Jan 21, 2019

From Riches to Rags: Tragedy in Bangladesh and a Rejected Class Action in Ontario

Das v. George Weston Limited, 2018 ONCA 1053 (CanLII)


Canadians are a compassionate nation, and they are concerned about wrongs committed by Canadian companies in foreign lands. In a novel development in the past few years, some claims have been brought in Canadian courts by foreign plaintiffs for damages they suffered in their home countries. These claims have been mainly against Canadian mining companies.

One of the principles of tort law under the Anns-Cooper test used by Canadian courts is that responsibility for damages should be based in part on public policy considerations. It is probably a valid statement of public policy that Canada does not want to be enriched by unethical or harmful practices carried out by Canadian companies in third world countries. Therefore, public policy considerations would not be an impediment to this type of claim. On one theory, the aim of tort law is not only to compensate the victims in the current case, but also to serve as a deterrent so that others will not commit similar wrongful acts.

In the cases of the Canadian mining companies, the alleged wrongs were carried out by their wholly owned foreign subsidiaries. A rather different type of claim has been brought against the Canadian retail giant Loblaws in connection with an industrial tragedy in Bangladesh, the collapse of the Rana Plaza building in 2013. That collapse is believed to be the deadliest accidental structural failure in world history, killing 1,130 people and injuring 2,520.

It is reported that the building had not been constructed nor approved for industrial uses. Incredibly, at the time it collapsed, a ninth story was being added to the top of it, making an already bad situation even worse.

A class action claim was brought in Ontario against Loblaws and its affiliates. The plaintiffs’ motion to certify this claim was rejected in Das v George Weston Limited, 2017 ONSC 4129 (CanLII). The decision of the motion judge was upheld by the Court of Appeal in Das v. George Weston Limited, 2018 ONCA 1053 (CanLII).

The issues are complex, as the motion judge’s decision ran to 140 single spaced pages, and the Court of Appeal added another 70. Numerous expert witnesses were brought, including experts on foreign law. It was an expensive undertaking. At the end of the day, the plaintiffs (who were backed financially by the Law Foundation of Ontario) were ordered to pay costs of over $1.6 million to the defendants.

The victims have not gone entirely uncompensated. A fund based on donations has been set up for them, and immediately after the collapse Loblaws itself made a voluntary payment to them.

Background to the Claim

Loblaws had no operations of its own in Bangladesh. However, a company named New Wave was under contract to it to manufacture clothing in the Rana Plaza building. It was alleged that Loblaws purchased about 50 percent of the output of New Wave for its Joe Fresh brand. There was some disagreement about this on the motion, but the Court of Appeal stated that “at the time of the collapse, Loblaws purchased about 50% of the garments New Wave produced.” (2018 ONCA 1053, at para. 7).

The plaintiffs argued that this effectively put Loblaws in control of New Wave, and therefore made it responsible for any negligence committed by New Wave. They claimed that Loblaws had a duty of care to the employees of New Wave and should be obligated to compensate them financially for the injuries they suffered.

The motion judge, Justice Perell, rejected certification for two main reasons. The first was due to his ruling that the case, even if tried in Ontario, had to be decided on the basis of the law of Bangladesh. The statute of limitations in Bangladesh requires claims to be brought within one year (rather than two years, as under the law of Ontario). The claim was filed well past the one-year deadline. If correct, that is sufficient to dispose of the plaintiffs’ claim.

Next, Justice Perell proceeded to go through the requirements for certification under class actions law. One of the requirements is that there be a valid cause of action known to law. He concluded that there was not: under the principles of tort law (regardless of whether it is the law of Ontario, England or Bangladesh), Loblaws did not have a duty of care to the workers. He decided that a customer of a manufacturer, even if it is a very large one, cannot on that ground be deemed to control the manufacturer.

Justice Perell went to considerable lengths to distinguish between what is a factual allegation in a pleading (which the judge is required to accept as true at this stage) and what is a legal argument in a pleading. On that basis, he concluded that there was no cause of action, though the other factors needed for certification of a class action were met.

The next two sections will briefly touch on some salient points related to these two issues.

Bringing the Law of Bangladesh to Ontario

There is an old legal principle that deals with this issue, and proof of its age is that it is in Latin: lex loci delicti, “the law in the tort’s location.” This rules states that, regardless of where the trial is conducted, the governing law to be applied in a tort case is the law of the place where the tort was committed. Historically, there had been contrary jurisprudence on this issue in Canada, but the Supreme Court affirmed the principle of lex loci delicti in Tolofson v. Jensen, [1994] 3 SCR 1022, 1994 CanLII 44 (SCC), and it has stood since then.

The important policy consideration in Tolofson was the desire to have a clear rule that would ensure predictability for interprovincial traffic accidents. Nevertheless, in Tolofson Justice La Forest did leave open the possibility that an exception could be made where applying the rule rigidly would lead to injustice:

I have already indicated, of course, that I view the lex loci delicti rule as the governing law. However, because a rigid rule on the international level could give rise to injustice, in certain circumstances, I am not averse to retaining a discretion in the court to apply our own law to deal with such circumstances. I can, however, imagine few cases where this would be necessary.

This is an exception for which the plaintiffs argued, but it was rejected by both the motion judge and the Court of Appeal. One of the arguments made by the plaintiffs was that the law of Bangladesh includes Islamic sharia law, which would lead to unequal division of payments among male and female plaintiffs. Both the motion judge and the Court of Appeal held that this was not a problem. If it got to the stage of apportioning payments to plaintiffs (which it would not, given the rest of the ruling) one could discard the parts of Bangladeshi law that were inimical to Canadian values, while retaining the remainder. That is, one may cherry-pick and use only the parts of Bangladeshi law that are appropriate.

The suit was brought as a class action under the aegis of Ontario’s Class Proceedings Act. As far as one is aware, Bangladesh has no class actions legislation, and even if it did it would not be used for a proceeding in Ontario. The Class Proceedings Act falls into the category of procedural law, and one would not attempt to import the procedural law of a foreign jurisdiction to run a trial in Ontario. By contrast, the law of limitations is considered to fall into the category of “substantive law.” The decision of Ontario’s courts was that the substantive law of Bangladesh ought to be applied, as it was the location where the damage occurred.

The most significant difference between Ontario law and the law of Bangladesh is in the limitations period, as noted by Justice Perell:

[208] It is no secret that the Defendants’ motivation for seeking a ruling that the action is governed by Bangladesh law is not driven by any great difference in the tort law in Ontario and in Bangladesh but rather is motivated by the circumstance that the Plaintiffs commenced their action more than one year after the tragic events at Rana Plaza.

By hindsight, class counsel perhaps made a strategic mistake, arguing that the tort consisted of decision making by Loblaws that occurred in Ontario. The judge rejected that reasoning, and it left the plaintiffs in a weaker position. The Court of Appeal solidly backed this finding:

[87] In his analysis, the motion judge relied on the principle of tort law that there is no actionable wrong without injury. He reasoned that the alleged duty was owed to the people in Bangladesh who were killed or injured there. The impugned decisions, it was alleged, resulted in those deaths and injuries. The wrong therefore occurred in Bangladesh.

A stronger argument might have been made by focussing on the special nature of a class action and the injustice of applying a limitation period from a 1908 statute in British India to an Ontario class action more than 100 years later. The practical difficulties leading to delays in filing a claim are vastly different.

One could well argue that in this case the rule in Tolofson v. Jensen ought to have been distinguished on its very different facts. Tolofson v. Jensen dealt with claims in automobile accidents that occurred in provinces different than the one where the claimants lived. That is a regular occurrence in Canada, as people frequently drive across provincial borders, and it makes sense to have a clearly defined rule. Nobody will be subjected to injustice by it, as there is easy access for residents of any province to experts in the law of another province. Indeed, there are now some personal injury law firms that have franchises in multiple provinces.

By contrast, in the unusual situation where a Canadian corporation commits a wrong in another country, justice may well require the law of Canada to be applied. For example, there are still some countries with laws that permit what is effectively state-sanctioned slavery by prisoners or conscripts. Would one allow a Canadian company that uses slave labour in its foreign mine to hide behind such a foreign law?

The one-year limitation period on filing a claim in Bangladesh may not appear to be as cruel as slavery, but it too can lead to significant injustice. The court has conceded that a class action in Ontario might in principle be allowable, but has made it impossible in practice by ruling that Bangladesh’s one year limitation period should apply.

Given the logistical problems of getting a complex class action of this type underway, and operating at such a distance, it is not surprising that no claim was filed in Ontario within that one-year window. An expensive piece of litigation such as this needs to get funding from a group such as the Law Foundation of Ontario. The latter has a committee of five experts that reviews applications, and they have to be convinced that there is some chance of success. That application and review process inevitably takes time, and has to be added onto the usual steps in preparing a case.

It is certainly not a foregone conclusion, based on Tolofson v. Jensen, that Bangladesh’s limitation period should apply. Once one has acknowledged that it is permissible to cherry-pick from the law of Bangladesh, it appears just that a one-year limitation period should not be imported for an Ontario class action.

The law of Bangladesh and other foreign jurisdictions (England and Wales) might still be applied to deciding to the more fundamental question of whether there is a cause of action. That is discussed in the next section.

Plain and Obvious that there is no Cause of Action?

The criterion for rejecting a claim as not disclosing a cause of action is supposed to be a difficult one to meet. It has to be “plain and obvious” that there is no cause of action: Hunt v. Carey Canada Inc., [1990] 2 SCR 959, 1990 CanLII 90 (SCC).

Beside the limitation period, the law of Bangladesh was also canvassed on the second issue, that of whether there is a cause of action. Bangladesh was part of India under British dominion until 1947. The tort law in Bangladesh draws on the English common law tradition, and leading decisions on tort law from England are persuasive in Bangladesh – as indeed they are in Canada.

This issue led to further escalation of the expense of the case, as both sides also had to retain expert witnesses on the tort laws of Bangladesh and England, including retired justices of the Bangladeshi supreme court, and eminent law professors from England.

That is perhaps ironic, given the open-ended nature of tort law. Justice Perell, as is his wont in momentous cases, began his decision with quotations from the classics: one from the Book of Luke on “who is my neighbour,” and the other on the same question from Lord Atkin’s speech in Donoghue v. Stevenson, [1932] AC 562 at p. 580. That is of course the decision that revolutionized British tort law, by allowing the plaintiff’s motion to proceed on a claim that the producer who allowed the snail into the bottle of ginger ale might have a duty of care as the consumer’s “neighbour.”

It continues to be said that “the categories of negligence are not closed,” as McLachlin C.J.C. said in Cooper v. Hobart, [2001] 3 SCR 537, quoting the aforementioned Lord Atkin. This makes it perhaps pointless to speculate about how the Supreme Court of the United Kingdom might think differently, if at all, from the Supreme Court of Canada on a novel tort claim of this type. It is often difficult to predict how either will rule on a complex tort matter.

In either country, one regularly encounters tort claims where the Court of Appeal was unable to predict how its Supreme Court would rule, and gets overruled as a result. In the last three duty of care decisions of the Supreme Court of Canada, the Court of Appeal’s decision was entirely overturned in two of them, and partly in the third.

Similarly, in Donoghue v. Stevenson, the House of Lords overturned the ruling of the Court of Sessions, Scotland’s appeal court. Justice Perell, cognizant that the law sometimes moves forward to recognize new duties of care, nevertheless cautions that there is a limit on how much it will move. In contrasting the biblical view of the neighbour principle and the legal view, he noted that Lord Atkin “made it clear that the legal answer to the duty of care question is much more restricted and circumscribed than the ethical answer provided by Jesus to the “who is my neighbour” question.” (2017 ONSC 4129 at para. 398).

Justice Perell was perhaps more aware of the futility of trying to distinguish between English and Ontario tort law on this issue. He made a point of emphasizing his view that there was no cause of action either under English law or Ontario law. By contrast, the Ontario Court of Appeal went through the ambitious process of stepping into the robes of the Court of Appeal of England and Wales and reviewing some recent English decisions in considerable detail.

On a motion to certify a class action, appeal courts are supposed to grant considerable deference to the motion judge. On that principle, the plaintiffs’ appeal was largely doomed to fail. In this instance the motion judge, Justice Perell, had expressed very negative views about the factual basis of the claim. He made highly critical comments about the reasoning displayed in the plaintiffs’ statement of claim:

[22] There are numerous examples where the Plaintiffs plead conclusions or beg a question that the court must decide. To illustrate, the Plaintiffs plead the conclusion that Loblaws had control over Pearl Global and New Wave’s workplace because Loblaws could take remedial steps under the agreements it had with them, which agreements required compliance with Loblaws’ CSR standards. Thus, in their factum, the Plaintiffs submit that the court is obliged to decide this motion on the basis that if Loblaws and Bureau Veritas had taken action within their control, then the disaster at Rana Plaza would never have occurred. I disagree that I am obliged to decide a motion based on a tautological assertion.

This was endorsed by the Court of Appeal:

[74] … while the material facts that are pleaded in the statement of claim are assumed to be true for purposes of a motion to strike, bald conclusory statements of fact and allegations of legal conclusions unsupported by material facts are not....

[76] While the principles are easily stated, the record before the motion judge was voluminous. The Fourth Amended Statement of Claim contained 261 paragraphs. The parties also filed 73 volumes of evidence and compendiums, including extensive evidence on foreign law. The motion judge thoroughly reviewed all of the material. His findings regarding what pleadings were facts, and what were legal conclusions or attempts to characterize facts to fit a legal theory, were fair and proper and fell within his purview.

On its review of the English case law, the Court of Appeal relied partly on the English expert witnesses, but also did some review of its own. Feldman J.A. modestly observed that “Canadian courts routinely consider English jurisprudence when applying domestic law in the absence of expert evidence on the English jurisprudence” (2018 ONCA 1053 at para. 173). Having considered these cases, she concluded that the plaintiffs’ claim had no chance of success:

[179] Simply put, the pleaded facts in this case do not amount to the type of relationship or control over New Wave’s operations by Loblaws that has been found in English law to be sufficient to establish proximity or assumption of responsibility, and to thereby impose a duty of care to protect against harm by third parties.

[181] Even if the relationship factors could be viewed as sufficient to give Loblaws some control over New Wave’s operations, and to establish some degree of proximity, there is no pleaded basis to extend that control to structural issues with Rana Plaza, or to find that Loblaws assumed responsibility for any structural issues.

A Possible Alternative Argument

There is no doubt that the claim that Loblaws ought to be responsible for the workers of its contractor in Bangladesh is a novel one that would extend tort law beyond its current bounds. However, the facts of this case are unusual and create the possibility of an element of control that does not normally exist in the relation between arm’s length business associates. The fact that Loblaws had come to purchase such a large portion of the output of New Wave might be able to support an argument that there was proximity between it and New Wave’s workers. It is certainly not a foregone conclusion, but it should require a trial to weigh all the evidence, rather than rejecting the claim at the certification stage.

It is recognized in many areas of law that a legal relationship is not immutably defined by simple categorization, but depends also on quantitative factors. For example, in corporate law a small shareholder will have no liability for torts committed by a corporation, while the corporate veil may be pierced to make a controlling shareholder liable.

One of the points noted by Justice Perell is that there is an important legal difference between active misfeasance and a mere omission to act. Tort law is very reluctant to impose liability for the latter. Justice Perell went on to note (at para. 401) that Loblaws did not “construct Rana Plaza or cause it to collapse.” Is the correctness of the statement that Loblaws did not cause it to collapse plain and obvious, or is it something that ought to have been determined at a trial?

A ninth floor for Rana Plaza was under construction to make the building even larger -- quite remarkably, this was going on while thousands of people toiled on the floors below. Evidence at a trial might have been able to prove that the extra stress from adding this ninth floor caused the collapse. That ninth floor would also have been occupied by New Wave. It might have been proven that this ninth floor would not have been constructed but for the expanding production that Loblaws orders was bringing to New Wave. It is possible that it could have been shown that Loblaws knew that its order was leading to this construction, and therefore it knew or ought to have known that it was increasing the risk of a collapse.

If a customer places an order that it knows will create unsafe conditions for the employees of the manufacturer, does it not have a duty of care? Should it just be able to wash its hands of the matter, and rely on its legal categorization as a mere customer? Should there be a gradation of responsibility, depending on the quantitative issue of how large a customer it is? Arguably, it is a logical development of tort law to say that being a very large customer that has a causal impact creates proximity with the workers in the factory.

Assigning this level of responsibility would extend Canadian tort law, but it would not open the floodgates. One of the principles of the Anns-Cooper test is the public policy one of avoiding “the spectre of indeterminate liability:” Cooper v. Hobart, [2001] 3 SCR 537, 2001 SCC 79 (CanLII), at para. 54.

Most Canadian importers are not in a situation where they take so large a portion of the production of a manufacturer that they would be seen to have this level of responsibility. Where a Canadian retailer wants to avail itself of such a high degree of involvement with a foreign manufacturer, it would not be a prohibitive expense to require it to maintain the appropriate supervision to ensure that it operated safely.

In the case under discussion, Loblaws was already aware of some degree of responsibility and had retained a company that undertakes "social audits" (its co-defendant in the case). Loblaws had not retained it to check the structural integrity of the building, which was apparently available as an option at a modest extra cost. The majority of Canadian consumers would probably say that they prefer to pay a few cents more for their garments to prevent this type of tragedy. By hindsight, Loblaws management and shareholders may feel the same way.


This case was quite complex, as seen from the fact that it required a decision of 140 pages on the certification motion. One is entitled to wonder whether it was really plain and obvious that the plaintiffs had no chance of success that would have entitled them to a trial of the issues. That would have allowed a more thorough review of the evidence about the relation between the defendants and their Bangladeshi supplier than was feasible on a certification motion. Unfortunately, in the context of class actions, there is a tendency for the meaning of “plain and obvious” to become rather stretched, making it a substantial hurdle to certification.