Cheers! Where Nobody Knows Your Name: An Analysis of Commercial Host Liability in Canadian Tort Law
Nathan McLean, 2L, Volume 81 Articles Editor
In this first edition of Forum Conveniens' "Term Papers into Blog Posts" program, Articles Editor Nathan McLean converted his tort law essay into this blog article. Contact forumconveniens.lawreview@utoronto.ca if you too would like to turn one of your course essays into a post on our blog.
In the landmark Jordan House Ltd v Menow, [1974] SCR 239, 1973 CanLII 16 (SCC) (Jordan House) decision, Laskin J, as he then was, asserted that the law did not impose “a duty on every tavern-owner to act as a watchdog for all patrons who enter his place of business and drink to excess” (250). Since then, the scope of the duty for commercial hosts has markedly expanded: see, for example, Schmidt v Sharpe, (1983) 27 CCLT 1 (Ont HC) (Schmidt) and Stewart v Pettie, [1995] 1 SCR 131, 1995 CanLII 147 (SCC) (Stewart).
The basic fact pattern for commercial host liability cases is as follows: a tavern served alcohol to a patron that it ought to have known was intoxicated. Then, the patron engaged in some activity (often driving) that caused self-injury and/or injury to a third party, and the tavern did not take reasonable steps (such as providing alternative transportation) to prevent the injury. The tavern is held liable.
While the issue in Childs v Desormeaux, 2006 SCC 18 (CanLII) (Childs) squarely concerns social host liability, much of the decision is spent distinguishing social hosts from commercial hosts to justify liability for the latter. Childs synthesizes the commercial host liability jurisprudence and provides a clear picture of the law. To discuss the general principles that animate commercial host liability rather than the minutiae that distinguish individual cases, I therefore use Childs as the bedrock of my analysis.
In Childs, McLachlin CJ constructed her argument for the non-liability of social hosts against the backdrop of commercial hosts’ assumed liability. In this article, I argue that Childs fails to distinguish social and commercial hosts successfully. If the twin host liabilities are insufficiently different, the question then becomes whether both, or neither, should be held liable. McLachlin CJ’s argument is convincing in explaining the rule against assigning liability for social hosts, but too convincing to preserve the law’s current assignment of liability for commercial hosts.
In the paragraphs that follow, I critically assess the reasons McLachlin CJ provides for assigning liability to commercial hosts and examine alternative rationales for imposing the duty—most significantly, through the economic analysis approach. I ultimately conclude that the relationship between commercial hosts and their patrons should not give rise to the positive obligations that the law currently imposes.
Analysis
The discussion of commercial host liability begins with whether the issue is framed as an act or a failure to act. The framing of commercial host liability as a positive act implies that the host was negligent in serving alcohol to an intoxicated patron and should be liable for all foreseeable consequences that flow. Under this approach, as noted by Lewis N. Klar in “The Role of Fault and Policy in Negligence Law”, a duty of care would also be imposed on social hosts who serve alcohol to guests (Klar at 37). The SCC rejected this formulation of the issue, stating that overservice is not the basis for host liability (Childs at para 33).
Instead, liability forms when the host fails to act in taking preventative measures that ensure intoxicated guests are not dangers to themselves or those they may encounter. As a legal question, overservice is not dangerous per se; instead, the danger relates to the subsequent inaction in permitting the inebriated guest to drive. Childs therefore characterizes liability as prima facie nonfeasance.
A positive duty to act may arise where there is a special relationship of proximity between the host and guest. In Childs, McLachlin CJ found three situations where a host’s position is sufficiently proximate to their guests such to give rise to a positive duty to act. Namely, where the host: i) is in a paternalistic position; ii) is the controller/creator of an inherent risk; or iii) serves a public function or operates a commercial enterprise (Childs at para 38). A tavern falls into the third category.
McLachlin CJ distinguished social from commercial hosts on three grounds. First, she concluded that commercial hosts have an advantage over social hosts because they are more aware of how much alcohol someone has consumed (Childs at para 18). This is an unsubstantiated generalization. It cannot be said that a social host who invites a small group of people to their home and watches how much they drink the entire time is less aware than, for instance, a bartender at a nightclub who serves hundreds of patrons (and often in a situation where people are buying drinks for others); see, for example, Williams v Richard, 2018 ONCA 889 (CanLII) where the facts present a situation where a social host monitoring alcohol consumption is simple. In fact, the confines of social gatherings arguably allow many social hosts to have a better indication of their guests’ intoxication levels than commercial hosts. Even commercial hosts who can monitor their guests’ intoxication levels by using consumption as a proxy are typically unaware of prior drinking. While the expectations of the host, rather than the actual awareness of a patron’s drinking, is the fundamental focus of the legal inquiry, reasonable expectations should be grounded in reality. The capacity of a host, whether commercial or social, to accurately monitor their guests’ consumption levels depends more on circumstances – such as size and prior drinking – than the presence of a regulatory regime in the commercial setting.
The second differentiation McLachlin CJ drew between social and commercial hosts is about reasonable reliance; see Cooper v Hobart, 2001 SCC 79 (CanLII) where the proximity analysis considers, inter alia, expectations and reliance, assessing what people are implicitly invited to do (para 34). She concluded that the regulatory regime and broader social expectations regarding the commercial sale of alcohol create a context wherein guests can reasonably rely on establishments to take care of them (Childs at para 46). But is this really the case? Does the quid pro quo of alcohol consumption at a tavern actually include the establishment’s care if the patron becomes intoxicated (see Klar at 36)? While consideration (and expectation) and reliance are discrete, if a patron does not expect something as part of the service, it is difficult to argue that they reasonably rely on it. A commercial vendor of alcohol is neither a “watchdog” (Jordan House at 250) nor is it a daycare for drunk adults. Patrons frequent these venues to socialize, eat, drink, and enjoy themselves. This is in contrast with, for instance, the detention centre in the famous Dorset Yacht Co Ltd v Home Office, [1970] UKHL 2 case which was held liable when its detainees escaped and caused property damage. There, a key raison-d’être of the defendant was to confine known dangers away from the public. McLachlin CJ’s characterization of the expectations governing the relationship between taverns and their patrons is a quixotic prescription of what (she believes) ought to be, or at least an uncited statement about the current reality.
The two other exceptions which induce a positive duty to act—the creation of an inherent risk and a paternalistic relationship—both invite reliance in a way that the commercial relationship does not. For instance, someone who invites another person to an inherently dangerous situation may be reasonably relied on to rescue the other if the danger arises. Likewise, a supervising adult is implicitly relied on to take care of a child. The duty that commercial hosts owe to their patrons has no parallel reliance invitation. Far from being built on a close connection between the plaintiff and the defendant where reliance is implicit by virtue of the relationship between parties (as the first two duties are), McLachlin CJ suggests that commercial hosts’ duties arise from public expectations (Childs at para 40).
Moreover, to incur an injury because of reliance, plaintiffs must change their position based on that reliance. If commercial hosts did not owe a duty, there is an absence of any evidence that suggests patrons would refrain from consuming as much alcohol or that the public (in the event that the plaintiff is a third party) would drive less as a measure of precaution. Reliance infers patrons or the public would have avoided the harm if they were aware the tavern would not fulfill its duty. If taverns seldom fulfill their positive obligations, that is evidence in itself that both patrons and the public do not rely on them (see, for instance, Clapp et al. (2009) “Blood alcohol concentrations among bar patrons: A multi-level study of drinking behavior”).
It is worth understanding what the parallel tort of negligent misrepresentation considers reasonable reliance. To establish liability in cases of negligent misrepresentation, a plaintiff must prove that they actually relied on a misleading representation (i.e., the plaintiff did something they would not have otherwise done). In host liability cases, there does not seem to be anything akin to holding out, or a specific transaction that resulted in a loss, making it unclear what would constitute the relevant detrimental reliance. Patrons and the public essentially conduct themselves as if these obligations did not exist (regardless of whether or not the jurisdiction also imposes a statutory duty on commercial hosts); intoxicated patrons habitually find safe ways home without a modicum of assistance from the tavern. Therefore, the argument for proximity based on reliance is unconvincing.
The final distinction that McLachlin CJ draws between taverns and social hosts is the profit motive present in the commercial context. Klar provides a compelling counterargument, suggesting that “there is nothing inherently special about profiting from an activity,” and alluding to a situation in which both “commercial and gratuitous” drivers owe their passengers a duty of care (Klar at 36). Profit is not typically determinative in imposing a duty of care. Childs suggests that these cases are exceptional because of the “perverse incentive” (i.e., profit) to overserve (Childs at para 22). Under this reasoning, commercial hosts could escape liability when they provided free drinks. This would be an absurd distinction to make. Moreover, the duty of care extends to “bring your own wine” establishments (irrespective of whether there is a corkage fee) where potential overservice is not necessarily connected to the profit motive, undermining the correlation between profit and legal duty: see “Responsible Service Tip Sheet: Bring Your Own Wine Endorsement” issued by the Alcohol and Gaming Commission of Ontario in 2018. Finally, social hosts may also have non-profit incentives to overserve—for example, to reduce the inhibitions of their guests or garner social reputation. Fundamentally, the profit justification is unsatisfactory as a rationale for imposing a duty on commercial hosts.
The strongest argument for assigning liability to commercial hosts follows the economic analysis approach. By serving alcohol to patrons, a commercial host is participating in an activity with associated risks to the patron and the public. To incentivize the commercial host to take precautions and mitigate the risk of injury, they should share in the liability if injury occurs. The profit factor is most relevant as a means of measuring incentives, rather than as an independent motivation for imposing a duty, as in Childs. If one is concerned about creating an economic incentive structure to protect against the dangers of overservice, it is also important to consider the profit that the commercial host earns from serving alcohol. Imposing liability makes a tavern internalize the cost of overservice proportionate to the risks they fail to protect against.
The economic analysis approach to negligence cases is captured in the Learned Hand formula, as described by Richard Posner in the Economic Analysis of Law, 2nd ed: “the defendant is guilty of negligence if the loss caused by the accident, multiplied by the probability of the accident’s occurring, exceeds the burden of the precautions that the defendant might have taken to avert it” (122). The cost of precautions in commercial host liability cases is often low while the expected accident cost is high. While certain cost factors vary from case to case, the economic standard implicit in the Learned Hand formula would assess liability in most commercial host liability cases. In fact, liability would likely be assigned to social hosts who face similarly trivial burdens and substantial expected injury costs. Let us revisit the Learned Hand formula in the general case of social hosts. Because commercial hosts have actual profit to gain from overserving alcohol while social hosts do not – considering the marginal loss of profit as a burden of precaution, the social host would be subject to more liability. This outcome is particularly curious when contrasted with the Childs decision which rejects liability for social hosts under the assumption of liability for commercial hosts. This is not to suggest that the analysis is inherently flawed; it simply illustrates that the economic approach fails to account for the imposition of positive obligations alongside other factors that comprise the relationship between parties. There is fundamentally no distinction between nonfeasance and misfeasance under the scrutiny of a pure economic analysis.
This economic analysis approach could be construed as advancing a policy-based argument for distributive justice. Whereas corrective justice “links two parties and no more because a relationship of correlativity is necessarily bipolar,” as described by Ernest J. Weinrib (351), distributive justice spreads a burden to potential parties based on a distributive criterion (349). In cases of host liability, the burden could be shared between intoxicated patrons and the tavern that failed to protect them (and, by extension, the public at large). Assigning liability to commercial hosts uses tort law as an apparatus for “distributing burdens in accordance with conduct or character” (Arthur Ripstein, Private Wrongs at 85) based on the view that overservice is immoral/risky and that all parties involved should bear responsibility for its consequences.
It is also worth considering whether tort law serves a broader social function of deterring “wrongful” conduct. When a court holds a commercial host liable, it forces it to internalize the costs resulting from its failure to take accident-avoidance precautions. The award of damages disincentivizes what it has determined is negligent conduct and encourages taverns to take measures to curtail excessive risk in the future. This implements an economic approach—a tavern will take precautions to abate accident costs when the costs exceed the burden of precaution. In theory, a socially-optimal result is therefore realized; see Ripstein at 265 where he suggests that the economic analysis is “unable to explain why the measure of those damages should be the plaintiff’s injury.”
However, if deterrence is the true objective, there is no reason for a connection to damages. Tort law is necessarily backward-looking—it will only assess liability in response to damages; conversely, deterrence is a forward-looking purpose. Policymakers seeking to discourage risky behaviour should not wait for damage to have occurred. If overservice and the ensuing failure to protect against actions flowing from it is deemed sufficiently immoral/hazardous, it should not be a tort but a regulatory offence prohibiting the wrongful conduct. A tavern is no less morally guilty if inebriated patrons manage to drive home safely than if they got into an accident. The law’s ability to pursue the deterrence goal should not depend on a stroke of chance.
Conclusion
At its core, the question of whether to impose liability on commercial hosts is a question about what the aim of tort law ought to be. The economic approach will assess liability in most cases. On the other hand, barring an invitation of reliance that yields a positive obligation, a rights-based approach will only hold a commercial host liable when their conduct interferes with the autonomy of a patron. Does advancing public policy goals like welfare maximization and deterrence fall within the ambit of tort law? If so, why not impose liability for sexual assaults or batteries that occur in the wake of overservice – more common consequences of intoxication than drunk driving accidents (see, for instance, Abbey et al.’s “Alcohol and Sexual Assault” and Statistics Canada’s “Impaired driving in Canada”)? I propose that there are more efficient and appropriate ways to pursue these goals than the inherently retrospective tort law; these include, inter alia, insurance programs, taxation, broader regulatory schemes, and criminal law.
I endorse the view that the tort law should seek to implement corrective justice. It follows that commercial hosts should generally not be held liable for failing to prevent their intoxicated patrons from engaging in dangerous activities. In Childs, McLachlin CJ states, “the consumption of alcohol, and the assumption of the risks of impaired judgment, is […] a personal choice and an inherently personal activity” (Childs at para 45); it is free of external interference. She underscores the fact that a partygoer “does not park his autonomy at the door” (para 45); the same is true for patrons of taverns. As the Chief Justice notes, “the common law is a jealous guardian of individual autonomy” (para 31). It both protects people from others’ interference with their autonomy and is loath to interfere with peoples’ autonomy. It would take an exceptional relationship of invited reliance to justify imposing a positive duty to act. The relationship between commercial hosts and their patrons should not meet this high bar.
[1] I would like to thank Professor Peter Benson and Professor Arthur Ripstein for their helpful feedback and guidance.