Vicarious liability hits a dead end down under: Australia confines liability to employment relationships in Bird v DP (a pseudonym) [2024] HCA 41

Vicarious liability in English and Australian law pre-Bird

Vicarious liability is the principle form of secondary liability in the law of tort that holds an employer liable for an employee’s tort where i) there is a relationship capable of giving rise to vicarious liability (‘stage 1’), and ii) the tortfeasor commits a tort in the course of that relationship (‘stage 2’).[8] While this principle is common to Australian law and English law, it has evolved very differently in each jurisdiction. Given the HCA’s extensive comparative reasoning in Bird regarding stage 1 (mostly to highlight the perceived inadequacies of the English position), an overview of each approach is helpful to set the scene.

In English law, the Supreme Court of the United Kingdom (UKSC) has ‘relentlessly expanded the circumstances in which vicarious liability arises’ (Goudkamp and Plunkett).[9] At stage 1, both relationships of employment and those which are merely ‘akin to employment’ are capable of grounding liability.[10] At stage 2, liability for both torts committed in the course of employment and those merely ‘in close connection’ with the course of employment can be attributed to the employer or quasi-employer.[11] However, while the court’s unpredictable treatment of the stage 2 ‘close connection’ test continues to attract trenchant academic criticism, its application of the stage 1 ‘akin to employment’ test has escaped comparatively unscathed. A relatively settled picture has emerged from the decisions in Christian Brothers, Cox v Ministry of Justice,[12] and BXB: the court will engage in a detailed factual comparison between the relationship in question and a standard employment relationship,[13] then consider the underlying policy behind vicarious liability as a ‘final check’[14] on the result.[15] Any doubt which Armes v Nottinghamshire County Council[16]may appear to cast on this picture will be addressed in the evaluation section, where I argue that the decision in that case does not materially alter the stable and principled operation of the ‘akin to employment’ test.

On the contrary, Australian law has taken a much more restrictive approach with respect to both stage 1 and stage 2 of the doctrine. At stage 2, this was made abundantly clear in Prince Alfred College Inc v ADC,[17] where the HCA held explicitly that the Australian counterpart to the ‘close connection’ test at stage 2 (where a defendant is liable where they provide the ‘occasion’ rather than a mere ‘opportunity’ for the commission of the wrong)[18] would not impose liability on the facts of Mohamud v WM Morrison Supermarkets Plc,[19] though liability was found on these facts in the UKSC. 

Regarding stage 1, the position was less clear prior to Bird due to the absence of case law decided directly on this point.[20] For instance, though in CCIG Investments Pty Ltd v Schokman the HCA affirmed the strict rule that a relationship of employment is a necessary precursor to a finding of vicarious liability,[21] that case was decided on the basis of the stage 2 test. Moreover, while cases which were decided on the basis of the stage 1 do not employ an ‘akin to employment’ test, they do indicate the possibility of a more generous approach to liability through a purposive approach to ‘employment.’ For instance, in Hollis v Vabu Pty Ltd, the majority remarked that distinguishing between employees and independent contractors ‘is informed by a recognition of the fundamental purposes of vicarious liability … in the context of one of the many particular relationships that has developed.’[22]Similarly, in Sweeney v Boylan Nominees Pty Ltd, their Honours found that the defendant company was an independent contractor by examining particular features of the relationship.[23] There, too, the focus was on whether the relationship’s essential features, rather than legal category, could give rise to vicarious liability. The purposive tenor of these cases indicate the possibility of a less restrictive approach to liability which could extend to quasi-employment relationships.

Bird v DP is therefore a significant decision which confirms that Australian law will not extend this purposive approach to an ‘akin to employment’ test of liability under stage 1, and the determinative question remains whether there is a relationship of employment (as defined by statute and common law principles which inform the content and construction of a contract of employment).[24]

The decision in Bird v DP

The respondent (‘DP’) was taught at his Catholic primary school by a man named Bryan Coffey, who also served as the assistant parish priest at the local church. In 1971, at the age of five, DP was assaulted and sexually abused at his parents’ home on two separate occasions by Coffey. As an adult, DP brought proceedings in the Supreme Court of Victoria in 2020, claiming damages for psychological injuries sustained due to the assaults committed by Coffey. DP alleged that the Roman Catholic Diocese of Ballarat (“the Diocese”) which the parish church belonged to was vicariously liable for Coffey’s actions, and also liable for the breach of a non-delegable duty to exercise reasonable care in its authority, supervision and control of Coffey’s conduct. As the Diocese is not a legal person, DP brought the proceeding against the nominated defendant Paul Bird, the current Bishop of Ballarat.

The main issue was that DP could not establish that the priest was an employee or agent of the Diocese. When faced with the same issue in Christian Brothers on a very similar fact pattern (involving sexual abuse by teachers who were not employed by the religious organisation being sued), the UKSC’s response was to find that inter alia the hierarchical structure, teaching activity undertaken under the organisation’s direction, and behavioural rules of the organisation imposed on members were sufficient to ground a claim in vicarious liability, despite the absence of a contract of employment.[25] In Lord Philips’ leading judgement, this was justified by the fact that the same primary objectives which support the imposition of liability in ‘true’ employment cases were present in cases with particular similarities.[26]

The first-instance and Court of Appeal decisions in DP v Bird adopted very similar reasoning. At first-instance, Forrest J found that though Coffey was not an employee or agent of the Diocese, their relationship was still capable of giving rise to liability based on an assessment of ‘the totality of the relationship’[27], including the ‘close nature of the relationship’ between the Bishop, Diocese and local Catholic community, ‘general control’ of the Diocese over Coffey’s role and duties, Coffey’s ‘pastoral role’ in the Catholic community, and the ‘relationship … of intimacy and imported trust’ between DP, his family, Coffey and the Diocese.[28] 

Likewise, the Victorian Court of Appeal noted that while Coffey was not an employee of the Diocese ‘in a strict legal sense’, the Bishop exercised control over Coffey that was ‘at least as great as, if not greater than, that enjoyed by an employer’. The court concluded that Coffey was ‘not acting independently of the Diocese, but as a representative of it’, and was ‘an emanation of the Diocese’.[29] For these reasons, the relationship gave rise to vicarious liability under the first stage. The second limb of the test was also satisfied (at both first-instance[30] and in the Court of Appeal[31]), and the Diocese was thus held vicariously liable.

Departing from both lower courts and the approach in Christian Brothers, the HCA allowed the appeal and found that the Diocese was not vicariously liable for the actions of the tortfeasor. A majority of the Court found that an employment relationship was a ‘necessary precursor’ to grounding vicarious liability under the first limb of the test,[32] and refused to expand its boundaries to include relationships ‘akin to employment.’[33]

Discussion

The majority judgement by Gageler CJ, Gordon, Edelman, Steward and Beech-Jones JJ advanced three main arguments to support this position, many of which were also endorsed by Jagot J in her concurring judgment. It will be argued that each is unsound.

First, the majority argued that Australian law recognises no clear justification for the imposition of vicarious liability, and so has no foundation by which to extend its existing scope. The majority emphasised that enterprise risk is not an accepted overarching justification in Australian law, despite its prominence in Canada and the UK.[34]

However, it is doubtful whether the lack of a clear justification for vicarious liability means there is no solid foundation to redraw its boundaries. Though it is generally accepted that there is no unitary justification for vicarious liability, this does not mean there is no basis whatsoever for its existence. If that were the case, abolition would be preferable; this is not a route contemplated by the HCA. The better view is that vicarious liability pursues a combination of policy-based justifications including enterprise liability, loss-allocation, and accident prevention, among others.[35] 

Crucially, this pluralist view of vicarious liability already has a solid foundation in Australian case law as discussed above: it appears in the purposive construction of employer-employee relationships at stage 1.[36] Indeed, it was conceded in Jagot J’s concurring judgement that ‘employment’ in the context of vicarious liability necessarily involves consideration of the substantive characteristics of employment relationships rather than precise legal labels.[37] As Gleeson J argued in her powerful dissent, per Personnel Contracting, such characteristics may include ‘the right of control by a putative employer’,[38] ‘integration of the worker’s role into the enterprise’,[39] the employer’s ‘capacity … to compensate a plaintiff for harm suffered’,[40] and the employer’s role in ‘setting in motion … the events in which the tortious conduct occurred.’[41] These ‘employment’ characteristics correspond to the theoretical justifications of control, enterprise liability, loss compensation and risk creation which were noted by Lord Philips in Christian Brothers as underlying principles behind vicarious liability.[42]

It is, therefore, clear that the Australian imposition of vicarious liability is founded on several essential features which pursue particular policy goals, as opposed to the mere presence of an employment contract. Given this, Gleeson J was correct to point out that it is anomalous not to even consider imposing liability when these essential features arise outside the classic employment context. For instance, in Bird, after the majority identified that there was no legal employment relationship, they disregarded Coffey’s position as an ‘emanation of the Diocese’[43] entirely, failing to acknowledge that he was subject to its control and able to exercise its influence. Had there been an employment contract present, the case may have been determined by critical engagement with these features under the purposive approach to constructing “employment”. The discrepancy here reveals the deeply unprincipled line drawn between substantially identical relationships which differ only on the presence or absence of a contract of employment.

Second, it was argued that the akin to employment test adopted in the UK is liable to produce ‘uncertainty and indeterminacy’ in the law, as it depends on an assessment on what is fair, just, and reasonable with respect to contestable policy choices and the allocation of risk, even though there is wide scope for disagreement on these matters. The essence of this objection is a rejection of ‘palm tree justice’.

Yet, this is an incorrect characterisation of how the ‘akin to employment’ test operates in English law. As discussed above, the types of relationships that attract liability under the test in English law are determined primarily by detailed, fact-specific analogy with standard employment cases rather than an abstraction of what is fair, just and reasonable. Lord Burrows explained this in BXB, where he noted that under the test ‘a court needs to consider carefully features of the relationship that are similar to or different from a contract of employment,’ such as how integral the tortfeasor’s work is to the organisation, the extent of the D’s control over the tortfeasor, and any hierarchy or seniority.[44]  Lord Reed made the point more explicitly in Cox, stating that ‘where the criteria [of a relationship akin to employment] are satisfied, it should not generally be necessary to re-assess the fairness, justice and reasonableness of the result in a particular case. Such an exercise … would be liable to lead to uncertainty and inconsistency.’[45]

Admittedly, this objection may have had greater force in the wake of Armes, where the court took a more purposive approach to stage 1. Namely, they considered in depth the presence of relevant contextual features which might justify imposing liability. Nonetheless, the court’s approach still retained detailed factual analysis and structured its measured consideration of context with reference to the five ‘policy incidents’ outlined by Lord Philips in Christian Brothers rather than a generic conception of ‘fair, just and reasonable.’[46] The reasoning in Armes thus also cannot fairly be described as ‘palm tree justice.’ Even if the purposive approach taken in that case did fit that description, the approach endorsed by Lord Burrows in BXB, (a descriptive rather than purposive comparison of the relationship with a contracted employment) is now the leading approach to determining liability in English law under stage 1. This determination is a coherent extension of the employment relationship category which grants workers greater certainty, as their protection in tort law is based on the substance of their relationship rather than any disruptions to legal relationships by opportunistic employers or the vagaries of the labour market.

Thirdly, the majority argued that importing the ‘akin to employment’ test into Australian law would threaten the distinction between independent contractors and employees which is ‘too deeply rooted to be pulled out’[47].  While they correctly recognised that this distinction has been retained in the UK,[48] they doubted its continued force in light of the multifactorial approach to grounding liability outside employment relationships. In her concurring judgment, Jagot J also emphasised that recognising a “[relevant] class of legal actor being neither an employee nor an independent contractor” for the purposes of vicarious liability would be an extension of existing Australian law.

However, this view overlooks the fact that excluding vicarious liability in independent contractor cases has been a carefully maintained bright line in English law, notwithstanding other expansions of vicarious liability.[49] This has been possible since the distinction is generally not just one of form, but one of substance, too. For instance, in Barclays Bank, the tortfeasor, a doctor, was held to be an independent contractor rather than ‘anything close to an employee’[50] as he was free to refuse offered examinations, was in business on his own account, and was not paid a retainer to accept referrals.[51] These features, which are characteristic of many independent contractors, weigh normatively against a finding of vicarious liability against the bank. The doctor’s freedom to refuse examinations signifies the lack of control held by the bank, while the fact he was on business on his own account demonstrates the inapplicability of enterprise liability attributable to the bank. This reveals that by following the principles which underpin vicarious liability, there is a justifiable basis to distinguish independent contractors from ordinary employees. These distinctions are not present in cases where the court identifies essential factual and normative features ‘akin to employment’, such as foster parents working in a council placement[52] or prisoners working for free in a prison.[53]

To the extent that an independent contractor-hirer relationship does meaningfully resemble an employee-employer relationship in a particularly exceptional case (perhaps even due to an opportunistic employer’s deliberate goal to avoid liability), courts should not rush to conclude that crossing the bright line signals a descent into incoherence. Exceptionally departing from a bright line rule to reach a result consistent with the underlying justifications of vicarious liability strikes a more than acceptable balance between certainty and the demands of fairness. Provided that the court remains attentive to the substantive features which generally distinguish standard employees from independent contractors, there is nothing to suggest that imposing vicarious liability in such cases will systematically erode the distinction. That is a fallacious assertion of a slippery slope from the Bird majority.[54] While obviously a possibility, the coherent normative distinction between independent contractors and employee-type cases outlined above, as well as the existing line of English cases, weigh against any significant likelihood of such an outcome. 

Conclusion

In light of the above, the plurality’s refusal to moderately extend vicarious liability to relationships “akin to employment” does not seem justified. As cogently established in Gleeson J’s dissent, there was no direct authority binding the HCA to restrictively recognise only employment relationships for the purposes of vicarious liability. The plurality themselves admitted that their decision was instead founded on the view that ‘abandoning the threshold requirement of a relationship of employment … does not fit within the body of accepted rules and principles.’ Yet, as demonstrated, the purposive construction of employment revealed in cases like Hollis and Sweeney show that there are justifications for vicarious liability which stretch beyond the mere employment contract context. Alternate arguments for the existing limitation motivated by preserving certainty were founded on an incorrect understanding of how the ‘akin to employment’ test operates in English law - that is, by factual analysis drawing close comparisons with the essential features of a standard employment contract. Lastly, the HCA’s fears of collapsing the independent contractor-employee distinction were overstated; they did not adequately credit the principled preservation of the distinction in English law. As such, the plurality’s final decision is to be regretted. It is hoped that Gleeson J’s dissent has left the door open for a different conclusion should the same issue come before the HCA again. 

Footnotes:
 

[1] [2024] HCA 41 (hereafter, subsequent references in footnotes to ‘Bird’ refer to the judgment of the HCA, not those of lower courts which will be referred to in full in discussion below).

[2] Doe v Bennett, 2004 SCC 17, [2004] 1 SCR 436.

[3] S v A-G [2003] NZCA 149, [2003] 3 NZLR 450 [64].

[4] Catholic Child Welfare Society v Institute of the Brothers of the Christian Schools [2012] UKSC 56, [2013] 2 AC 1 (‘Christian Brothers’).

[5] Hickey v McGowan [2017] IESC 6, [2017] 2 IR 196.

[6] Ng Huat Seng v Munib Mohammad Madni [2017] SGCA 58, [2017] 2 SLR 1074.

[7] Christian Brothers (n 4) [47] (Lord Phillips JSC), quoting JGE v Trustees of the Portsmouth Roman Catholic Diocesan Trust [2012] EWCA Civ 938, [2013] QB 722 [57] (Ward LJ). 

[8] Construction, Forestry, Maritime, Mining and Energy Union v Personnel Contracting Pty Ltd (2022) 275 CLR 165 (‘Personnel Contracting’) 239 [191] (Gordon J); Trustees of the Barry Congregation of Jehovah’s Witnesses v BXB [2023] UKSC 15, [2024] AC 567 (‘BXB’) [58] (Lord Burrows JSC).

[9] Goudkamp, J. and Plunkett, J. (2017) ‘Vicarious liability in Australia: on the move?’, Oxford University Commonwealth Law Journal, 17(1), 162.

[10] BXB (n 8), [58] (Lord Burrows JSC).

[11] ibid.

[12] [2016] UKSC 10; [2016] AC 660 (‘Cox’)

[13] Christian Brothers (n 4) [56]-[59], [89]; Cox (n 13) [32], [34], [36]-[37]; BXB (n 8), [66]-[68].

[14] BXB (n 8), [82] (Lord Burrows JSC). 

[15] Christian Brothers (n 4) [34] (Lord Phillips JSC); Cox (n 13) [41]-[42] (Lord Reed JSC); BXB (n 8), [82 (Lord Burrows JSC)].

[16] [2017] UKSC 60; [2017] 3 WLR 1000 (‘Armes’). 

[17] [2016] HCA 37. 

[18] ibid [80]. 

[19] [2016] UKSC 11, [2016] AC 677.

[20] Bird (n 1) [101] (Gleeson J).

[21] CCIG Investments Pty Ltd v Schokman (2023) 97 ALJR 551 [12]-[14], [64]-[66]. 

[22] Bird (n 1) [116] (Gleeson J), quoting Hollis v Vabu Pty Ltd (2001) 207 CLR 21 [59]. 

[23] Bird (n 1) [121] (Gleeson J), citing Sweeney v Boylan Nominees Pty Ltd (2006) 226 CLR 161.

[24] WorkPac Pty Ltd v Rossato (2021) 271 CLR 456 at 477 [56]-[58], 478-479 [62]; Construction, Forestry, Maritime, Mining and Energy Union v Personnel Contracting Pty Ltd (‘Personnel Contracting’) (2022) 275 CLR 165 at 226 [162], 228 [172].

[25] Christian Brothers (n 4) [56] (Lord Phillips JSC). 

[26] ibid [34]-[35]. 

[27] DP (a pseudonym) v Bird [2021] VSC 850 [213] (Forrest J).

[28] ibid [278]. 

[29] Bird v DP (a pseudonym) [2023] VSCA 66 [125], [128]–[129] (Beach, Niall and Kaye JJA).

[30] DP (a pseudonym) v Bird [2021] VSC 850, [213] [280] (Forrest J). 

[31] Bird v DP (a pseudonym) [2023] VSCA 66, [148] (Beach, Niall and Kaye JJA). 

[32] Bird (n 1) [45] (Gageler CJ, Gordon, Edelman, Steward, and Beech-Jones JJ). 

[33] ibid [47].

[34] ibid [48], [50].

[35] See Marco Cappelletti, ‘A Pluralist View of Vicarious Liability in Tort’ (2024) 140 LQR 61 for detailed consideration.

[36] Hollis v Vabu Pty Ltd (2001) 207 CLR 21 [59]; Sweeney v Boylan Nominees Pty Ltd (2006) 226 CLR 161. 

[37] Bird (n 1) [231] (Jagot J): ‘It is not to the point that a person may have been characterised as an employee or independent contractor for other purposes, such as taxation purposes.’

[38] ibid [98] (Gleeson J).  

[39] ibid.

[40] ibid [90].

[41] ibid [90].

[42] Christian Brothers (n 4) [35] (Lord Phillips JSC) 

[43] Bird v DP (a pseudonym) [2023] VSCA 66 [129] (Beach, Niall and Kaye JJA).

[44] BXB (n 8) [58].

[45] Cox (n 12) [41]. 

[46] Armes (n 16) [56], [59]-[65].

[47] Bird (n 1) [63], [241]. 

[48]See, eg, Woodland v Essex CC [2013] UKSC 66, [2013] AC 537 [3] (Lord Sumption JSC); Barclays Bank plc v Various Claimants [2020] UKSC 13 [2020] AC 973 [22] (Lady Hale JSC); Trustees of the Barry Congregation of Jehovah’s Witnesses v BXB [2023] UKSC 15, [2024] AC 567 [58] (Lord Burrows JSC). 

[49] Woodland (n 17) [3] (Lord Sumption JSC) (‘The boundaries of vicarious liability have been expanded by recent decisions … but it has never extended to the negligence of those who are truly independent contractors’); Barclays (n 43.) [19] (Lady Hale JSC). 

[50] Barclays (n 43) [28] (Lady Hale JSC).

[51] ibid.

[52] Armes (n 42.). 

[53] Cox (n 41.). 

[54] Bird (n 1) [66].