Melbourne Law School - SSRN papers

Melbourne Law School

Legal Studies Research Paper No. 265

Social Hosts’ responsibility for their intoxicated guests: Where courts fear to tread Ian Malkin & Tania Voon

This paper can be downloaded without charge from the Social Science Research Network Electronic Library at: http://ssrn.com/abstract=1021634

JOBNAME: No Job Name PAGE: 1 SESS: 1 OUTPUT: Wed Feb 7 12:00:12 2007 /journals/journal/tlj/vol15/09−15012

Social hosts’ responsibility for their intoxicated guests: Where courts fear to tread Ian Malkin* and Tania Voon† The recent Supreme Court of Canada case, Childs v Desormeaux, provides a valuable opportunity to reassess the High Court of Australia’s position regarding the liability of commercial hosts for the injurious conduct of their intoxicated clientele, as well as the likely Australian approach to social host liability. In Canada, while commercial hosts owe a duty of care to their impaired patrons, social hosts ordinarily do not. In Australia, courts would be reluctant to impose a duty on either type of host. Moreover, the trend in Australia away from imposing liability in tort makes social hosts just one more example of potential tortfeasors with little incentive to engage in responsible conduct, thereby threatening public safety. Difficulties with the potential use of Australian tort law as a deterrent are highlighted by the High Court’s likely response to an innocent passenger like Zoe Childs, rendered a paraplegic at 18 years of age by a partygoer’s drunk driving.

I Introduction On 5 May 2006, the Supreme Court of Canada handed down its decision in Childs v Desormeaux,1 ruling for the first time on social hosts’ responsibility for injury that their intoxicated guests cause to third parties. The case arose from a New Year’s Eve party hosted by Dwight Courrier and his de facto spouse Julie Zimmerman, where guests brought their own alcohol.2 One guest, Desmond Desormeaux, left the party at some time before 1:30 am, driving his own car and taking two passengers: his girlfriend and another friend.3 On the way home, Desormeaux was driving the wrong way down a street and collided head on with another car carrying four people, including Zoe Childs. One of the four was killed and everyone else in both cars was injured, including the 18-year-old Childs, who became a paraplegic.4 Desormeaux’s blood alcohol concentration at the time of the accident was estimated at 0.225, when the legal limit was 0.08;5 in Australia it is generally 0.05.6 He pleaded * Associate Professor, University of Melbourne Law School. † Senior Lecturer, University of Melbourne Law School. We would like to acknowledge the helpful assistance provided by Professor Harold Luntz and the anonymous reviewers for the TLJ. All errors of course remain ours. 1 [2006] 1 SCR 643. 2 Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [11]. 3 Ibid, at [3], [24]–[26], [72]; Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [12]. 4 Childs v Desormeaux [2006] 1 SCR 643 at [2]; Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [1]–[5], [36]. 5 Childs v Desormeaux [2006] 1 SCR 643 at [3]; Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [72]; Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [13]. These numbers refer to grams per 100mL. 6 See, eg, Road Transport (Alcohol and Drugs) Act 1977 (ACT) s 19(1) and Dictionary,

1


2 (2007) 15 Torts Law Journal

guilty in criminal proceedings and was sentenced to 10 years’ imprisonment.7 Childs sued Desormeaux, Zimmerman and Courrier for damages.8 Before this case, the liability in negligence of commercial hosts for the conduct of intoxicated patrons was already ‘firmly entrenched in Canadian law’.9 However, ‘no conclusive authority respecting social host liability’ existed,10 despite consideration of the issue in some lower courts.11 The everyday nature of the events leading to the accident made this decision all the more significant: consuming alcohol at social functions in private homes is commonplace. Indeed, this was one of the underlying concerns of the courts in this dispute.12 Writing for the Supreme Court,13 McLachlin CJ held that, ‘as a general rule, a social host does not owe a duty of care to a person injured by a guest who has consumed alcohol’.14 The court therefore upheld the lower courts’ decisions to dismiss Childs’ action.15 How would the High Court of Australia have decided Childs v Desormeaux? In all likelihood it, too, would have refused to impose a duty of care on Zimmerman and Courrier. Indeed, Australian courts have shown an unwillingness to find even commercial hosts responsible in tort for injury caused by (or to)16 their clientele, and a case of social host liability in circumstances similar to Childs would be lucky to make it to the High Court. Nevertheless, the Canadian case provides a fitting opportunity to evaluate the likely Australian position on social host liability and explore the implications of that position for the role of tort in this country. We begin by briefly noting legislative changes enacted in recent years throughout Australia that limit the value of tort in many circumstances and corresponding judicial trends, in order to explain the legal context in which the issues arising in Childs would be considered. We then analyse the Supreme Court of Canada’s reasoning in Childs, which reflects a particular understanding of the position of social hosts and an approach to establishing duty of care that differs markedly from that applied in Australia. Next, we

7 8 9

10 11

12 13 14 15 16

‘prescribed concentration’; Road Transport (Safety and Traffic Management) Act 1999 (NSW) s 9(2) and Dictionary, ‘low range prescribed concentration of alcohol’; Road Traffic Act 1961 (SA) s 47B(1) and s 47A, ‘prescribed concentration of alcohol’; Road Safety Act 1986 (Vic) s 49(1)(b) and s 3(1), ‘prescribed concentration of alcohol’. Childs v Desormeaux [2006] 1 SCR 643 at [3]; Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [5]. Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [3], [13]. J Neyers and U Gabie, ‘Canadian tort law since Cooper v Hobart: Part I’ (2005) 13 TLJ 302 at 313. See also Childs v Desormeaux (2003) CanLII 25348 (unreported, Ont SCJ, Chadwick J, 10 March 2003, 2003 CarswellOnt 799) at [19]; Childs v Desormeaux [2006] 1 SCR 643 at [8]. Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [29]. Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [75]; Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [30]; E Chamberlain, ‘Alcohol Provider Liability in Canada and the United Kingdom: Legal and Cultural Influences’ (2004) 33 Common L World Rev 103 at 109–16. Childs v Desormeaux [2006] 1 SCR 643 at [42]; Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [112]. Bastarache, Binnie, LeBel, Deschamps, Fish and Abella JJ concurring. Childs v Desormeaux [2006] 1 SCR 643 at [1]. Ibid. Cole v South Tweed Heads Rugby League Football Club Ltd (2004) 217 CLR 469; 207 ALR 52.


Social hosts’ responsibility for their intoxicated guests

3

compare the likely outcome in Australia, examining the legal approach and policy concerns relevant to a case like Childs. We conclude by noting the (lost) potential value of tort in this context and suggest a pragmatic proposal for analysing and responding to these kinds of cases: enhance the focus on breach, in order to promote community safety without over-burdening the public.

II The Australian context The question of whether a duty would be found in a case like Childs should be assessed in the light of recent Australian judicial and legislative trends. With respect to the latter, in 2002 and 2003, most Australian jurisdictions enacted major legislative changes to substantive principles underlying tort law, reducing the capacity for plaintiffs to litigate successfully with respect to personal injuries they have sustained.17 The changes were introduced in a manner that was, for major ‘law reform’ purposes, exceptionally swift.18 Yet the degree of change was ‘unprecedented’:19 a ‘sea-change’.20 The statutory changes are intended to serve a number of objectives, including reducing the cost of liability insurance and heightening personal responsibility and risk-taking behaviour as rationales for denying liability. Other notable consequences of the changes are the curtailment of the quantum of damages that theoretically could have been awarded under traditional common law principles, the reduction of limitation periods, and the provision of additional protections to professionals. 17 See Civil Law (Wrongs) Act 2002 (ACT); Civil Liability Act 2002 (NSW); Personal Injuries (Liabilities and Damages) Act 2003 (NT); Personal Injuries (Civil Claims) Act 2003 (NT); Civil Liability Act 2003 (Qld); Personal Injuries Proceedings Act 2002 (Qld); Civil Liability Act 1936 (SA); Civil Liability Act 2002 (Tas); Wrongs Act 1958 (Vic); Civil Liability Act 2002 (WA). These changes are discussed extensively elsewhere: for a comprehensive review, see B McDonald, ‘Legislative Intervention in the Law of Negligence: The Common Law, Statutory Interpretation and Tort Reform in Australia’ (2005) 27 Sydney L Rev 443. 18 Nicholas Mullany comments critically on the speed of the changes and momentum generated by them: N Mullany, ‘Tort Reform and the Damages Dilemma’ (2002) 25 UNSW L J Forum 876. See also P Underwood, ‘Is Ms Donoghue’s Snail in Mortal Peril?’ (2004) 12 TLJ 39; J Carter and E Peden, ‘A Contract Law Perspective of the Civil Liability Amendment (Personal Responsibility) Bill 2002 (NSW)’ (2002) 54 Plaintiff 19 (describing the NSW amendments as ‘rushed and ill-conceived’); S Clark, C Harris and R McInnes, ‘Tort Reform Take Two’ [2003] L Society J 54 at 58 (‘It is hard to believe, and indeed our American colleagues are amazed, that the entire process has occupied less than a year’). 19 K Booth and J Varghese, ‘Opinion: The Rush to Law Reform in Personal Injuries’ (2003) 28(5) Alt LJ 210. Senator Coonan, Minister for Revenue and Assistant Treasurer, boasted, ‘Such extensive law reform in a limited timeframe is unprecedented in the history of Australian insurance law and, taking into account the complexity of Australia’s multiple jurisdictions, perhaps a first for the common law world’: Commonwealth of Australia, Reform of Liability Insurance Law in Australia, Treasury, Canberra, 2004, p 11. 20 J Dietrich, ‘Liability for Personal Injuries Arising from Recreational Services: The Interaction of Contract, Tort, State Legislation and the Trade Practices Act and the Resultant Mess’ (2003) 11 TLJ 1. Rob Davis describes the changes as an ‘orgy of tort reform’: R Davis, ‘The Tort Reform Crisis’ (2002) 25 UNSW L J Forum 865 at 869. While the intent of these changes in general terms is the severe restriction of the common law, their actual effect is more disparate. Many are of great significance, especially where they affect what courts previously had discretion to do at common law. Others simply restate or clarify the pre-existing positions at common law.



Some kinds of activities are not affected by the new provisions. Although this differs from jurisdiction to jurisdiction, in general terms, claims with respect to dust-related diseases, injuries resulting from smoking or other use of tobacco products, civil liability for intentional acts done with intent to cause injury or death, or sexual assault or other sexual misconduct are excluded from the new regimes and remain subject to the old law. So, too, do workplace-related claims, which are subject to their own restrictions. While some jurisdictions exempt motor accident claims from the scope of these legislative changes, others do not.21 Even in jurisdictions where these statutory changes do not have a direct impact on a case like Childs, in that it involved a motor vehicle accident, they provide a background against which any attempts to impose liability may well be judged. It is true that, on one view, courts should not take into account statutory developments that are not directly applicable to the case at hand.22 Moreover, Australian courts may regard themselves as free to expand tort law in areas where the legislatures have not explicitly restricted it.23 On the other hand, Ipp JA has asserted that, ‘when legislatures throughout the country have legislated or have foreshadowed restricting liability for negligence . . . it would be quite wrong to expand, by judicial fiat, the law of negligence into new areas’.24 We consider it unlikely that courts would ignore the over-arching message signalled by legislatures throughout the country, at least as a general reflection of community values — an undeniably important feature in assessments of liability in the tort of negligence. In any case, prior to these legislative changes, judicial decision-making had already begun to focus more on individuals’ responsibility for their own misfortune, as illustrated in the orders and reasoning in cases like Agar v Hyde,25 Romeo v Conservation Commission of the Northern Territory,26 and Woods v Multi-Sport Holdings Pty Ltd.27 The need to ‘pull back’ on tort awards was also reflected in the attitude of some judges writing extra-judicially. One of the most noted proponents of the need for reform and a shift in the culture of blame, Spigelman CJ, considered how negligence law evolved in the ‘welfare state’: ‘The practical operation of the tort of 21 Injuries sustained in the context of motor vehicle accidents, which were already subject to various restrictions, are excluded from these legislative changes in the Northern Territory, Victoria and Western Australia; no such exclusions exist in South Australia, Queensland or the Australian Capital Territory. In New South Wales and Tasmania, the Acts’ coverage is mixed with respect to motor vehicle accidents. Even though most host liability cases would involve driving, in which case these provisions would not apply in some jurisdictions, this is not the only circumstance in which liability could arise. For example, an impaired visitor injured when diving into a host’s swimming pool could come within the scope of the recently enacted tort ‘reforms’. This would be more akin to the facts of a case like Cole (rather than Childs), where the intoxicated person was also the injured plaintiff. In Russell v Edwards (2006) Aust Torts Reps 81-833 (NSW CA), the court grappled with the effect of s 50 of the Civil Liability Act 2002 (NSW) in the context of a diving injury at the defendant’s home (the plaintiff’s intoxication precluded recovery). 22 McDonald, above n 17, at 454. 23 See, eg, Harriton v Stephens (2004) 59 NSWLR 694 at 722 per Mason P (dissenting). 24 Ibid, at 746. 25 (2000) 201 CLR 552; 173 ALR 665. 26 (1998) 192 CLR 431; 151 ALR 263. 27 (2002) 208 CLR 460; 186 ALR 145.



5

negligence sometimes gives inadequate weight to the conduct of the plaintiff.’28 More recent judicial trends are aptly reflected in the High Court decision in Cole v South Tweed Heads Rugby League Football Club Ltd,29 the Australian case that is in some respects closest to Childs. As that case has been evaluated extensively elsewhere,30 here we merely set out its most important features as context for understanding what might happen to Childs in Australia. Cole involved the potential liability of a commercial enterprise rather than a social host. In that case, a customer attended a rugby club’s champagne breakfast. She commenced drinking at 9:30–10:00 am, consuming a great deal of alcohol throughout the day. She was last served at about 12:30 pm; at 3:00 pm, the club refused to supply her with more alcohol as she appeared intoxicated; at 5:30 pm, she was asked to leave and the manager offered to get her a courtesy bus or taxi home. She refused; her new-found companions said they would look after her; she left the premises. At 6:20 pm, she walked carelessly on to the road 100 metres north of the club and was struck by an oncoming vehicle; her blood alcohol reading was 0.238.31 She sued the driver and the club. Gleeson CJ and Gummow, Hayne and Callinan JJ held in the club’s favour. Gleeson CJ stated that the allegations against the club involved its supposed failure to restrain or prevent the plaintiff from engaging in voluntary behaviour, raising significant autonomy and privacy concerns.32 Similarly, Callinan J contended that, ‘except for extraordinary cases, the law should not recognise a duty of care to protect persons from harm caused by intoxication following a deliberate and voluntary decision on their part to drink to excess’.33 The reasoning of Gleeson CJ and Callinan J was thus consistent with the general trend away from imposing liability.34 In contrast, Kirby J in dissent held that the club owed a relevant duty to its patron. His Honour emphasised that, aside from providing legal remedies for injured persons 28 Hon J J Spigelman, ‘Negligence: the Last Outpost of the Welfare State’ (2002) 76 ALJ 432 at 433. See also Reynolds v Katoomba RSL All Services Club Ltd (2001) 53 NSWLR 43. 29 (2004) 217 CLR 469; 207 ALR 52 (Cole). 30 See, eg, the comprehensive article by G Orr and G Dale, ‘Impaired Judgements? Alcohol Server Liability and “Personal Responsibility” after Cole v Tweed Heads Rugby League Football Club Ltd’ (2005) 13 TLJ 103. 31 Cole (2004) 217 CLR 469; 207 ALR 52 at [6]–[8], [21]–[22], [51]. 32 Ibid, at [14]. Gleeson CJ asserted at [18]: ‘The consequences of the appellant’s argument as to duty of care involve both an unacceptable burden upon ordinary social and commercial behaviour, and an unacceptable shifting of responsibility for individual choice.’ 33 Ibid, at [121]. Callinan J continued: The voluntary act of drinking until intoxicated should be regarded as a deliberate act taken by a person exercising autonomy for which that person should carry personal responsibility in law. . . . [T]o extend the duty to the protection of patrons from self-induced harm caused by intoxication would subvert many other principles of law and statute which strike a balance between rights and obligations, and duties and freedoms. 34 Although Gummow and Hayne JJ appeared to favour the views of Gleeson CJ and Callinan J, they did not commit themselves on the duty question: ibid, at [59], [81]. They ultimately refused to find liability because, even assuming that the club owed a duty, the other elements of the tort of negligence had not been established, namely breach (as the club offered the plaintiff transport home) and causation (as any failure to monitor the plaintiff’s alcohol consumption did not cause her harm).



where it is ‘fair and reasonable and consonant with legal principle’,35 tort law ‘exists to set standards in society, to regulate wholly self-interested conduct’,36 and, in essence, to protect the vulnerable. Two key differences between the circumstances in Cole and Childs are immediately apparent. First, Cole arose in a commercial setting and Childs in a purely social setting. Secondly, the plaintiff’s intoxication in Cole contributed to the accident and her injury, whereas the plaintiff in Childs was entirely blameless in the accident. Some might therefore conclude that Cole is easily distinguishable and not really relevant in determining how Childs would be decided in Australia. However, in our view, the reasoning in Cole provides an important indication of the High Court’s likely approach to a case like Childs. In both commercial and social settings, the majority is likely to highlight the misbehaviour of the intoxicated party, rather than the responsibility of the host. Tort’s primary function is often said to be to compensate, which many would argue it does not do particularly well. As compensation is based on proof of fault, it is, by definition, only available to a very limited number of accident victims.37 In its ‘defence’, some have asserted that tort serves, or has the potential to serve, additional important roles, such as educator and deterrent.38 We argue for the increased use of tort as a means for encouraging safety by imposing potential liability in circumstances like those in Childs (and Cole). If a duty were found, commercial and social hosts would properly fear the consequences of being held liable for not hosting a ‘safe’ party. This would highlight the host’s responsibility, based on control and actual or constructive knowledge of the risks associated with guests’ impairment through alcohol or drugs. This responsibility exists regardless of whether the injured plaintiff is also blameworthy. Despite the importance Kirby J attached to the commercial nature of the relationship in Cole, it is arguable that his reasoning, with its emphasis on standard-setting, deterrence and safety, could be used to support the imposition of a duty of care in a case like Childs. In turn, ultimate liability would be determined by breach evaluations, where the fact-finder can more appropriately engage in assessments of reasonableness.39 We endorse Kirby J’s rather acerbic suggestion that the majority’s determination in Cole reflects a ‘withered view of community and legal neighbourhood’.40 Courts 35 Ibid, at [91]. 36 Ibid. At [93], he stated: ‘There are many decisions . . . that support the general proposition that a person in control of licensed premises owes a duty of care in negligence to take reasonable precautions in the circumstances not to contribute to a danger to others.’ 37 H Luntz and D Hambly, Torts: Cases and Commentary, revised 5th ed, LexisNexis Butterworths, Sydney, 2006, p 54. 38 A Linden, Canadian Tort Law, 7th ed, LexisNexis Canada, Toronto, 2001; B Feldthusen, ‘The Canadian Experiment with the Civil Action for Sexual Battery’ in N Mullany (Ed), Torts in the Nineties, Law Book Company, Sydney, 1997, p 274. 39 With respect to breach arguments, Kirby J stated that the club should have prevented the supply of alcohol earlier, called the police if necessary, and arranged for her transport home. A trial judge in a case like Childs could engage in similar assessments of what would be reasonable in the circumstances (such as preventing the intoxicated guest from leaving or calling a taxi). 40 Cole (2004) 217 CLR 469; 207 ALR 52 at [93].



7

should not so readily allow responsibility to be abdicated, inappropriately elevating notions of autonomy, at the expense of enhanced community safety. Rather, they should be signalling to social and commercial hosts that their acts and omissions can have an enormous impact on the community. Before turning to the Canadian position on social host liability, we wish to point out one important contextual difference between Canada and Australia that may explain why this issue has arisen less frequently in Australia.41 Typically, motor vehicle third party liability insurance policies in Canada are subject to upper limits, whereas in Australia compulsory motor vehicle insurance against liability for personal injury extends to the full extent of the driver’s or owner’s liability. This may encourage plaintiffs in Canada more than Australia to seek other defendants (such as social hosts) to achieve full recovery. Nevertheless, in our view, the opportunity to set standards and increase safety by imposing liability on social hosts is still relevant in Australia. For example, in jurisdictions like Victoria, the Transport Accident Commission (TAC) may seek contribution from third parties such as social hosts.42 Although the TAC might choose not to do so in the case of uninsured social hosts, it could exercise its rights of subrogation even in those circumstances in order to promote its important safety objectives.

III Canada’s approach to social hosts A The role of social hosts and the facts in Childs Childs v Desormeaux, on a first reading, might seem a weak test case for social host liability. The Supreme Court explained that the ‘only alcohol served by the hosts was three-quarters of a bottle of champagne in small glasses at midnight’43 and emphasised ‘the absence of evidence that the hosts in this case in fact knew of Mr Desormeaux’s intoxication’.44 Yet the lower court decisions reveal a murkier factual picture on both counts. As the Supreme Court acknowledged, aside from the champagne just mentioned, this was a ‘BYOB’ event (which the Supreme Court and the Court of Appeal for Ontario translated as ‘Bring Your Own Booze’ and the Ontario Superior Court of Justice, rather more delicately, as ‘bring your own alcoholic drinks’).45 Weiler JA, who delivered the judgment of the Court of Appeal,46 identified as a ‘distinguishing feature of this case . . . that Courrier and Zimmerman did not supply or serve alcohol to their guests’, such that they did not actively contribute to the risk of Desormeaux causing injury on the road.47 In contrast, Chadwick J in the Superior Court observed that the ‘BYOB situation places more of a duty on the host because they do not know how much their guests 41 We thank Professor Harold Luntz for bringing this issue to our attention. 42 Transport Accident Act 1986 (Vic) Pt 6 Div 3. 43 Childs v Desormeaux [2006] 1 SCR 643 at [4]. See also Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [69]; Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [55]. 44 Childs v Desormeaux [2006] 1 SCR 643 at [30]. 45 Ibid, at [4]; Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [1]; Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [6]. 46 O’Connor ACJO and Sharpe JA agreeing. 47 Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [74] (see also [8]).



are consuming’.48 Moreover, certain evidence before Chadwick J suggested that someone at the party was serving alcohol brought by the guests.49 Whether guests served themselves alcohol is an important question. Even though the hosts supplied little alcohol, if they (or another guest acting on their behalf) served the available alcohol to others this could bring the circumstances closer to a commercial setting. Commercial establishments like bars and restaurants both sell and serve alcohol. The selling creates a ‘perverse incentive’50 to serve too much, because the host profits financially from the sale. The serving creates an opportunity to monitor consumption.51 Both factors may contribute to an expectation that commercial hosts accept some responsibility for their clients’ safety and conduct towards others. As for whether hosts Courrier and Zimmerman knew that Desormeaux was intoxicated, the Supreme Court’s assertion of an ‘absence of evidence’52 to this effect is perhaps inconsistent with Chadwick J’s findings in the Superior Court. In the Court of Appeal, Weiler JA insisted that ‘the trial judge did not find that the social hosts knew Desormeaux was impaired’.53 But Chadwick J specifically stated that he was ‘satisfied on all of the evidence [that] Desmond Desormeaux would be showing obvious signs of impairment when he left the party’.54 As discussed further below, Courrier and Zimmerman did not dispute that they had contact with Desormeaux just before he left.55 The logical conclusion (although not expressly articulated by Chadwick J) is that they would therefore have observed these signs of impairment. These obvious visible signs of Desormeaux’s impairment should be put into the following context. Courrier and Zimmerman were well aware of Desormeaux’s habit of excessive drinking; he had been an alcoholic for 20 years before the accident and was convicted of impaired driving in 1991 and 1994.56 Courrier, who described Desormeaux as a mere ‘acquaintance’, had lived with Desormeaux for two or three months in 1992 and agreed that he had seen Desormeaux two or three times a month for the last 20 years.57 Desormeaux described Courrier as his best friend, a characterisation of their relationship that Chadwick J accepted.58 Weiler JA discounted the significance of the hosts’ knowledge of Desormeaux’s habits, stating that ‘the social hosts’ knowledge of Desormeaux’s propensity to drink is but one factor to consider in determining whether the social hosts knew that Desormeaux was intoxicated when he left the party and [was] about to drive’.59 The Supreme Court adopted a similar line, as outlined further below. This differed somewhat from its approach to the knowledge of commercial hosts, who seem to be subject to additional obligations based in part on the expectation that 48 49 50 51 52 53 54 55 56 57 58 59

Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [97] (emphasis added). Ibid, at [28]–[29], [34], cf [44], [47]. Childs v Desormeaux [2006] 1 SCR 643 at [22]. Ibid, at [18]; Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [33]. Childs v Desormeaux [2006] 1 SCR 643 at [30] (see also [28]). Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [53] (see also [62], [64]). Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [73]. Ibid, at [35], [60]. Ibid, at [12], [15]. Ibid, at [10], [17]. Ibid, at [10]. Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [39].



9

they ‘possess special knowledge about intoxication’.60 The Supreme Court made little of the final contact between Courrier, Zimmerman and Desormeaux, but this contact may be more significant than the hosts’ knowledge of Desormeaux’s previous drink driving offences. Chadwick J recounted the following details based on the evidence of Maureen O’Brien, Desormeaux’s girlfriend, who was injured in the accident: When they were leaving Dwight Courrier asked Desmond Desormeaux ‘are you ok brother’ to which Desmond Desormeaux responded ‘no problem’. Dwight Courrier did not ask Desmond Desormeaux to stay overnight as he had done before nor did he offer to call a taxi.61

Chadwick J appeared to find O’Brien a credible witness,62 and Courrier apparently did not dispute this exchange with Desormeaux. Unless we assume that Courrier was merely asking a routine question as a responsible host, he presumably had some reason to question Desormeaux in this way, such as Desormeaux’s appearance upon leaving, his past conduct, or the amount he had drunk or the way he behaved at the party. This arguably conflicts with the Supreme Court’s insistence that the hosts did not and could not have been expected to know that Desormeaux was intoxicated. Chadwick J did not accept Zimmerman’s claim that she invited Desormeaux and his guests to stay overnight when she saw them just before they left,63 and the judge also apparently rejected Courrier’s suggestion that he asked them to do so.64 Further, Chadwick J found that ‘Zimmerman was more than pleased’ to see Desormeaux and his friends leave and that Courrier ‘deliberately did not pay any attention to how much . . . the Desormeaux group . . . consumed while on the premises’.65 This suggests a worrying indifference to the welfare of Desormeaux and his friends, in contrast to past occasions when Courrier and Zimmerman had invited Desormeaux to stay with them so that he did not have to drive.66 In comparing the factual analysis of the three courts, it is worth remembering that Chadwick J in the Superior Court was closest to the evidence and his factual findings should ordinarily stand. The Supreme Court’s rather different understanding of the facts in Childs v Desormeaux and its limited engagement with the factual details suggest a reluctance to liken social with commercial hosts and an expectation that guests under the influence in a social rather than commercial context should take care of themselves. In this regard, the court would presumably agree with Courrier’s assertion that ‘people are responsible for their own actions’.67 This conclusion provides a useful background for understanding the Supreme Court’s analysis of the duty of care issues in the case. 60 61 62 63 64 65 66 67

Childs v Desormeaux [2006] 1 SCR 643 at [18]. Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [35]. Ibid, at [37]. Ibid, at [60], [62] (see also [65], [67]). Ibid, at [65], [69]. Ibid, at [62], [67] (emphasis added). Ibid, at [23]. Ibid, at [64].



B No duty of care for social hosts? Overview As mentioned earlier, all three courts in Childs v Desormeaux rejected Childs’ claim against Courrier and Zimmerman. However, the reasons of Chadwick J in the Ontario Superior Court of Justice differed markedly from those of the higher courts. Chadwick J accepted that Childs’ injuries were reasonably foreseeable and that the relationship between Desormeaux and the hosts was sufficiently proximate to found a duty of care.68 On this basis, he assessed the hosts’ responsibility for the damage (or the accident, on these facts) at 15%.69 However, Chadwick J nevertheless concluded that ‘there is good policy reason not to expand tort law to include the social host’ and therefore dismissed Childs’ action.70 In contrast, the Supreme Court and the Court of Appeal for Ontario refused to find the requisite foreseeability and proximity and therefore did not need to examine countervailing policy considerations.71 While the trial judge viewed the facts as establishing a duty of care, but policy as closing off that possibility, the Supreme Court and Court of Appeal denied that the facts created a duty in this case, but left the door open to the imposition of such a duty in future. Below, we explain each step leading to the Supreme Court’s conclusion, as a basis for understanding the likely differences in the Australian approach. In the Supreme Court, McLachlin CJ explained that the general two-stage test in Canada for determining whether a duty of care exists72 derives from the decision of the House of Lords in Anns v Merton London Borough Council.73 Put simply, stage one of the Canadian test asks whether the parties are in a sufficiently close relationship or sufficiently proximate to justify a duty being imposed. In answering this question, Canadian courts will consider reasonable foreseeability and ‘factors going to the relationship between the parties’, as well as the similarity between the case at hand and previous cases in which a duty has arisen.74 If the answer at stage one is yes, a prima facie duty of care arises. In that case, stage two of the test applies, which asks whether ‘broader policy considerations’ negate the prima facie duty.75 The simplicity of the test as just stated masks the great difficulties involved in answering the two main questions. The caselaw is far from clear,76 as exemplified in Childs v Desormeaux itself. Although the three courts ostensibly applied this test, only the trial court reached the second stage of analysis. The two higher courts held 68 69 70 71 72 73 74 75 76

Ibid, at [94], [98], [104]. Ibid, at [105]. Ibid, at [138]. Childs v Desormeaux [2006] 1 SCR 643 at [26], [48]; Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [75], [77]. Childs v Desormeaux [2006] 1 SCR 643 at [11]. [1978] AC 728 at 742; [1977] 2 All ER 492. Childs v Desormeaux [2006] 1 SCR 643 at [12], [15] (referring to Odhavji Estate v Woodhouse [2003] 3 SCR 263 at [52]; City of Kamloops v Nielson [1984] 2 SCR 2; Cooper v Hobart [2001] 3 SCR 537). Childs v Desormeaux [2006] 1 SCR 643 at [12]. See, generally, Neyers and Gabie, above n 9.


Social hosts’ responsibility for their intoxicated guests 11

that no duty arose at the first stage and therefore deemed it unnecessary to proceed to the second stage.77 Novel duty In applying the general test to the circumstances of the present case, McLachlin CJ first considered ‘whether claims against private hosts for alcohol-related injuries caused by a guest constitute a new category of claim’.78 Like the two lower courts, the Supreme Court held that these claims fall within a different category from claims against commercial alcohol providers.79 Had they been part of the same category, it would have been easier to pass stage one of the test,80 because Canada has long held that commercial hosts owe a duty not only to their intoxicated clients,81 but also to third parties injured by these clients.82 McLachlin CJ provided three reasons for the conclusion that Childs was requesting the imposition of a novel duty of care, despite the apparent analogy with commercial hosts: First, commercial hosts enjoy an important advantage over social hosts in their capacity to monitor alcohol consumption. . . . Second, the sale and consumption of alcohol is strictly regulated by legislatures, and the rules applying to commercial establishments suggest that they operate in a very different context than private-party hosts. . . . Third, the contractual nature of the relationship between a tavern keeper serving alcohol and a patron consuming it is fundamentally different from the range of different social relationships that can characterize private parties in the non-commercial context.83

These three factors also influenced the court’s subsequent analysis of social hosts under stage one of the test, as discussed further below. The ability of social hosts to monitor alcohol consumption would seem to depend on the context, but the Supreme Court maintained that commercial hosts have an incentive to monitor and their patrons expect them to do so.84 Most of the court’s reasoning focused on the second factor. It reiterated that, unlike commercial hosts, social hosts are not expected to enforce limits on alcohol consumption.85 However, this assertion (even if correct) seems circular: social hosts are not subject to a duty of care because they are not expected to take responsibility for guests’ drinking. Although it did not say so, perhaps the court also had in mind the view expressed by Chadwick J that ‘it should be left to the legislature to determine a social host[s] liability and also to properly compensate the innocent victims’.86 In identifying the ‘contractual nature’ of the relationship between 77 Childs v Desormeaux [2006] 1 SCR 643 at [14]. 78 Ibid, at [15]. 79 Ibid; Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [33]; Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [76]. 80 Childs v Desormeaux [2006] 1 SCR 643 at [15]. 81 Jordan House Ltd v Menow [1974] SCR 239; Crocker v Sundance Northwest Resort Ltd [1988] 1 SCR 1186. 82 Stewart v Pettie [1995] 1 SCR 131. See also Chamberlain, above n 11, at 105–8. 83 Childs v Desormeaux [2006] 1 SCR 643 at [18], [19], [22]. 84 Ibid, at [18]. 85 Ibid, at [21]. 86 Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [138].



commercial host and client, the Supreme Court was primarily concerned with the profit-making incentive to oversupply,87 as already mentioned. However, according to Adjin-Tettey: There is no justification for making profitability the distinguishing factor between commercial and social host liability. . . . Profitability should inform the standard of care expected in a particular case and not the existence of a duty.88

While this may not be consistent with the views of Kirby J in Cole89 (for example), it has some merit. Arguably, profitability more appropriately influences standard of care and breach issues, such as the practicability of precautions to be undertaken by the reasonable defendant in response to a foreseeable risk of injury, in accordance with the Wyong ‘approach’.90 We further contend that, to avoid uncertainty, courts should examine the extent to which social hosts supply alcohol, monitor alcohol consumption or contribute to intoxication in determining whether they have breached their duty of care rather than in assessing whether a duty exists in the first place.91 We return to this issue below. Foreseeability Having held that imposing a duty of care on social hosts would be more than a mere extension of the existing duty on commercial hosts, the Supreme Court proceeded to apply stage one of the Anns test to the facts at issue in order to decide whether Courrier and Zimmerman had a duty of care towards Childs. In surprisingly brief reasoning, the court found that Childs’ injury was not reasonably foreseeable.92 McLachlin CJ disagreed with the trial judge’s suggestion that Desormeaux’s history of drink driving affected the foreseeability of the injury, stating, ‘[t]he inferential chain from drinking and driving in the past to reasonable foreseeability that this will happen again is too weak to support the legal conclusion of reasonable foreseeability’.93 As highlighted earlier, the Chief Justice also refused to accept that Desormeaux was exhibiting signs of intoxication when he was seen by Courrier and Zimmerman upon leaving. Ironically, McLachlin CJ justified this stance on the basis that Chadwick J declined to make an explicit finding to this effect and that it was not for the Supreme Court to interfere with the factual basis of the case.94 Nonfeasance Although the Supreme Court could have stopped here, it chose instead to explain its view that, even if the injury had been reasonably foreseeable, ‘no duty would arise because the wrong alleged is a failure to act or nonfeasance 87 Childs v Desormeaux [2006] 1 SCR 643 at [22]. 88 E Adjin-Tettey, ‘Social Host Liability: A Logical Extension of Commercial Host Liability?’ (2002) 65 Saskatchewan L Rev 515 at 522, 524. 89 Cole (2004) 217 CLR 469; 207 ALR 52 at [91]. 90 Wyong Shire Council v Shirt (1980) 146 CLR 40; 29 ALR 217. 91 For a similar view, see F Kelly, ‘Before you host a party read this: Social host liability and the decision in Childs v Desormeaux’ (2006) 39 UBC L Rev 371. 92 Childs v Desormeaux [2006] 1 SCR 643 at [30]. 93 Ibid, at [29]. 94 Ibid, at [30].



13

in circumstances where there was no positive duty to act’.95 At the outset, it is worth noting that the characterisation of Childs v Desormeaux as a case of nonfeasance is itself debatable. The Supreme Court regarded Childs’ case as being that Courrier and Zimmerman should have ‘interfered with the autonomy of Mr Desormeaux by preventing him from drinking and driving’.96 Yet one could equally describe the hosts as engaging in positive acts: not only organising the BYOB event (as the court acknowledged),97 but also inviting Desormeaux, providing alcohol and possibly having alcohol served to guests, and asking if Desormeaux was OK upon departure.98 These actions provide highly relevant context for the alleged failure to prevent Desormeaux from driving. In other words, this looks more like an omission in the course of positive conduct than a pure omission. Or, as Neyers and Gabie contend, ‘Childs is a case of “pseudo-nonfeasance” (inaction after the innocent creation of a risk) rather than a case of “pure nonfeasance” (no risk creating activity)’.99 The Supreme Court identified three situations that bring ‘legal strangers into proximity’ and impose on defendants a positive duty to act.100 First, where a defendant deliberately invites third parties to participate in an activity involving an inherent risk that the defendant has created or controls (such as ‘a dangerous inner-tube sliding competition’).101 The court correctly noted that the BYOB party of Courrier and Zimmerman was not such an activity.102 Secondly, where the parties are in a paternalistic relationship such as parent and child.103 Again, despite the history of Desormeaux and the hosts, the court correctly concluded that they were not in such a relationship.104 Thirdly, where defendants exercise a public function or engage in commercial enterprises including public responsibilities.105 By definition, social hosts do not perform this kind of role.106 The Supreme Court emphasised that these three situations are not necessarily exhaustive; rather, they involve particular features that may create a positive duty to act.107 Therefore, the Supreme Court went on to consider whether social hosts might have a positive duty to act even though they fall outside these three situations. McLachlin CJ pointed out that the three situations have in common three factors: creation or control of risk by the defendant, the autonomy of competent people to engage in risky behaviour and reasonable reliance by those people on the defendant.108 The Chief Justice concluded that these factors do not suggest that social hosts have a positive 95 96 97 98 99 100 101 102 103 104 105 106 107 108

Ibid, at [26] (see also [32], [47]). Ibid, at [32]. Ibid, at [33]. Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [35]. Neyers and Gabie, above n 9, at 315. Childs v Desormeaux [2006] 1 SCR 643 at [34]. Ibid, at [35], [42]. Ibid, at [42]. Ibid, at [36]. Ibid, at [42]. Ibid, at [37]. Ibid, at [42]. Ibid, at [34]. Ibid, at [38]–[40].



duty to act.109 Holding a BYOB party is not inherently risky, a partygoer ‘does not park his autonomy at the door’ and guests do not expect their hosts to monitor their alcohol consumption.110 McLachlin CJ concluded that ‘hosting a party at which alcohol is served does not, without more, establish the degree of proximity required to give rise to a duty of care on the hosts to third-party highway users who may be injured by an intoxicated guest’.111 The court indicated that liability might be imposed in particular factual circumstances, namely where ‘the host’s conduct implicates him or her in the creation or exacerbation of the risk’,112 which might occur where ‘a host . . . continues to serve alcohol to a visibly inebriated person knowing that he or she will be driving home’.113 The Court of Appeal too suggested that liability could arise where ‘a social host knew that an intoxicated guest was going to drive a car and did nothing to protect the innocent third party users of the road’.114 Policy Although the Supreme Court found it unnecessary to apply stage two of the Anns test (because it found no prima facie duty under stage one),115 its views on the policy issues surrounding social host liability are reflected in its reasoning. In particular, McLachlin CJ pointed out that ‘the common law is a jealous guardian of individual autonomy’,116 explaining the court’s expectation that guests take responsibility for themselves and its reluctance to impose a duty of care on social hosts in most circumstances. The Ontario Superior Court of Justice articulated other policy considerations weighing against the imposition of a duty. These include the burden this would place on social hosts due to the difficulty of controlling guests’ drinking,117 the likely absence of insurance for social as compared with commercial hosts for this kind of liability,118 and the potential for the legislature to regulate such matters instead of the courts,119 not to mention the numerous actions that could flow from such a decision. On the other hand, the fact that the Supreme Court left open the possibility of imposing a duty on social hosts in appropriate circumstances indicates that other considerations besides individual autonomy are at play. Indeed, the Court of Appeal suggested that a duty might be desirable to deter drink driving. Weiler JA took note of evidence from the intervenor, Mothers Against 109 110 111 112 113 114 115 116 117

Ibid, at [41], [43]. Ibid, at [44]–[46]. Ibid, at [47]. Ibid. Ibid, at [44]. Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [90] (see also [10], [76]). Childs v Desormeaux [2006] 1 SCR 643 at [48]. Ibid, at [31]. See also Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [78]. Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [112]–[114], [121]. See also Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [83]. 118 Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [136]. See also Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [79]–[80]. 119 Childs v Desormeaux (2002) 217 DLR (4th) 217 (Ont SCJ) at [135]–[138]; cf Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [86]–[89].



15

Drunk Driving Canada,120 highlighting the effectiveness of education, new criminal laws and civil liability in reducing drunk driving.121 Adjin-Tettey also advocates social host liability as a means of promoting ‘socially responsible behaviour on the part of hosts’.122

IV Childs in Australia A Introduction In this section, we explain the current approach of Australian courts to establishing a duty of care in a novel context and predict how the High Court of Australia would apply that approach to the factual circumstances in Childs. Unlike the Supreme Court of Canada, the High Court of Australia has refused to follow Anns v Merton LBC123 since the judgments of Mason CJ, Deane and Brennan JJ in Sutherland Shire Council v Heyman.124 Further, the High Court no longer subscribes to the framework propounded by Deane J125 and endorsed by a majority of the court in a series of cases,126 having discarded it, rather unceremoniously, in Sullivan v Moody.127 The court made clear that Anns’ more recent incarnation as a three-stage test in Caparo Industries Plc v Dickman128 is not to be followed in Australia. If a case like Childs arose in Australia, courts would instead employ the ‘no formula’ approach,129 with a focus on ‘salient features’, to assess whether a duty was owed. It is likely that the use of the salient features approach to determine the duty question in Childs would yield the same result in Australia as that reached in Canada.130 However, the vagaries of this recently endorsed approach could perhaps lead to a different result, because many of the relevant features are undeniably value-laden. Moreover, each feature’s relative weight vis-a`-vis others cannot be determined (let alone predicted) with precision. Evaluating the relative weight of the features in analysing the totality of the relationship may not lead all assessors — such as judges and commentators — to the same end result.131 120 See Childs v Desormeaux (2003) 231 DLR (4th) 311 for the motion to intervene in the Ont CA. 121 Childs v Desormeaux (2004) 239 DLR (4th) 61 (Ont CA) at [82], [88]. 122 Adjin-Tettey, above n 88, at 546. 123 [1978] AC 728; [1977] 2 All ER 492. 124 (1985) 157 CLR 424; 60 ALR 1. 125 Jaensch v Coffey (1984) 155 CLR 549; 54 ALR 417; Sutherland Shire Council v Heyman (1985) 157 CLR 424; 60 ALR 1; Stevens v Brodribb Sawmilling Co Pty Ltd (1986) 160 CLR 16; 63 ALR 513. 126 Cook v Cook (1986) 162 CLR 376; 68 ALR 353; Burnie Port Authority v General Jones Pty Ltd (1994) 179 CLR 520; 120 ALR 42. 127 (2001) 207 CLR 562; 183 ALR 404. 128 [1990] 2 AC 605; [1990] 1 All ER 568. With respect to policy, the default position in Anns and Caparo differs: in terms of which party must prove the dictates of policy, it favours plaintiffs in Anns and defendants in Caparo. See Lord Hoffmann in Stovin v Wise [1996] AC 923; [1996] 3 All ER 801. 129 By definition this seems to be an approach rather than a ‘framework’. For a detailed discussion of this approach, see section IVD, below. 130 Arguably, concerns of the judiciary and commentators with respect to the appropriate methodology to be used to assess duty questions are overwrought. 131 The NSW Court of Appeal considered North American authorities including Childs in obiter



B Defining the duty Just as in Canada, in Australia, if a binding precedent already governs the type of case under scrutiny, the duty question is considered settled.132 Accordingly, as a first step in determining whether social hosts owe a duty to road users in circumstances like Childs, an Australian court would have to decide whether this is a novel kind of case or a set of circumstances that falls within an existing duty. The ‘scope’ of a given duty has taken on increased significance in recent years.133 Most clearly illustrated in Modbury,134 the fundamental problem involves identifying the relevant duty, to ensure that a duty holding is not ‘piggy-backed’ on a given relationship that is not in fact applicable in the circumstances of the case. For example, while a host may owe a duty of care to a guest in some respects — for example, to prevent the guest from slipping on a sticky floor, in the well-recognised duty category of occupier to entrant — this is not necessarily sufficient to found a duty of care to prevent injury when a guest drinks alcohol in one’s home, leaves the premises, and is subsequently injured by a motor vehicle or injures someone else. Should the court ask whether an occupier owes a duty to exercise reasonable care so as to prevent all harm caused to or by an entrant as a result of activities that take place on the host’s premises, or should it ask whether a social host owes a duty to exercise reasonable care so as to prevent harm from a guest driving home from a party when obviously drunk? In assessing this situation, the decision regarding how to frame the relevant duty and how much factual specificity to include in doing so135 can sometimes have a profound effect on the result, as indicated by the reasoning of the majority and dissenting judges in Cole. In that case, Gleeson CJ asserted that the duty question should not be cast at too high a level of abstraction.136 Thus,

132

133 134 135

136

dicta in Parissis v Bourke [2004] NSWCA 373 (unreported, 23 November 2004, BC200408095), SLR [2005] HCATrans 672 (2 September 2005). The court held that householders in this case did not owe a duty of care to a guest who suffered burns at a barbecue party. Tobias JA nevertheless recognised that social host liability ‘is both complex and unresolved in both the United States and Canada’ (at [9]). Bryson JA stated: ‘While I acknowledge and appreciate counsel’s industry and assistance, I do not finally find any persuasive material in the extensive references to North American law’ (at [64]). Nevertheless, an ultimate appellate court like the High Court could revisit or overrule a principle it had laid down earlier. For example, in D’Orta-Ekenaike v Victoria Legal Aid (2005) 223 CLR 1; 214 ALR 92, the High Court recently had occasion to reconsider the principle of advocates’ immunity applied in Giannarelli v Wraith (1988) 165 CLR 543; 81 ALR 417, though it chose not to depart from it. See P Cane, ‘The New Face of Advocates’ Immunity’ (2005) 13 TLJ 93; cf R Tobin, ‘An uncommon common law: Barristers’ immunity in New Zealand’ (2006) 14 TLJ 224. Cole (2004) 217 CLR 469; 207 ALR 52. See also Gummow J in Vairy v Wyong Shire Council (2005) 223 CLR 422; 221 ALR 711 at [77], where he utilised a scope of duty approach (without deciding on that basis). Modbury Triangle Shopping Centre Pty Ltd v Anzil (2000) 205 CLR 254; 176 ALR 411. See Cole (2004) 217 CLR 469; 207 ALR 52. Appeal courts have expanded their reach by adding in facts and considerations that were traditionally relegated to fact-finders’ evaluations of standard of care and breach. Put simply, questions that would ordinarily be characterised as factual have become legal: cf Neindorf v Junkovic (2005) 222 ALR 631 at [55] per Kirby J (dissenting). (2004) 217 CLR 469; 207 ALR 52 at [1].



17

he declined to formulate any general proposition regarding the circumstances in which a supplier of alcohol in either a commercial or social setting might be under a duty to exercise care to protect a guest from physical injury due to alcohol consumption.137 Instead, he maintained, ‘[t]he question is whether there was such a duty in the circumstances of this case’;138 his answer was no.139 Unfortunately, this fact-specific approach is unhelpful in providing a principle to govern future conduct. By way of contrast, in his dissent, McHugh J formulated the duty question in more general terms. He contended that what was at issue was the duty owed by an occupier,140 which ‘is not confined to protecting entrants against injury from static defects in the premises’, but ‘extends to the protection of injury from all the activities on the premises’.141 The duty includes an affirmative obligation to monitor the behaviour of customers on the premises and to prevent harm. Accordingly, the club owed the plaintiff a duty to take care not to expose her to risk of injury. In this way, factually disparate scenarios can fall within the broadly cast duty. Rather than precluding a duty altogether, the customer’s conduct went to questions of contributory negligence.142 His Honour’s reasoning seems equally apt in a case like Childs, assuming the element of control is met. Here, it would be reasonable to ask whether a social host owes a duty to take reasonable care to prevent a guest from causing harm. However, the likelihood of applying such an approach would depend on the composition of the bench and the particular judge’s views on the level of abstraction that should be used when framing the question.

C Reasonable foreseeability While foreseeability must be satisfied as a threshold matter, in Australia it has been universally held to be insufficient to found a duty of care because it is too undemanding. Without additional constraints, it would allow for the imposition of a duty of care in an inordinately large number of circumstances, frequently unwarranted.143 While the defendant must reasonably foresee a general kind of harm to a class of persons, it need not foresee the precise sequence of events.144 The foreseeability required is now described as a ‘real 137 Ibid, at [9]. 138 Ibid. 139 Ibid, at [18]. He stated that ‘the appellant cannot succeed in this case unless there is a general duty upon a supplier of alcohol, at least in a commercial setting, to take such care. I do not accept that there is such a general duty’ (at [17]). 140 Ibid, at [30]. 141 Ibid, at [31]. This ‘general’ formulation is consistent with his approach in Jones v Bartlett (2000) 205 CLR 166; 176 ALR 137 and Vairy v Wyong Shire Council (2005) 223 CLR 422; 221 ALR 711 at [20]. See also Kirby J in Neindorf v Junkovic (2005) 222 ALR 631 at [50]; C Coventry, ‘You Had Better Watch Out: Liability of Public Authorities for Obvious Hazards in Footpaths’ (2006) 14 TLJ 81. 142 Cole (2004) 217 CLR 469; 207 ALR 52 at [32]–[34]. 143 Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404; Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337. The duty question in Graham concerned the authorities (local and state government); the suppliers clearly owed a duty of care in this case, though the majority held that they were not in breach. 144 Chapman v Hearse (1961) 106 CLR 112; [1962] ALR 379.



and not far-fetched possibility’.145 To strengthen its content, in Tame v New South Wales; Annetts v Australian Stations Pty Ltd,146 McHugh J took pains to highlight the reasonableness (and policy-oriented) aspect of the foreseeability inquiry. The assertion that, on its own, reasonable foreseeability is not sufficient to justify imposing a duty in novel (and sometimes controversial) circumstances is well-rehearsed. Clearly, the determination of what else is necessary is more taxing. Although the High Court offered differing views for a period of time, a consensus has emerged among a majority of the court, alluded to earlier:147 ‘no single formula’ governs all cases; Deane J’s proximity ‘formula’, which commanded the attention of the Australian courts for over 10 years, is considered unhelpful and should not be followed; further, Caparo’s148 three-stage test (in some respects, not unlike Anns v Merton LBC, with an opportunity to canvass policy considerations openly) should not be followed. Caparo’s third element — ‘is it fair, just and reasonable’ to impose a duty? — has been castigated by the High Court as unhelpful and too closely tied to the vagaries of individual judges’ personal values and morality.149 Even Kirby J, in Graham Barclay Oysters Pty Ltd v Ryan, reluctantly agreed to relinquish his advocacy of the Caparo approach,150 although in Woolcock Street Investments Pty Ltd v CDG Pty Ltd he again referred to the Caparo test with decided affection, hinting at a possible future re-endorsement.151 In any event, if a case like Childs arose in Australia, the threshold foreseeability requirement would be readily satisfied. It is clearly a real and not far-fetched possibility that guests consuming alcohol at a party might leave and injure themselves or others on the road as a result of their impaired driving. The fact that the Supreme Court of Canada did not find this element satisfied is, to be blunt, rather astonishing, given that foreseeability in this context requires so little imagination.

D A ‘multi-faceted inquiry’: salient features Background to the salient features approach Certain scenarios have long troubled the common law with respect to duties of care, including cases involving pure omissions, the conduct of others, mental harm, pure economic loss and statutory authorities failing to exercise powers to act. As a consequence, Australian courts have offered a wide array 145 Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404. 146 (2002) 211 CLR 317; 191 ALR 449. 147 Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404; Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337; Woolcock Street Investments Pty Ltd v CDG Pty Ltd (2004) 216 CLR 515; 205 ALR 522. 148 Caparo Industries Plc v Dickman [1990] 2 AC 605; [1990] 1 All ER 568. 149 The Caparo test is said to be insufficiently precise and too discretionary. Its third limb states that ‘the situation should be one in which the court considers it fair, just and reasonable that the law should impose a duty of a given scope upon the one party for the benefit of the other’: Caparo Industries Plc v Dickman [1990] 2 AC 605 at 617; [1990] 1 All ER 568. 150 Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337 at [238]. 151 Kirby J, dissenting in Woolcock Street Investments Pty Ltd v CDG Pty Ltd (2004) 216 CLR 515; 205 ALR 522, stated: Caparo ‘obliges a transparent consideration of the issues of legal policy’ (at [158]). While noting that he had been compelled by High Court decisions to abandon Caparo ‘for the time being’ (ibid), he added that it is still used elsewhere.



19

of sometimes perplexing reasoning in their attempts to provide workable frameworks that can readily be applied to later circumstances. In a novel kind of case, the ‘salient features’ approach is meant to evaluate the totality of the relationship between the plaintiff and defendant. It is a ‘multi-faceted inquiry’ to determine if a duty of care is owed.152 The court must explore and weigh the features that are significant to the type of case at hand. Some may point towards the existence of a duty of care, others away. An assessment of the duty of care question using these features is said to be based on principle and doctrine,153 as opposed to mere judicial whim (supposedly the policy considerations in Caparo). Some features are particularly relevant to certain kinds of cases. For instance, indeterminacy is a particularly relevant factor in cases involving pure economic loss. It could be argued that this approach unpacks and makes transparent what was incorporated in the notion of ‘proximity’. For example, it could be used to re-evaluate a case like Sutherland Shire Council v Heyman,154 where it was held (in Deane J’s view) that there was no duty of care because of the absence of proximity. Using the methodology of the modern High Court, the fact that the Heymans did not obtain a certificate from the local authority could be viewed as the absence of demonstrated vulnerability, a feature salient to the duty analysis. This approach is arguably preferable to an over-reliance on ‘proximity’, which on occasion tended to obscure rather than elucidate the court’s reasoning. As was the case in Sutherland Shire Council v Heyman, the use of proximity (in that case, ‘causal proximity’) had to be unravelled and explained by examining factors such as ‘reliance’ and ‘assumption of responsibility’. Today, these factors would be examined as ‘salient features’, without having to squeeze them (sometimes rather unnaturally) into the proximity framework. Some judges advocate an ‘incremental approach’ to deciding a ‘novel kind of case’. The way in which this approach operates may be illustrated by the well-known case of Lowns v Woods,155 in which a doctor was held to owe a duty of care to a non-patient in the doctor’s professional setting (albeit on the basis of Deane J’s then-binding proximity framework). Whether such a duty extends to a social setting could be assessed incrementally, evaluating the principle and reasoning from Lowns v Woods in the light of the changed context — for example, moving from the doctor’s office to a restaurant.156 An incremental assessment would necessarily involve an explicit or implicit examination of the salient features of the situation, such as vulnerability, control, indeterminacy and defensive practices. Therefore, it may be difficult to distinguish an incremental approach from a ‘salient features’ approach.157 152 Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337 at [149] per Gummow and Hayne JJ. 153 See Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404. Although the High Court did not use the term ‘salient features’, this phrase encapsulates the way the court canvassed a series of factors essential to its determination. 154 (1985) 157 CLR 424; 60 ALR 1. 155 (1996) Aust Torts Reps 81-376 (NSW CA). 156 Not unlike moving from the context of commercial enterprise to that of social host. 157 In Sullivan, the incremental movement was used to assess a novel duty claim by reference



Moreover, the two methods may well lead to the same result, as in Perre v Apand Pty Ltd.158 Omissions, knowledge and control Although the fundamental approach to determining whether a duty is owed is markedly different in Australia and Canada, in both jurisdictions the context of omissions necessitates close scrutiny as to whether a duty of care is owed. Like those in Canada, Australian courts are reluctant to impose duties of care in the context of pure omissions. In his encapsulation of the principle, Brennan J asserted in Sutherland Shire Council v Heyman159 that ‘a man on the beach is not legally bound to plunge into the sea when he can foresee that a swimmer might drown’.160 There are, of course, exceptions. Brennan J noted: Some broader foundation than mere foreseeability must appear before a common law duty to act arises. There must also be either the undertaking of some task which leads another to rely on it being performed, or the ownership, occupation or use of land or chattels to found the duty.161

While Brennan J suggested what must be present to find a duty to act affirmatively, he did not set out the criteria in terms quite as categorical as in Childs. Nevertheless, cases involving alleged nonfeasance present Australian courts with the same underlying concerns highlighted by the Supreme Court of Canada in Childs, related in part to the (purported) need to protect a potential defendant’s autonomy and issues of knowledge and control.162 The defendant’s knowledge and awareness of risk appear to be increasingly relevant factors when determining whether or not a duty of care should be imposed in novel circumstances.163 Also of considerable importance are the defendant’s control over the risk and the particularity of that risk.164 This helps explain the result in Pyrenees Shire Council v Day, where a duty of care was

158

159 160 161 162 163

164

to features that pointed away from that duty, such as coherence of the law and indeterminacy: Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404. (1999) 198 CLR 180; 164 ALR 606. Although Gummow J employed a ‘salient features’ approach and McHugh J proceeded ‘incrementally’, both examined the same factors, including vulnerability, knowledge and indeterminacy, to determine whether a duty of care was owed in this pure economic loss context. (1985) 157 CLR 424; 60 ALR 1. Ibid, at CLR 477. Ibid (referring to Windeyer J in Hargrave v Goldman (1963) 110 CLR 40 at 66; [1964] ALR 377). See also Stovin v Wise [1996] AC 923; [1996] 3 All ER 801, where Lord Hoffmann discussed political, moral and economic reasons why duties to act affirmatively are rarely imposed. The absence of this feature was significant to the court’s reasoning in Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337, where no duty was placed on the state and the local authority; by comparison, its presence in Crimmins v Stevedoring Industry Finance Committee (1999) 200 CLR 1; 167 ALR 1 and Pyrenees Shire Council v Day (1998) 192 CLR 330; 151 ALR 147 was particularly important to the court’s duty holdings. The ‘fragmented’ nature of the defendants’ alleged control proved fatal as a salient feature (among others) in Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337 (with respect to the state and the local authority) and Agar v Hyde (2000) 201 CLR 552;



21

held to exist, in contrast to Graham Barclay Oysters Pty Ltd v Ryan,165 where it was not. The particularity of the defendant’s knowledge of the risk has had an important effect on the results in several cases involving omissions (albeit in contexts involving statutory authorities).166 In a case like Childs, the court would have regard to the hosts’ knowledge. The effect of this factor would depend on the significance attached to the pre-existing relationship of the hosts and guest: that is, how much importance would be given to their history and the knowledge the hosts had of the guest and his or her proclivity to drink too much? How much actual knowledge of the guest’s drinking at that particular time did the hosts have? How much ought the hosts to have known with respect to potential risks, based on the guest’s condition? The guest’s proclivity to drink to excess and drive, highlighted by a criminal record for impaired driving, would weigh in favour of holding that a duty exists. On the other hand, if the prior knowledge was downplayed, as it was in Childs in the Supreme Court of Canada, this would favour the hosts. Also of particular significance is the defendant’s control of the risk. The general rule that no duty of care arises to prevent a third party from harming another person stems partly from the fundamental principle, noted above, that the common law does not ordinarily impose liability for omissions. The defendant’s lack of control over the conduct of a third party was particularly important in Modbury Triangle Shopping Centre Pty Ltd v Anzil.167 In that case, the plaintiff (an employee of one of the defendant’s tenants, a video store) was criminally assaulted by third party assailants, while in a car park owned by the defendant. The court held that there was no relevant duty of care because the defendant had no control over the assailants’ behaviour and no knowledge of what they planned to do. Gleeson CJ applied Dixon J’s oft-cited comment from Smith v Leurs168 that it is exceptional for there to be a duty on a person to control the conduct of another; for a duty to exist, it must be on the basis of control and a special relationship.169 Although one can sometimes look to reliance to help establish the duty, no relevant reliance was held to be present in Modbury. Further, the defendant did not assume responsibility for the security of persons in the plaintiff’s position: its capacity to decide when

165 166 167 168 169

173 ALR 665; on the other hand, it was a particularly weighty factor supporting the imposition of a duty of care in Crimmins v Stevedoring Industry Finance Committee (1999) 200 CLR 1; 167 ALR 1. (2002) 211 CLR 540; 194 ALR 337. Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337; Crimmins v Stevedoring Industry Finance Committee (1999) 200 CLR 1; 167 ALR 1; Pyrenees Shire Council v Day (1998) 192 CLR 330; 151 ALR 147. (2000) 205 CLR 254; 176 ALR 411. This case is discussed extensively in J Dietrich, ‘Liability in negligence for harm resulting from third parties’ criminal acts: Modbury Triangle Shopping Centre Pty Ltd v Anzil’ (2001) 9 TLJ 152. Smith v Leurs (1945) 70 CLR 256; [1945] ALR 392. This was established, rather exceptionally, in Dorset Yacht Co Ltd v Home Offıce [1970] AC 1004; [1970] 2 All ER 294, but held not to exist in Hill v Chief Constable of West Yorkshire [1989] AC 53; [1988] 2 All ER 238, which distinguished Dorset Yacht. Both the High Court in Modbury and the Supreme Court in Childs held that the requisite control and special relationship were not present.



the car park would be lit was not the same as a responsibility to protect persons by doing so. The fact that little alcohol was actually provided by the hosts in Childs would weigh against a duty holding, falling short of satisfying the control feature.170 Similarly, reliance would in all likelihood be absent, given that the hosts did not appear to be monitoring their guests’ consumption and that Desormeaux did not appear to rely on them to play such a role. The control and knowledge regarding the condition and state of the premises, present in a typical occupiers’ liability case, were not present in Modbury, as that case involved strangers’ criminal behaviour. The unpredictability of criminal behaviour is one reason why, as a general rule, courts do not impose a duty to prevent harm from the criminal conduct of a third party, even if the risk of such harm is foreseeable. Aside from special relationships, such as employer and employee171 or school and pupil, the predictability of the criminal behaviour would have to exist at a heightened level to impose a duty of care. This was not satisfied in Modbury. Arguably, because of the history of the hosts and guest in Childs, this special element was present. However, this is again debatable, depending on the weight and importance attached to their history. While this could favour the plaintiff on the facts in Childs, it is unlikely to be decisive, especially in the light of the recent general trend of High Court decision-making in the sphere of negligence.172 Vulnerability The plaintiff’s vulnerability, or inability to protect her or himself, has emerged in the Australian caselaw as a factor of great importance. As a salient feature, vulnerability certainly favours the plaintiff in Childs-like circumstances: road users can do little to avoid the harm caused by impaired drivers. They are ‘vulnerable’, in both the commonplace and modern legal sense of the word. While the guest (impaired driver) might be the person best able to ‘control’ the situation so as to avoid this harm, if he or she is in a state of intoxication this may not be possible. The host, on the other hand, may be in a position to prevent the harm. In Cole173 vulnerability was missing. Throughout the case, the majority characterised the impaired plaintiff-pedestrian as the catalyst of her own harm or author of her own misfortune, quite unlike the blameless passenger Childs. In relative terms, the plaintiff in Childs more closely resembles the plaintiffs in Hill v Van Erp,174 Perre v Apand Pty Ltd,175 Crimmins v Stevedoring 170 Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337 (concerning the state and local authority as defendants); Crimmins v Stevedoring Industry Finance Committee (1999) 200 CLR 1; 167 ALR 1; Pyrenees Shire Council v Day (1998) 192 CLR 330; 151 ALR 147. 171 For example, Chomentowski v Red Garter Restaurant Ltd (1970) 92 WN (NSW) 1070; English v Rogers (2005) Aust Torts Reps 81-800 (NSW CA). See also the duty of care on prison authorities to protect inmates from attacks by fellow-inmates, as in New South Wales v Bujdoso (2005) 222 ALR 663. 172 See text above nn 25–27. 173 (2004) 217 CLR 469; 207 ALR 52. 174 (1997) 188 CLR 159; 142 ALR 687. 175 (1999) 198 CLR 180; 164 ALR 606.



23

Industry Finance Committee176 and Pyrenees Shire Council v Day,177 where the claimants were unable to do anything to avoid the harm that befell them, as opposed to the unsuccessful plaintiffs in Esanda Finance Corporation Ltd v Peat Marwick Hungerfords (Reg)178 and Woolcock Street Investments Pty Ltd v CDG Pty Ltd.179 In the latter cases, the commercially experienced and sophisticated plaintiffs were characterised as lacking vulnerability because they could have taken measures to protect their interests. Policy Although policy is not, strictly speaking, a factor that is meant to be used in determining duty questions,180 several considerations, such as the indeterminacy of the class181 potentially affected by the defendant’s conduct (as opposed to its ascertainability)182 sound remarkably like policy considerations. Similarly, the potential (and purported) effect on the defendant’s autonomy and how her or his freedom would be constrained seem more akin to policy than principle, yet these are well-accepted considerations, even outside Caparo’s and Deane J’s respective frameworks. These features seem to be as value-laden as any ‘fair, just and reasonable’ assessment could be, no less connected to the beliefs of any one judge than any other matter raising policy concerns. The High Court has attempted to differentiate determinations based on public policy (which it considers unacceptable) from those grounded in legal policy, which are said to be principle (and therefore appropriate influences). One could well argue, though, that, just like public policy considerations, those founded on legal policy are based on personal views of community mores and values, such as the effects a duty holding would have on the administration of justice.183 A significant policy-oriented salient feature that has been considered in several cases, and which could be said to be relevant to social host liability, is that the imposition of a duty of care would lead potential defendants to engage in deleterious defensive practices.184 Even in cases where this feature has justified in part the refusal to impose a duty of care, such as Hill v Chief 176 177 178 179

180 181 182 183

184

(1999) 200 CLR 1; 167 ALR 1. (1998) 192 CLR 330; 151 ALR 147. (1997) 188 CLR 241; 142 ALR 750. (2004) 216 CLR 515; 205 ALR 522. With regard to the plaintiff’s supposed ability to protect itself, see B Codd, R Hinchy and V Nase, ‘An Alternative View of Woolcock Street Investments v CDG Pty Ltd’ (2004) 12 TLJ 194. Cf Sutherland Shire Council v Heyman (1985) 157 CLR 424; 60 ALR 1. According to the High Court in Sullivan v Moody (2001) 207 CLR 562 at 579; 183 ALR 404. See, eg, Hill v Chief Constable of West Yorkshire [1989] AC 53; [1988] 2 All ER 238. With respect to indeterminacy, see J Beach, ‘Indeterminacy: The Uncertainty Principle of Negligence’ (2005) 13 TLJ 129. Perre v Apand Pty Ltd (1999) 198 CLR 180; 164 ALR 606. Giannarelli v Wraith (1988) 165 CLR 543; 81 ALR 417; D’Orta-Ekenaike v Victoria Legal Aid (2005) 223 CLR 1; 214 ALR 92. On the similarities between principle and policy, see J Stapleton, ‘The Golden Thread at the Heart of Tort Law: Protection of the Vulnerable’ (2003) 24 Aust Bar Rev 135. Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404; Hill v Chief Constable of West Yorkshire [1989] AC 53; [1988] 2 All ER 238; Giannarelli v Wraith (1988) 165 CLR 543; 81 ALR 417; D’Orta-Ekenaike v Victoria Legal Aid (2005) 223 CLR 1; 214 ALR 92. A



Constable of West Yorkshire,185 the argument is unpersuasive. It is little more than a reflection of a particular set of values and value judgments, resting on little or no empirical evidence.186 Moreover, it is doubtful whether social hosts would respond to a duty holding in a widely reported significant case by undertaking undesirable defensive practices. For example, it is unlikely that such a duty would result in people hosting fewer social occasions. On the assumption that tort has an educative function, if a duty were found in a well-publicised decision, the upshot would be safer parties and social occasions and a heightened sense of responsibility with respect to risks for which one could be held partly to blame. The ‘defensive practices’ argument could therefore be recast as ‘safer practices’ or ‘heightened care’, whereby the duty leads to changed modes of behaviour resulting in lessened risks of harm. Further, one should not equate the existence of a duty of care with breach or civil liability. The effect on the ‘coherence of the law’ also can be relevant in some circumstances. Specifically, if a duty were held to exist, how would it affect other causes of action or relevant statutory frameworks?187 For example, in Sullivan v Moody,188 the High Court convincingly reasoned that the imposition of a duty of care could undermine child protection efforts and statutory provisions imposing responsibility on relevant actors and agencies, unequivocally establishing the child’s welfare as of paramount importance.189 In Cole, coherence of the law should have been invoked to support a duty of care in the commercial context, but curiously the majority seemed to consider this of little importance. Establishing a duty of care would have been consistent with regulatory laws imposed on commercial alcohol providers, enhancing rather than conflicting with other statutory and criminal regimes. While the specific content of the criminal and civil standards might differ, the regimes nevertheless attempt to fulfil similar objectives: to inculcate a culture of safety. However, analogous regulatory regimes do not exist in social host circumstances, so the ‘coherence of the law’ consideration seems to be somewhat neutral. Certainly, the fact that this feature was downplayed in Cole suggests it would be ignored in a case like Childs.

185 186 187

188 189

related factor referred to in Sullivan v Moody and Hill v Chief Constable of West Yorkshire is the (speculated) negative effect on resource depletion and misallocation if a duty existed. [1989] AC 53; [1988] 2 All ER 238. Indeed, in Giannarelli v Wraith (1988) 165 CLR 543 at 578; 81 ALR 417 at 438, in reference to the Canadian position, Wilson J was dismissive of the fact that little empirical evidence demonstrated the need to retain advocates’ immunity. The effect that a duty holding would have had on defamation law was of significance to the High Court’s assessment in Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404 and Tame v New South Wales; Annetts v Australian Stations Pty Ltd (2002) 211 CLR 317; 191 ALR 449. The effect on the statutory framework sitting alongside the alleged duty was relevant to the no-duty holdings in Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540; 194 ALR 337 (with respect to the local authority and state) and Tame (with respect to the police officer). Sullivan v Moody (2001) 207 CLR 562; 183 ALR 404. See also D v East Berkshire Community Health NHS Trust [2005] 2 AC 373 (HL); [2005] 2 All ER 443. But see E Handsley, ‘Sullivan v Moody: Foreseeability of injury is not enough to found a duty of care in negligence — but should it be?’ (2003) 11 TLJ 1.



25

V Conclusion: Dampening tort’s potential value Actions for damages in negligence can serve useful purposes in some kinds of cases: successful litigation and the threat of instituting proceedings arguably have positive effects on behaviour. For example, proceedings brought by ‘passive smokers’, even in lower Australian courts, have yielded remarkable responses by all those operating self-contained environments, such as workplaces, where they might be held liable. Similarly, marginalised members of the community — including the homeless, some indigenous plaintiffs, prisoners and drug users — may attempt to effect change through tort, one of the few mechanisms in which they could possibly be empowered, using open public processes to subject powerful wrongdoers to judicial scrutiny.190 Similarly, tort’s role as a potential standard-setter, regulator and educator could be improved by imposing a duty of care in cases like Childs. Unfortunately, in the climate of legislative changes across Australia and given recent judicial tendencies, the High Court is unlikely to capture this opportunity. Instead, it would probably hold that no duty of care exists because of concerns about rising insurance premiums and the general absence of social host insurance, along with the expectation that guests and patrons will look after themselves and be responsible for harm caused to others. Tort’s efficacy would thereby be dulled and its value as a mechanism that might have had something useful and significant to say about the enhancement of safety would be reduced, if not extinguished. The likely High Court approach would have tangible, detrimental effects on community safety. Ultimately, the court’s message is that a commercial host (and a social host, in all likelihood) need not monitor potentially unsafe behaviour that could result in grave harm. Yet it is the commercial or social host who provides the environment in which a serious risk of harm is allowed to flourish; she or he has the opportunity and responsibility to diminish this risk. One can imagine the collective sighs of relief in commercial premises after the High Court handed down its decision in Cole, with its curtailment of liability in the commercial context and its emphasis on the drinker’s sole personal responsibility. Many would-be social hosts would no doubt be relieved if the High Court reached a similar result to that of the Supreme Court of Canada in a case like Childs (as predicted). On the other hand, if a duty of care was held to exist and liability was established in one case, it would not necessarily result in such a finding in another; holding that there is a duty of care does not equate with liability. The trial judge would have to consider all the circumstances in analysing breach and standard of care. If a duty was imposed on social hosts, reasonable hosts would have ample opportunity to be exonerated from fault and liability, for example by removing car keys or calling a taxi to ensure that their guests arrived home safely. This is something that neither courts nor social hosts should fear.

190 See I Malkin, ‘Tort Law’s Role in Preventing Prisoners’ Exposure to HIV Infection while in Her Majesty’s Custody’ (1995) 20 MULR 423. See, eg, Henry v Thomson [1989] 2 Qd R 412; New South Wales v Bujdoso (2005) 222 ALR 663.