Vicarious Liability: The Two Approaches
This is a special guest contributor article.
On June 17, 1999, the Supreme Court of Canada handed down two decisions dealing with the same topic. One of them, Bazley v. Curry, has become the leading Canadian case on vicarious liability; the other one, Jacobi v. Griffiths is less well-known. A quantitative indication of their different fates is that on CanLII, Bazley has roughly twice the citations as Jacobi. My more anecdotal impression is that Bazley has also been cited far more frequently by courts outside Canada.
This difference need not mean anything significant. Often, when an appellate court issues two or more decisions on the same day dealing with the same legal question, one decision provides a general statement of principle and the other decisions, invariably shorter, apply it. But in the case of Bazley and Jacobi, the two decisions reflect quite different views on the topic. A clear indicator of this difference is that Justice McLachlin who authored the decision in Bazley, dissented in Jacobi.
To understand the two decisions and their difference, it helps to go back to first principles. The simplest cases of vicarious liability involve employees committing torts at the behest of an employer. A bank employee instructed by her employer to defraud clients by giving them false information is an example. Cases like this raise no serious issues, as it is clear that the employee is doing the employer’s bidding. Only slightly more difficult are cases in which an employee messes up while performing a task given by the employer. Think of an employee of a parcel delivery company, who in the course of making a delivery, drives carelessly and injures a pedestrian. Even if it can be shown that the employee did not follow the employer’s instructions, the tort here is incidental to the employee doing her job.
Bazley and Jacobi were more difficult because they dealt with the scope of vicarious liability when the employee was acting against the professed aims of their job. Both cases dealt with sexual abuse of youths by employees who held positions of authority in educational facilities. Though neither case dealt with abuse by priests, it was around that time that stories began coming to light of countless cases of sexual abuse by people in positions of educational or religious authority, sometimes stretching decades. Even when the tort liability and criminal responsibility of individuals was not in doubt, the real battle, in the courtroom and beyond, often focused on questions of institutional liability. Bazley was one of the first cases, in Canada and abroad, to chart a path to imposing vicarious liability on the employing institutions in such cases.
The crux of the problem that Bazley, Jacobi, and numerous other cases like them have all raised, is whether vicarious liability should be found in cases where the employment relationship provides an opportunity to commit a tort but the tortious acts are clearly in conflict with the ends of the employment, and in fact explicitly prohibited by law and the employer. It is in these cases that the different approaches of Bazley and Jacobi often lead to different outcomes.
Justice Binnie’s majority opinion in Jacobi reflected the more traditional to this question. As he put it, his decision was an attempt “to explain the existing case law, not to provide a basis for its rejection.” Central to that older case law was what’s known as the Salmond test, named after New Zealand tort scholar and judge John Salmond. This approach focuses on whether the employee’s tortious acts can be seen as part of doing his job. As long as the employee could still be seen as doing his job—still making that parcel delivery, still engaged in instruction—vicarious liability was appropriate. If, on the other hand, the employee “went rogue,” taking advantage of the employment relationship for personal benefit, there was no room for vicarious liability. Justice Binnie’s judgment in Jacobi was an attempt to revamp and adapt this approach to modern times
Justice McLachlin’s approach in Bazley was different. In her words, vicarious liability is “generally appropriate where there is a significant connection between the creation or enhancement of a risk and the wrong that accrues therefrom.” Here the connection is not so much between the aims of the employment and the employee’s actions; instead, the key question is whether the job gave the employee an opportunity he would not otherwise have had to commit the tort.
The difference is not just in the “test,” in whether to prioritize the relationship between employment tasks and the tort, or between the employment opportunity and the tort. The two approaches also reflect different philosophies. While neither approach rests on a finding of an employer’s fault, the more traditional approach places greater emphasis on a positive connection between the tasks the employer gave the employee—the job requirements—and the commission of the tort, to make it possible to attribute the tort of the employee to the employer. If we can see the employee’s tort as part of her job, we have the connection that makes it appropriate to make the employer responsible for it.
The other approach thinks of vicarious liability in more functional terms. For proponents of this approach, whatever its historical origins, vicarious liability should be understood more as a legal mechanism created to make sure that those injured by a wrong do not end up without compensation. Of course, some connection of the employer to the tort is needed (or else, one could turn to anyone with means to compensate the employees’ victims), but the connection here is not based on anything the employer did or did not do, and more to do with its position. On this view, it is appropriate to impose liability on the employer in such cases, not simply because it has deeper pockets, but because it is in a unique position to deal with the possibility of an employee’s tort by trying to prevent it from happening (by greater monitoring and training of employees), by spreading the costs of this liability to all consumers, or by purchasing liability insurance. At its most extreme, the direct liability of the employee almost drops out of the picture, while vicarious liability becomes one of the costs of an employer’s business.
Though this approach does not necessitate adopting the opportunity test, the two approaches go well together. If the primary aim of vicarious liability is to guarantee compensation to those wronged by a tort, the focus of vicarious liability shifts from tests for attributing employees’ torts to employers to tests that emphasize adequate responses to risk creation.
In many cases, the two approaches will yield the same outcome, so it is easy to miss, or downplay, their differences. Some cases of criminal activity unauthorized by the employer could lead to vicarious liability under Justice Binnie’s approach. Binnie, after all, was on the panel that unanimously decided that vicarious liability could be imposed in Bazley. But this is because of the unique circumstances of that case, where the job gave the abusive defendant such control over the youths under his supervision that there was a sufficiently strong link between what the employer was asking the employee to do and the abuse.
By contrast, Jacobi showed that the differences between the two approaches are not merely verbal. Unlike Bazley, where the abuse took place in the workplace, in Jacobi the defendant Griffiths built relations of trust with some children under his supervision, then invited them to his home and abused them there. For Binnie, the combined effect of all these considerations was that there was no place for vicarious liability. It is not just that the actions were the exact opposite of the job Griffiths was hired to do; the fact that they took place outside working hours and at the employee’s home meant they were an entirely private affair. For McLachlin, what mattered most was that the job gave Griffiths the unique opportunity to meet and befriend the teenagers he abused, as well as conferred on him the authority that he could exploit later. This sufficed for vicarious liability, even though the torts took place outside work, in terms of both time and place. Though she insisted that the connection between the work and the tort has to be more than but-for causation—the opportunity provided by the employment had to be more than just meeting at work—this is generally a more expansive test for vicarious liability.
The difference between the two approaches had to be resolved somehow, and a few years after Bazley and Jacobi were decided, the Supreme Court took up another case of sexual abuse in an educational facility. The decision in E.B. v. Oblates of Mary Immaculate was once again authored by Justice Binnie and was joined by Chief Justice McLachlin. In what looks like an attempt to paper over the differences between the different approaches in the earlier decisions, Binnie quoted extensively from Bazley and even from McLachlin’s dissent in Jacobi. But despite the effort to present a united front, in the end, Binnie remained faithful to the more traditional approach he had adopted in Jacobi. He insisted that vicarious liability cannot rest simply on an examination “of the risk created by the enterprise,” which is a close formulation of the opportunity test. Rather, vicarious liability depends on “the job-created power and the nature of an employee’s duties as a fundamental component of determining if a particular enterprise increased the risk of the employee’s wrongdoing in relation to the claimant.” I take Binnie’s italics to indicate that the tort must be seen as the tortfeasor acting in his or her capacity as an employee.
Comparing the reasons and outcomes of Bazley, Jacobi, and Oblates would suggest that in the end the more traditional approach of the latter two decisions prevailed. However, as mentioned at the outset, over the years, it is Bazley that has come to be seen as the leading decision on vicarious liability. As a result, some lower courts seem to favour the more distributive approach expressed in it. This is well illustrated by the recent decision of British Columbia Court of Appeal in Insurance Corporation of British Columbia v. Ari.
If abuse in education facilities dominated the case-law a quarter-century ago, Ari reflects the perennial concern of our time, the viewing and disclosure of private information collected in private databases. The case deals with an employee of the ICBC, the Crown corporation that is the sole provider of compulsory motor vehicle insurance in British Columbia. One of its employees was a claims adjuster, who used her access to the ICBC’s database to retrieve and sell private information on customers.
One question the court in Ari had to decide, but on which I say nothing here, is whether the information sold by the employee was in fact private. I will assume that the court correctly decided that it was, and that in disclosing it to someone else, the employee breached plaintiffs’ privacy, a statutory tort in British Columbia. The other main question the court had to decide was whether ICBC could be vicariously liable for its employee’s torts. The BCCA answered it could.
In line with my earlier remarks, in its analysis of the Supreme Court’s jurisprudence, the Court of Appeal made Bazley the central case on vicarious liability, never mentioning Jacobi. Incorrectly in my view, it reduced the difference between Bazley and Oblates to their different “fact-patterns,” rather than my claim that the two cases rest on different philosophies with different doctrinal implications. With the two cases now seemingly adopting the same view, it was open for Justice Griffin, writing for a unanimous panel, to treat the Bazley approach, as the one adopted by the Supreme Court, and based on it conclude that the employer was vicariously liable.
It is true that at one point, the court suggested that the employee’s “unlimited access to the searchable electronic database made her work more efficient and thereby furthered ICBC’s aims,” but obviously she was not doing so when she accessed private information for personal gain. Indeed, the court’s own decision said that these actions were done “for no apparent business reason.” It is clear from the decision that the main grounds for the result were policy reasons: “Unless the employer is liable, the injured party might not be able to recover compensation…and it is the employer who is best positioned to absorb any losses, whether through insurance, higher prices, or otherwise.” The imposition of vicarious liability is thus justified for providing “an adequate and just remedy and deterrence.”
There will be those who will argue that courts commit some kind of category mistake when they rely on policy arguments in tort law. That is not my view, although this is not the place to argue for it in any detail. For present purposes, suffice it to say that courts have for centuries relied on such arguments in all areas of law, including tort law, and that there is no “principled” basis (a word beloved by proponents of this view) to exclude such considerations from some areas of law. As far as Canadian tort law is concerned, there is no doubt that such considerations have been central to the formulation of virtually every doctrine. Those who criticize such arguments for being inconsistent with what tort law is invoke their idealized conception of what they want tort law to be with the reality of what tort law in Canada (and elsewhere) is.
Still, that such considerations are permissible, does not tell us how they play out in concrete cases. My discussion above suggests that at the very least the two approaches have lived side by side for some time now, and quite possibly that precedent favours a more restrictive approach than the one adopted in Ari. Even those who favour a more expansive approach to vicarious liability can acknowledge that it would be good for another restatement of the law on the issue, one that will try to avoid the temptation of blurring the difference between the two approaches. Cases like Ari show that the difference between them is not merely “academic”: It is more difficult to impose vicarious liability in a case like Ari based on the majority’s view in Jacobi.
Ari also demonstrates the problem with casual appeals to policy. In justifying the imposition of vicarious liability for its deterrent effect, the Court of Appeal explained that “the imposition of liability…encourages employers to reduce the risk and to engage in protective measures against the risk.” This is too quick. What “deterrence” means in this context is greater monitoring of employees by their employers. Some of ICBC’s submissions suggest that such monitoring was not possible, perhaps because some employees had to access the databases where private information is collected on a regular basis. One unintended consequence of Ari may be that the court has just given employers permission to surveil their employees’ work, perhaps down to their every keystroke. It is somewhat ironic that a decision that champions the importance of protecting privacy did not consider its potential effects on employee surveillance, a topic of considerable privacy implications. This does not mean courts should not consider the consequence of their decisions; quite the contrary. Nor does it mean that the deterrence rationale is wrong. But if “deterrence” is more than just a ready-to-wear slogan for every case of vicarious liability, its specific implications for particular cases must be considered in every case.