.

23 7AA Charles Alan Wright et. al, Federal Practice and Procedure §1777 (1998).

24 Id. §1751 (emphasis added).

25 S. Gordon McKee & Martha Cook, Class Actions in Canada: 2005 State of the Union – Growing, Expanding, Developing, 73 Def. Couns. J. 31, 32 (2006).

26 See, e.g., Class Actions Act, S.S. 2001, c.-12.01, §6 (Sask.); Class Proceedings Act, S.B.C., ch. 21 (1995) (B.C.); Class Proceedings Act, S.A. 2003, c. C-16.5, §5 (Alta.); Class Actions Act, S.N. 2001, c. C-18.1, §5 (Nfld.); Class Proceedings Act, S.M. 2002, c. 14, § 4 (Man.).

27 W. Canadian Shopping Ctrs., Inc. v. Dutton, [2001] 2 S.C.R. 534, ¶ 39.

28 Gordon McKee & Robin Linley, The Evolving Landscape for Pharmaceutical Product Litigation in Canada, 73 Def. Couns. J. 242, 243 (2006) ("In Quebec, the threshold for certification is even lower than in the other Canadian provinces.").

29 W. Canadian Shopping Ctrs. [2001] 2 S.C.R. 534, ¶ 37 (citing Quebec Code of Civil Procedure, art. 1003).

30 Jean Saint-Onge & Glenn M. Zakaib, Cross-Border Class Actions—A Canadian Perspective, 744 Practicing Law Institute, Litigation and Administrative Practice Course Handbook Series 87, 93 (2006); John C. Kleefeld, Class Actions as Alternative Dispute Resolutions, 39 Osgoode Hall L.J. 817, 827-28 (2001) ("in Quebec, where a refusal to certify a proceeding as a class action can be appealed, while a decision to certify cannot").

31 McKee & Cook, supra note 26, at 32.

32 Saint-Onge & Zakaib, supra note 30, at 93.

33 Comité d'environement de la Baie, Inc. v. Société d'électrolyse et de chimie Alcan Ltée, [1990] 29 Q.A.C. 251.

34 Id. ¶ 21.

35 Id.

36 Id. ¶ 22 (emphasis in original).

37 Id. (emphasis in original).

38 Boulanger v. Johnson & Johnson Corp., [2007] 40 C.P.C. (6th) 170 (Ont. Sup. Ct.).

39 Id. ¶ 21.

40 Id. ¶ 28.

41 See, e.g., Hollick v. Metro. Toronto, [2001] 13 C.P.C. (5th) 1, ¶ 30 (Sup. Ct.) ("The question of preferability, then, must take into account the importance of the common issues in relation to the claims as a whole... I cannot conclude... that the drafters intended the preferability analysis to take place in a vacuum. There must be a consideration of the common issues in context."); Knight v. Imperial Tobacco Canada Ltd., [2006] 54 B.C.L.R. (4th) 204, ¶ 12 (Ct. App).

42 The Alberta legislation provides such an example. See Class Proceedings Act, S.A. 2003, c. C-16.5 (Alta).

43 Campbell v. Flexwatt Corp., [1997] 44 B.C.L.R. (3d) 343 ¶ 61 (Ct. App.) ("nowhere does the Act mandate that if an individual issue should predominate, an action must not be certified. Instead, the Act sets out a variety of factors to be considered. The existence of an individual issue is not necessarily determinative."); Cloud v. Canada, [2004] 73 O.R. (3d) 401 (Ct. App.).

44 See Watson, supra note 17, at 4 ("in certain respects the Canadian legislation is more liberal in facilitating class actions than its American counterpart."); Davidson, supra note 17, at 1442 ("[i]n many respects, particularly certification, the Ontario regime is more liberal."); McKee & Cook, supra note 25, at 32 ("The threshold for class certification in Canadian provinces is generally considered to be lower than in the United States."); Charles M. Wright & Matthew D. Baer, Price Fixing Class Actions: A Canadian Perspective, 16 Loy. Consumer. L. Rev. 463, 467 (2004) ("Canadian courts may be somewhat more inclined to certify classes").

45 McKee & Cook, supra note 25, at 32.

46 See Broussard v. Meineke Disc. Muffler Shops, Inc., 155 F.3d 331, 344- 45 (4th Cir. 1998).

47 Id.

48 Id.

49 [2007] 39 C.P.C. (6th) 153 (Sup. Ct.)

50 Id. ¶ 77

51 [2007] 40 C.P.C. (6th) 170 (Ont. Sup. Ct.)

52 Id. ¶¶ 28, 34-36.

53 Martin Redish & Adriana Kastaneck, Settlement Class Actions, The Case- Or-Controversy Requirement, and the Nature of the Adjudicatory Process, 73 U. Chi. L. Rev. 545, 557 (2006) (noting that the use of the class action device has enabled plaintiffs' lawyers to "extract a settlement far in excess of the individual claims' actual worth") (quotation omitted).

54 Regents of the Univ. of Calif. v. Credit Suisse First Boston (USA), Inc., 482 F.3d 372, 379 (5th Cir. 2007).

55 Watson, supra note 18, at 279; John C. Kleefeld, Class Actions as Alternative Dispute Resolutions, 39 Osgoode Hall L.J. 817, 827-28 (2001) ("In the class action context, [ ] certification is everything.").

56 See Watson, supra note 17, at 278.

57 Watson, supra note 17, at 273 ("The Canadian criteria for certification of a proceeding as a class action are relatively undemanding."); Wright, supra note 44, at 467 ("Canadian courts are somewhat more inclined to certify classes").

58 Arthur Miller, Of Frankenstein Monsters and Shining Knights: Myth, Reality, and the "Class Action Problem," 92 Harv. L. Rev. 664, 670 (1979).

59 Id. at 669 (citing the1966 amendments to the Federal Rules and their accompanying literature which can be found at 39 F.R.D. 69 (1966)).

60 Id. ("The Advisory Committee's objectives in rewriting the [federal class action rule] were rather clear. It had few, if any revolutionary notions about its work product. Although it was expected that the revision would operate to assist small claimants, the draftsmen conceived the procedure's primary function to be providing a mechanism for securing private remedies, rather than deterring public wrongs or enforcing broad social policies.").

61 Id. at 670.

62 Id. at 672-74 (discussing, among others, the Civil Rights Act of 1964, the availability of court-awarded attorneys fees in antitrust cases, the judicial expansion of the concept of "security," the passage of the Fair Credit Reporting Act).

63 Stephen B. Presser, How Should The Law of Products Liability Be Harmonized? What Americans Can Learn From Europeans, 2 Global Liability Issues (1-2) Feb. 2002.

64 Id. at 4 ("Counsel's fees are then taken from the settlement or judgment, and will generally be a substantial portion of the recovery, as much as 30- 40%, or more").

65 Stephen N. Subrin, Fishing Expeditions Allowed: The Historical Background of the 1938 Federal Discovery Rules, 39 B.C. L. Rev. 691, 739-740 (1998); Richard M. Dunn & Raquel M. Gonzalez, The Thing About Non-U.S. Discovery for U.S. Litigation: It's Expensive and Complex, 67 Def. Couns. J. 342, 342, 346-47 (2000) ("American lawyers are accustomed to the broad, sometimes intrusive, discovery available to litigants in the United States under federal and state rules of procedure.").

66 See Watson, supra note 17, at 277.

67 See Sherman, supra note 19, at 156; S. Gordon McKee, Why Development of Mass Torts in Canada is Important to Corporate America, vp Def. Couns. J. 32, 36 (2004).

68 See Watson, supra note 17, at 274.

69 Id.

70 McKee & Cook, supra note 25, at 42-43 (2006) ("Given the potential for alleged wrongs to occur across international boundaries, plaintiffs can be expected to rely more and more on cross-border discovery to gather the evidence necessary to support their motions for certification in Canada."). An example of this emerging cross-border coordination can be seen in the Vitapharm litigation. See Vitapharm v. F. Hoffman-LaRoche, Ltd, [2001] 11 C.P.R. (4th) 230 (Ont. Super. Ct.) (affirmed, [2002] 18 C.P.R. (4th ) 267 (Ont. Div. Ct). There, vitamin manufacturers were accused of fixing prices in both Canada and the United States. Lawsuits were pending in both countries, and the defendant sought to prevent the Canadian plaintiffs from gaining access to U.S. discovery. The court denied the defendant's request, noting that "as a result of the inexorable globalization and expanding international free trade and open markets, there will be an ever-increasing inter-jurisdictional presence of corporate enterprises. This is seen particularly in respect of American and Canadian business activity, given the extent of cross-border trade." Id. ¶ 27. Because access to the U.S. discovery could save time and allow plaintiffs to "more easily determine and discard what is not relevant for the purpose of the Canadian proceeding," id. ¶. 48, and because allowing the access did not cause unfairness to the defendants in the Canadian courts, id. ¶ 50, the Court permitted the sharing of information and denied the defendant's motion.

Originally published by The Federalist Society for Law and Public Policy Studies (www.fed-soc.org) in Engage: The Journal of the Federalist Society's Practice Groups.