| Original Full Text | Campbell, M. and Lindsay, B. (2024) Refining vicarious liability. Edinburgh Law Review, 28(2), pp. 174-206. (doi: 10.3366/elr.2024.0892) This is the author version of the work. There may be differences between this version and the published version. You are advised to consult the published version if you want to cite from it: https://doi.org/10.3366/elr.2024.0892 https://eprints.gla.ac.uk/319720/ Deposited on 14 February 2024 Enlighten – Research publications by members of the University of Glasgow http://eprints.gla.ac.uk Page 1 of 25 REFINING VICARIOUS LIABILITY A. INTRODUCTION Vicarious liability is the treatment, regardless of fault, of one defender as liable1 for another defender’s wrong.2 The basic inquiry involves examination of whether (i) the defenders stand in a relationship capable of giving rise to vicarious liability; and (ii) the wrongdoing is relevantly linked to that relationship.3 Unsatisfactory though the reality and its results may be,4 the contours of these core components are sculpted as new fact patterns come before the courts,5 and as judges emphasise differently,6 and compromise between,7 various policy goals.8 We have examined previously how vicarious liability has developed and is likely to develop.9 The elementary principles are quite well-settled, especially given a relative paucity of case law and literature. But room for refinement remains. We here further analyse issues of interest in Scotland and beyond concerning the basic inquiry (Sections B and C), and clarify the unimportance in this context of the distinctiveness of Scots law (Section D), before summarising our conclusions (Section E). B. RELATIONSHIPS GIVING RISE TO VICARIOUS LIABILITY The first stage of the basic inquiry concerns the relationships from which vicarious liability arises. It arises principally from employment (including employment pro hac vice) and relationships akin to employment. 10 This Section suggests ways better to understand the former, and prescribes11 a proper approach to the latter. One preliminary point. Practitioners know of the need to advise on vicarious liability as a risk.12 Working practices evolve,13 and it is recognised that organisations may seek to elude vicarious 1 In Scotland, the liability is joint and several: Law Reform (Miscellaneous Provisions) (Scotland) Act 1940, s 3; Bell v Alliance Medical Ltd [2015] CSOH 34, 2015 SCLR 676. 2 Benkert UK Ltd v Paint Dispensing Ltd [2022] CSIH 55, 2023 SLT 19 at para 53. What actually is attributed to the vicariously liable defender – always the second defender’s wrong, always her acts, or sometimes one, sometimes the other – has not been discussed at length in Scotland. Compare e.g. R Leow, Corporate Attribution in Private Law (2022) ch 4; CCIG Investments Pty Ltd v Schokman [2023] HCA 21 at paras 49-81. 3 C v Shaw [2023] CSOH 11, 2023 SLT 359 at paras 25-26; affirmed [2023] CSIH 36, 2023 SLT 1327. 4 R Stevens, Torts and Rights (OUP 2007) ch 11 esp 244, 259, 270, 274; J Goudkamp and D Nolan, Winfield & Jolowicz on Tort (20th edn, 2020) para 21.006. 5 Mohamud v Wm Morrison Supermarkets Plc [2016] UKSC 11, [2016] AC 677 at paras 10, 26, 39, 55-56. 6 Armes v Nottinghamshire County Council [2017] UKSC 60, [2018] AC 355 at para 56; noted B Lindsay (2018) 22 Edin LR 294. 7 E v English Province of Our Lady of Charity [2012] EWCA Civ 938, [2013] QB 722 at paras 20, 47-55, 60-61, 71-72, 81 (Ward LJ), 133 (Davis LJ). 8 Vaickuviene v J Sainsbury Plc [2012] CSOH 69, 2012 SLT 849 at paras 22-24, 35-36 (reversed but not on this point [2013] CSIH 67, 2014 SC 147), mentioning “general policy considerations which have been advanced to justify vicarious liability, including the discussion in” Majrowski v Guy’s and St Thomas’ NHS Trust [2007] UKHL 34, [2007] 1 AC 224 at para 9. 9 M Campbell and B Lindsay, “Vicarious Liability in Scotland” in P Giliker (ed), Vicarious Liability in the Common Law World (2022) 197; hereafter “Campbell and Lindsay”, and “VLCLW”, respectively. 10 Grubb v Shannon [2018] SCGLA 13, 2018 SLT (Sh Ct) 193 at paras 48-49, 55-56, 63, 66, 86, also mentioning dual vicarious liability, which we discuss elsewhere: Campbell and Lindsay (n 9) 200-203. 11 See J Smits, “What Is Legal Doctrine?” in R van Gestel, H-W Micklitz and EL Rubin (eds), Rethinking Legal Scholarship: a Transatlantic Dialogue (2018) 207 at 217-219. 12 K Patrick and MC Bowly, “Vicarious Liability of Employers: Supreme Court Provides Welcome News” (2020) 9(5) Compliance & Risk 10 at 13-14. 13 Commenting from the perspective of vicarious liability, see P Giliker, “Vicarious Liability and Corporations” in M Petrin and C Witting (eds), Research Handbook on Corporate Liability (2023) 274 at 276-278. Page 2 of 25 liability and other burdens by making persons who act beneficially on their behalf appear to be independent contractors.14 Thus, judges must prioritise reality over appearance when analysing the arrangements between defenders coming before them in vicarious liability cases. In the simplest terms, two propositions underpin this approach, each explained differently.15 First, how parties label their relationships is immaterial. This is because their view on a matter of law is, strictly, irrelevant.16 Secondly, judges may overlook the letter of defenders’ contractual – and presumably other – arrangements where they contradict reality, even if otherwise unchallenged: 17 whether a relationship gives rise to vicarious liability “turns on all the circumstances of the case”.18 This proposition is well-vouched by authority. A justification for it is harder to identify. But one could usefully start with and enlarge upon a principle, short of a sham doctrine requiring party collusion, that “rules of contract law ([such as] usual rules of contractual construction) are not to be conscripted in aid of transactions or documents that are deceitful or that mask the true agreement of the parties”.19 (1) Employment and employment pro hac vice One is not generally liable for the delict of one’s independent contractor.20 In contrast, it is well-established that employment gives rise to vicarious liability.21 Continuing to refine how we view the principal relationship from which vicarious liability arises 22 may enable practitioners to advise more reliably, and reduce the frequency with which employment is disputed in litigation. Furthermore, a firmer understanding of employment proper feeds positively into doctrine on the identification of employment pro hac vice, shared employment, and relationships akin to employment.23 Our present understanding may be advanced in at least two ways. 14 Grubb (n 10) at paras 58-59; Cox (n 41) at paras 29-31; E (n 7) at paras 58-60; Morrissey v Health Service Executive [2020] IESC 6, [2020] PNLR 17 at paras 12.11-12.16; J Fulbrook, “Reverberations from Uber v Aslam in personal injury claims?” [2021] JPIL 57 at 63-66, also mentioning that one could always instate vicarious liability for independent contractors. 15 Policy ideas influence developments at this stage of the vicarious liability inquiry but cannot be relied upon for want of consensus: Grubb (n 10) at paras 50-51, 58-59; E (n 7) at paras 50-60. Labour and tax law cases show that the approach suggested in the text nevertheless requires a satisfactory rationale: compare Uber BV v Aslam [2021] UKSC 5, [2021] ICR 657 at paras 69-87; with Revenue and Customs Commissioners v Atholl House Productions Ltd [2022] EWCA Civ 501, [2022] ICR 1059 at para 156. Note, however, that whether a relationship gives rise to liability does not depend on how it is seen from other perspectives: Cox (n 41) at para 11; Various Claimants v Barclays Bank Plc [2020] UKSC 13, [2020] AC 973 at para 29. 16 Even if occasionally helpful: Construction, Forestry, Maritime, Mining and Energy Union v Personnel Contracting Pty Ltd [2022] HCA 1, (2022) 398 ALR 404 at paras 66 (Kiefel CJ, Keane and Edelman JJ), 184 (Gordon J). 17 E.g. as a sham or as having been varied: compare ibid esp at paras 43, 59, in which it was decided that contract terms were an unchallenged, comprehensive, and so decisive, record of the parties’ relationship. See also ZG Operations Australia Pty Ltd v Jamsek [2022] HCA 2, (2022) 398 ALR 603; E Schofield-Georgeson and JR Munton, “Precarious Work in the High Court” (2023) 45 Sydney LR 219. 18 Natwest Markets Plc v Bilta (UK) Ltd [2021] EWCA Civ 680 at para 151; citing Mersey Docks and Harbour Board v Coggins & Griffith (Liverpool) Ltd [1947] AC 1 (HL) at 10, 15, 20, 21; in turn cited in Kerr v Hailes (Plant) Ltd 1974 SLT (Notes) 31 at 32. Based on this proposition one might overlook unreal contractual provisions denying any obligation to do or provide work, or instituting sweeping substitution entitlements. 19 P Bomball, “Contractual Autonomy, Public Policy and the Protective Aim of Labour Law” (2020) 44 Melbourne ULR 502 at 522-534 esp 529. 20 For an exception, see Esso Petroleum Co Ltd v Scottish Ministers [2016] CSOH 15, 2016 SCLR 539 at paras 14-24. 21 Somerville v Harsco Infrastructure Ltd [2015] SCEDIN 71 at para 18. 22 Grubb (n 10) at paras 63, 86. 23 For the third of these, this may be inferred from, and was tentatively noticed in, respectively, Hughes v Rattan [2022] EWCA Civ 107, [2022] 1 WLR 1680 at para 88; E (n 7) at paras 63-64. Page 3 of 25 One way concerns the identification or denial of employment in particular instances.24 Many relevant indicia are recognised.25 And well-known is the requirement to take “a broad overhead view” of whether employment exists, considering changing social, political, and working conditions.26 But there can be neither a test for all cases,27 nor a fixed list of factors to address.28 Whether a set of facts discloses employment may thus depend on rather unrestrained, multifactorial examinations. Refining such impressionistic exercises may generate greater certainty and increase the utility of precedents at no cost. The presence, absence, and relative significance – both abstract and concrete – of indicia of employment should be considered with external, structuring guidance in mind.29 When occasion arises, pleaders may benefit from reference to, and judges from emphasising, certain general questions. These include the extent to which a wrongdoer depends economically on a particular purported employer for the financial exploitation of her talents;30 whether skill or labour in return for remuneration is offered as opposed, for example, to an end result simpliciter; or whether a purported employee contracts to work as part and parcel of another’s business,31 rather than to act on her own account as an independent, or principal,32 perhaps conducting a separate business of her own.33 Channelling the weighing of the indicia of employment through such questions is a transparent way to assess their materiality. It may also assist with the separation from irrelevant matters of those features of an association between defenders which bear on whether they stand in an employer-employee relationship. A second way our understanding of employment will continue to improve is illustrated by turning to employment pro hac vice, from which it is also well-established that vicarious liability may arise.34 In short, practitioners who advise and argue, and judges who decide, must act readily upon the clear licence which the authorities give to keep the law moving.35 The cases disclose general questions to guide factual analysis. One is whether there has occurred a transfer of responsibility for conduct in the discharge of an eventual wrongdoer’s functions.36 Another is whether there has been a transfer of capacity to direct, or to delegate discretion as 24 Valuable insights may be gleaned from labour law scholarship: S Deakin, “Decoding Employment Status” (2020) 31 King’s LJ 180; P Bomball, “Vicarious Liability, Entrepreneurship and the Concept of Employment at Common Law” (2021) 43 Sydney LR 83; P Bomball, “The ‘Entrepreneurship Approach’ to Determining Employment Status: A Normative and Practical Critique” (2021) 44 UNSWLJ 1336; C Sutherland, “Judging the Employment Status of Workers: An Analysis of Commonsense Reasoning” (2022) 46 Melbourne ULR 281. 25 See those mentioned in Campbell and Lindsay (n 9) at 198; one may also instance, e.g. the extent of any substitution entitlements: MacFarlane v Glasgow City Council [2001] IRLR 7 (EAT). 26 Kilboy v South Eastern Fire Area Joint Committee 1952 SC 280 (IH) at 285-287 (Lord Cooper, with whom Lord Carmont and Lord Russell agreed). 27 Macdonald and Hayward v Glasgow Western Hospitals 1954 SC 453 (IH) at 465 (Lord Cooper (in Macdonald’s case), with whom Lord Carmont and Lord Russell agreed). 28 Dignon v Irving 1999 SCLR 601 (Sh Ct, Kil) at 620. 29 Personnel Contracting (n 16) at paras 33-39, 62 (Kiefel CJ, Keane and Edelman JJ), 185-189 (Gordon J). 30 Hall (Inspector of Taxes) v Lorimer [1994] 1 WLR 209 (CA) at 218. 31 Which need not be a sole or permanent post: Lee Ting Sang v Chung Chi-Keung [1990] 2 AC 374 (PC (HK)) at 384, 388. 32 United Wholesale Grocers Ltd v Sher 1993 SLT 284 (OH) at 287; Grubb (n 9) at para 72; citing the useful elaboration in E (n 7) at para 72. 33 But perhaps not: see Grubb (n 10) at para 99; Personnel Contracting (n 16) at paras 39 (Kiefel CJ, Keane and Edelman JJ, noting that an “‘own business / employer’s business’ dichotomy may not be perfect so as to be of universal application [because] not all contractors are entrepreneurs”), 180-183 (Gordon J, elaborating). 34 King v Fife Council 2004 Rep LR 33 (OH) at para 9. 35 See Kilboy (n 26) at 285 (Lord Cooper), 287 (Lord Keith); citing Short v J&W Henderson Ltd 1946 SC (HL) 24 at 34, where Lord Thankerton opined that the indicia of employment may require to be restated. 36 King (n 34) at paras 11-12, 14-15. Page 4 of 25 to, how work is done.37 It may well be that employment pro hac vice should come to be assessed on the same integration-type basis as dual vicarious liability.38 The former involves a complete transfer of responsibility for a wrongdoer, rather than the partial transfer giving rise to the latter.39 Meantime, however, the above guidance is helpful, and can accommodate desirable refinements in this area when the opportunity comes. Two examples. In the first place, the precise manner in which control over a wrongdoer is relevant, 40 recognised as regards employment,41 relationships akin to employment,42 and dual vicarious liability,43 should be affirmed in the context of employment pro hac vice. When considering transfer of control, emphasis should be on control over what a wrongdoer does, now how she does it.44 The status of control in the employment context is especially pertinent here. If one can be a general employer without detailed control over an employee, it is unclear why a borrowing employer should escape a finding of vicarious liability merely by taking someone – perhaps deliberately – over whom it exercises no such control. In the second place, there is a considerable body of authority to the effect that “[t]he more skilled the employee the more difficult it may be to demonstrate a transfer of control over the way in which he carried out his duties and therefore the more difficult it may be to demonstrate a transfer of responsibility for his acts and omissions”.45 As a generality, this proposition is unsound. “A brain surgeon may very well be an employee. A window cleaner is commonly self-employed.”46 Skill, ingenuity, and initiative, do not by themselves render a person’s becoming generally employed more difficult.47 They cannot by themselves render employment pro hac vice more difficult.48 37 McGregor v JS Duthie & Sons & Co Ltd 1966 SLT 133 (IH) at 141 (Lord Grant), 143 (Lord Strachan), 145 (Lord Walker). 38 As essentially seemed to occur when dual vicarious liability was recognised but employment pro hac vice was not in Natwest (n 18) esp at paras 151-156, 185-187; though the penultimate sentence in the second-cited paragraph range may neglect Hawley v Luminar Leisure Ltd [2006] EWCA Civ 18, [2006] IRLR 817. 39 Which we consider will likely be recognised by the Scottish courts when the occasion arises: Campbell and Lindsay (n 9) 200-203. 40 For general recognition that it is no longer crucial and is to be assessed alongside other factors, see King (n 34) at paras 11, 15. 41 Cox v Ministry of Justice [2016] UKSC 10, [2016] AC 660 at para 21; Various Claimants v Catholic Child Welfare Society [2012] UKSC 56, [2013] 2 AC 1 at para 36 (“CCWS”). 42 Armes (n 6) at paras 56, 62, 64-66. 43 Natwest (n 18) at paras 152, 181; CCWS (n 30) at para 45. 44 Compare, e.g. McGregor (n 37) at 141, 143, 145; Kerr v Hailes (Plant) Ltd 1974 SLT (Notes) 31 (OH); Mt Owen Pty Ltd v Parkes [2023] NSWCA 77 at para 49. 45 King (n 34) at para 11; and cases in which there was no defence of employment pro hac vice because wrongdoers were skilled e.g. Moir v Wide Arc Services Ltd 1987 SLT 495 (OH) at 497-498; Brogan v William Allan Smith & Co 1965 SLT 175 (OH) at 177-178, quoting and mentioning unreported dicta of Viscount Simonds in John Young & Co (Kelvinhaugh) Ltd v O’Donnell 1958 SLT (Notes) 46 (HL, Scot). 46 Hall (n 30) at 216. 47 Stagecraft Ltd v Minister of National Insurance 1952 SC 288 (IH) at 297-298 (Lord Cooper), 300-301 (Lord Jamieson), 302 (Lord Patrick); Zuijs v Wirth Brothers Pty Ltd (1955) 93 CLR 561 (HCA) at 569-573 (Dixon CJ, Williams, Webb, and Taylor JJ). See also the similar discussion, but of a narrower proposition about skilled persons, with a particular statutory backdrop, in Macdonald and Hayward (n 27) at 473-474 (Lord Cooper (in Hayward’s case)), 483-485 (Lord Carmont). 48 For a move in this direction, see Parkes v Mt Owen Pty Ltd [2022] NSWSC 909 at paras 81-102, in which a distinction was accepted between the provision of a skilled worker (a) “where the host employer does not generally employ persons with [particular] qualifications”, and (b) “to make up some shortfall temporary or otherwise in the host’s skilled workforce”. The facts of the case were of type (b), and there was employment pro hac vice. The proposition that if skill does not prevent employment it does not prevent employment pro hac vice was noted on appeal in apparently neutral fashion: [2023] NSWCA 77 at paras 34-39, 58. Page 5 of 25 (2) Relationships akin to employment The notion that a relationship akin to employment gives rise to vicarious liability is less novel in Scotland than some may think.49 The basic idea has gained more general acceptance in recent years under the influence of English learning.50 However, it appears that the law relating to what may be termed quasi-employment could be better understood.51 Existing coverage in the books consists largely of summaries of leading authorities.52 Instead of sketching another of these, therefore, we here cut through the cases, and prescribe what we see as the correct approach to quasi-employment cases. The Scottish courts are in a position to adopt it. Before proceeding, we note that the development of the relationship akin to employment renders unattractive any further erosion of the general principle against vicarious liability for the wrong of a true independent contractor in order, for example, to protect persons wronged by gig economy workers,53 difficult though it may remain to draw relevant distinctions.54 It is of course always necessary to identify a second defender to be vicariously liable along with a wrongdoing first defender.55 But the issue is to the fore in this context. The search for what may be termed a quasi-employer will sometimes be simple. In other cases, it may be more challenging than in standard situations of employment or shared employment. There may be multiple potential quasi-employers, actors therein, and representatives thereof,56 among whom relevant powers and responsibilities are distributed. Behind a wrongdoer there may, for example, be an overall governing religious body, satellite setups for the body, members from time to time of a particular congregation, and trustees57 from time to time of that congregation. Resistance on the part of these groupings in litigation could give rise to difficult questions. Most generally: who is or are the vicariously liable quasi-employer(s), and who should satisfy 49 Campbell and Lindsay (n 9) at 203-206; discussing Marshall v William Sharp & Sons Ltd 1991 SLT 114 (IH). 50 See judicial statements, argument, and references, in Grubb (n 10) at paras 60-63, noting the influence of litigation involving child abuse; Kennedy v Bonnici [2021] CSOH 106, 2022 SLT 63 at paras 55-56, obiter; X v Y [2023] CSOH 17, 2023 SC 235 at paras 13, 18, 21. Movement at common law in Australia has been more tentative: RC v The Salvation Army (Western Australia) Property Trust [2023] WASCA 29 at paras 141-156; Bird v DP [2023] VSCA 66 at paras 80-130 esp 82, 113-117; special leave granted [2023] HCATrans 145; but for relevant legislative developments, see C Beuermann, “Vicarious Liability in Australia”, in VLCLW (n 9) 73 at 98-100. 51 See, e.g. F McManus and others, Delict: A Comprehensive Guide to the Law in Scotland (3rd edn, 2021) paras 20.54-20.56, 20.77-20.92, in which relevant aspects of cases on point are mentioned only after discussion turns from relationships giving rise to vicarious liability to the distinct matter of whether wrongdoing is relevantly connected to the defenders’ relationship. See also A Tettenborn (gen ed), Clerk & Lindsell on Torts (24th edn, 2023) paras 6.35-6.36. 52 G Cameron, Thomson’s Delictual Liability (6th edn, 2021) para 12.8; H MacQueen and Lord Eassie (eds), Gloag & Henderson: The Law of Scotland (15th edn, 2022) para 25.09; EC Reid, The Law of Delict in Scotland (2022) paras 3.23-3.30. 53 See Barclays (n 15) at paras 24, 27, 29 (though as our use of further clarifies, we do not quite agree with the court’s view that there has been no erosion). See further, with discussion of vicarious liability, D Airey, “‘The Gig is Up’: The Politics and Policy of Compensating Harms Caused by Gig Workers”, in J Gardner and others (eds), Politics, Policy and Private Law, vol I (2023) 25; Lady Rose, “On the Job? The Law on Status in a Gig Economy”, Denning Lecture, Gray’s Inn, London (7 December 2023). 54 Compare J Gracie, “Vicarious Liability: No Longer ‘on the Move’” (2021) 26 Torts LJ 269 at 275, 278-279. 55 After it was written, the relevance of the discussion in this paragraph was confirmed by Trustees of the Barry Congregation of Jehovah’s Witnesses v BXB [2023] UKSC 15 at paras 59-64. 56 See the careful analysis required to determine who should be sued as representing the crown in X v Y (n 50) at paras 49, 54-67. 57 The trustees accepted liability in D v Bishop’s Conference of Scotland [2022] CSOH 46, 2022 SLT 816. Page 6 of 25 any judgment? Undifferentiated references to “the defenders” do not suffice.58 Even where a quasi-employer is identifiable, it is perhaps more likely than in standard situations to be an unincorporated association, the uncertain status of which as a defender could in Scotland be taken up by well-advised, properly funded litigants.59 After the identification of a wrongdoer’s potential quasi-employer, comes the question whether she stands in a relationship akin to employment with a wrongdoer. Scots law is yet to commit to a particular pattern of inquiry. It could usefully embrace the following. What is required is an examination in detail of the features of the defenders’ relationship “that are similar to, or different from”, one of employment. 60 Focus is on these, rather than just outward appearances.61 The relationship is akin to employment if it possesses the same fundamental qualities as those inherent in an employment relationship,62 so that if the dichotomy is factually apposite, 63 the wrongdoer more closely resembles an employee than an independent contractor.64 Thus the presence of the normal indicia of employment, or something close to them, is key.65 Relevant features of a working relationship between a wrongdoer and quasi-employer include: “[W]hether the work is being paid for in money or in kind, how integral to [an] organisation is the work carried out by the [wrongdoer], the extent of the [second defender]’s control over the [wrongdoer] in carrying out the work, whether the work is being carried out for the [second defender]’s benefit or in furtherance of the aims of [an] organisation, what the situation is with regard to appointment and termination, and whether there is a hierarchy of seniority into which the [wrongdoer’s] role fits.”66 58 Compare BXB v Watch Tower and Bible Tract Society of Pennsylvania [2020] EWHC 156 (QB) at para 1, [2021] EWCA Civ 356 at para 2, [2020] 4 WLR 42, in which the governing body agreed to satisfy any judgment against the trustees. 59 See Campbell and Lindsay (n 9) 217-219. There perhaps existed extra-legal reasons not to dispute vicarious liability in, e.g. T v The English Province of the Congregation of Christian Brothers [2020] SC EDIN 13, 2020 SLT (Sh Ct) 108; and a later case involving the same defender, counsel, and solicitors: AB [2022] SC EDIN 7. 60 BXB (n 55) at para 58(ii); Barclays (n 15) at para 27, stating that this was also key in CCWS (n 41), Cox (n 41), and Armes (n 6). The court in Barclays (n 15) at paras 18, 23, demonstrated this for CCWS and Armes, but not Cox (though see Cox (n 41) at paras 32-37). See also the efforts made in BXB (n 55) at paras 36, 43, 46. 61 See Hughes (n 23) at paras 86-88, though query the views there expressed that (1) the phrase “recognisably independent business”, used in Cox (n 41) at paras 24, 29-30, means a business “recognisable to someone with no knowledge of the contractual arrangements between the tortfeasor and the defendant”; and (2) Barclays (n 15) marks some sort of reversion to “focus … on the contractual arrangements between tortfeasor and defendant”. Compare Barness v Ingenious Media Ltd [2019] EWHC 3299 (Ch), [2020] PNLR 10 at paras 75-79; SKX v Manchester City Council [2021] EWHC 782 (QB), [2021] 4 WLR 56 at paras 45-46, 48; and, semble, Barclays (n 15) at para 22. 62 Ng Huat Seng v Munib Mohammad Madni [2017] SGCA 58, [2017] 2 SLR 1074 at para 63; quoted in Barclays (n 12) at para 26 to demonstrate a more general point. Note that a relationship akin to employment may be closer than an employment relationship: CCWS (n 41) at paras 56-58; and a non-commercial entity may be vicariously liable: Cox (n 41) at para 30. On vicarious liability for volunteers, see further P Morgan, “Recasting Vicarious Liability” [2012] CLJ 615. 63 See TVZ v Manchester City Football Club Ltd [2022] EWHC 7 (QB) at paras 310-311; Carr v Brands Transport Ltd [2022] EWHC 3167 (KB) at para 343, 357. 64 TVZ (n 63) at paras 240-246, 256, 259. 65 Blackpool Football Club Ltd v DSN [2021] EWCA Civ 1352 at paras 60, 62, 122, 127-128. 66 BXB (n 55) at paras 58(ii), 66-67; TVZ (n 63) at paras 261-263, 313-320: proportions of time wrongdoer spends with quasi-employer’s operation and elsewhere, shouldering of profit-loss-risk, duties of obedience, extent and nature of control over wrongdoer, wrongdoer’s involvement with core of quasi-employer’s operation; Barclays (n 55) at para 28: burden of administration, framing of tasks, payment structure, freedom to refuse work, who insures, whether purported quasi-employer is closer to being one in a portfolio of clients than an employer. On top of at least a more then vestigial degree of control over what the wrongdoer does, and a real degree of integration Page 7 of 25 In the usual manner, assistance is derived from parallels with, and general guidance in, previous authorities.67 The existence of quasi-employment may be determinable based on authority which applies more or less directly to a set of arrangements between defenders, examined in detail.68 Absent such authority, the matter is determinable by the use in broader fashion of authorities yielding incidental factual comparisons and general guidance from principle,69 such as that just quoted. So far, so good. Difficulty commences with the statement in a Supreme Court decision on an English appeal that in so-called “doubtful cases”, certain policy factors “may be helpful in identifying a relationship which is sufficiently analogous to employment to make it fair, just and reasonable to impose vicarious liability”70 First, the factors highlighted are unhelpful. Secondly, no policy reasoning of any type is necessary to the decision of quasi-employment cases.71 In the first place, the factors referred to by the Supreme Court are “policy reasons that usually make it fair, just and reasonable to impose vicarious liability on the employer when these criteria are satisfied”: “(i) the employer is more likely to have the means to compensate the victim than the employee and can be expected to have insured against that liability;[72] (ii) the [delict] will have been committed as a result of activity being taken by the employee on behalf of the employer; (iii) the employee’s activity is likely to be part of the business activity of the employer; (iv) the employer, by employing the employee to carry on the activity will have created the risk of the [delict] committed by the employee;[ 73 ] (v) the employee will, to a greater or lesser degree, have been under the control of the employer.[74] […] [These] incidents of the relationship between employer and employee … make it fair, just and reasonable to impose vicarious liability on a defend[er]. Where the defend[er] and the [wrongdoer] are not bound by a contract of employment, but their relationship into the quasi-employer’s operation, the court in Blackpool (n 65) at paras 66-69, 72, 78, 82, 101-104, 122-128 endorsed as a relevant factor the assignment of tasks to the wrongdoer which increases the risk of wrongdoing. The assignment of tasks may be at the root of some relevant considerations. But risk of wrongdoing is not itself a relevant criterion: BXB (n 55) at para 69. 67 Blackpool (n 65) at paras 104, 129, 131-135; TVZ (n 63) at paras 258-263, 302-303; BXB (n 55) at para 68. 68 We necessarily oversimplify here. See K Llewellyn, The Common Law Tradition (1960) 77-91, famously cataloguing sixty-four ways precedent may be handled or affect an instant case. 69 As in Carr (n 63) at para 361. 70 Barclays (n 15) at paras 14-18, 27; discussing CCWS (n 41). See also Blackpool (n 65) at paras 99-101; BXB (n 55) at para 58(iv): “difficult cases”. 71 By policy, we mean “substantive justifications to which judges”, and indeed other lawyers, “appeal when the standards and rules of the legal system do not provide a clear resolution of a dispute”: J Bell, Policy Arguments in Judicial Decisions (1983) 22-23. For further complexities, see, e.g. P Jaffey, “Policy and principle and the character of private law” (2020) 11 Jurisprudence 387. 72 On the weight of deep pockets and insurance, compare Cox (n 41) at para 20; with Armes (n 6) at paras 56, 63; and see BXB (n 55) at paras 42, 82. 73 On factors (ii), (iii), and (iv), “the beating heart of the policy assessment” (S Eggleton (2023) 139 LQR 193 at 197), see Cox (n 41) at paras 23-24, 29-31; and on the “core idea” encapsulated in these factors, see BXB (n 55) at paras 42, 47, 58(iv), 69, 82, also noting that factors (i) and (v) have “little, if any, force”. 74 On this factor, see Cox (n 41) at para 21; and on factors (ii)-(v), see Blackpool (n 65) at paras 77-78, 82, 101-104, 135. Page 8 of 25 has the same incidents, that relationship can properly give rise to vicarious liability on the ground that it is akin to that between an employer and an employee.”75 As the introductory and closing words of this passage confirm, the five factors are set up as policy reasons for having employers’ vicarious liability in the first place. A decision whether the contours of the doctrine should expand to take in a new situation from which an employment relationship proper is absent, but in which the core normative concerns of employers’ vicarious liability are present, is supposed to be assisted by an understanding of these reasons. 76 That such an understanding provides such assistance may, however, be doubted. Factor (i), deep pockets and insurance, explains why one might target a particular defender at all.77 It cannot be used to differentiate relationships of quasi-employment (or, indeed, employment) from those of principal and independent contractor. The same objection applies to factor (ii), causation of wrongdoing through action on another’s behalf, and factor (iv), creation of the risk of wrongdoing. These are unconcerned with any particular kind of relationship between a wrongdoer and a party sought to be made vicariously liable. Further, factors (ii),78 causation, (iii), integration into business activity, (iv),79 risk creation, and (v), control,80 will likely be present in any situation in which arguments of vicarious liability are contemplated. In many circumstances worth litigating – and supposedly meriting analysis through a policy lens because hard to analyse – perhaps only a little effort is needed to bring these factors out.81 Query, therefore, their potential to assist in cases posing special difficulty. In the second place, if the factors thought helpful by the Supreme Court are largely unfit for purpose, it is fortunate that they – and policy reasoning generally – are unnecessary. On the best account of the English authorities, a consideration of whether policy reasoning can be used in a quasi-employment case proceeds as follows. Such reasoning is not strictly82 permissible where a detailed examination of the parties’ relationship with the aid of precedents determines to the court’s satisfaction whether a relationship akin to employment exists.83 If not, policy reasoning is permissible.84 Unmistakeable instances aside,85 a clear example of a case in which policy reasoning is needed involves “a pattern of relationships with legal and factual elements which differ from those found in any of the authorities to which [the court is] referred”.86 At a measure of distance from this lie (1) from a legal perspective, “circumstances which have not 75 CCWS (n 41) at paras 35, 47 (reference omitted); and see the endorsement of loss distribution at para 34. 76 Barclays (n 15) at paras 16-18, 27; Ng Huat Seng (n 62) at para 62. 77 See CCWS (n 41) at para 34. 78 Blackpool (n 65) at para 102. 79 Armes (n 6) at para 77 (Lord Hughes, dissenting): “will in practice apply to virtually all situations in which A asks or authorises B to deal in some manner with C”. 80 DJ v Barnsley MBC, Sheffield County Court, 13 August 2021, at paras 34-35, 57-58: “business activity, risk, control and means [i.e. deep pockets] are more or less written into every fostering situation”. 81 See the admirably frank warning in the clear case of TVZ (n 63) at para 327: “[i]t is deceptively easy to apply [policy reasoning] in a way that leads to a conclusion that [vicarious liability] is fair, just and reasonable”. 82 The word “generally” has, however, been used, reflecting the “never say never” attitude which must prevail (R v Bottrill [2002] UKPC 44, [2003] 1 AC 449 at para 26) among appellate judges: Cox (n 41) at para 41; BXB (n 55) at paras 58(iv), 82. 83 SKX (n 61) at paras 50-54. In emphasising that this is a matter for the court, we do not unrealistically suggest that counsel would forbear to argue that hers is a doubtful case if she thinks she properly can, even knowing she may very well not succeed. 84 TVZ (n 63) at para 279(4). 85 Carr (n 63) at para 362: “unquestionable[]”; SKX (n 61) at para 53: “classic”; DJ (n 80) at paras 23, 33, 65: a doubtful case “that would delight the most sadistic law examiner ever born”; agreed with when the decision was affirmed on appeal [2023] EWHC 1815 (KB) at para 39. 86 JXJ v The Province of Great Britain of the Institute of Brothers of the Christian Schools [2020] EWHC 1914 (QB) at paras 136, 144. Page 9 of 25 previously been the subject of an authoritative judicial decision”;87 and (2) from a factual perspective, situations in which a pursuer can “establish the essential minimum of the employee relationship … but where uncertainty remains whether” the defenders’ relationship “is sufficiently analogous to that of a conventional employment relationship”.88 The matter is one of degree. The further removed a case from the clear example, the more caution is required in allowing policy reasoning to influence its resolution.89 This picture is familiar to lawyers acquainted with the use in Scotland and England of policy reasoning in the law of negligence when the question arises whether a person is under a duty of care.90 Where authority already gives an answer, the wisdom of the answer in policy terms has already been considered. It need not be re-considered unless a court is invited to depart from decided cases.91 Ordinarily, only where established principles provide no answer does the question become whether one should develop the law incrementally and by analogy with established authority, so that a duty of care is imposed.92 Particular consideration is given to (i) whether so doing would be fair, just, and reasonable;93 (ii) the needs to avoid incoherence and inappropriate distinctions; and (iii) the potential consequences of extending existing doctrine.94 Though identifying established categories of case may be challenging,95 it is clear that any move beyond them,96 starting with extensions of the same and not just the recognition of 87 Cox (n 41) at para 42, which was “such a case”, in that – as we read Lord Reed – the proposition that the ministry-prisoner relationship could give rise to vicarious liability was undecided. Note that Lord Reed considered policy by reference to the fair, just and reasonable construct (ibid at paras 42-44), not the five policy factors from CCWS (n 41) which we are here discussing. This is unsurprising. He saw the latter as “criteria for the imposition of vicarious liability”, reflecting “the various policy justifications for its imposition”: Cox (n 41) at para 41. So they had already informed his discussion of the details of the defenders’ relationship: see ibid at paras 32-39 (“the requirements laid down in [CCWS] are met”); Barclays (n 15) at paras 16, 20-22; BXB (n 55) at paras 42-44. 88 P Giliker, “Can the Supreme Court Halt the Ongoing Expansion of Vicarious Liability? Barclays and Morrison in the UK Supreme Court” (2021) 37 PN 55 at 61-62. See also DJ v Barnsley MBC [2023] EWHC 1815 (KB), [2023] PIQR P17 at paras 34, 39. 89 TVZ (n 63) at paras 327-328. It should not, of course, be thought that English judges will never mention policy if they decide that it is not strictly engaged. Most obviously, they may address it obiter (for appellate acknowledgement that this may be wise, see Benyatov v Credit Suisse (Securities) Europe Ltd [2023] EWCA Civ 140, [2023] ICR 534 at para 32) lest their initial assessments be held wrong on appeal. See Blackpool (n 63) at para 129; TVZ (n 63) at paras 219, 257, 321, 327; MXX v A Secondary School [2022] EWHC 2207 (QB) at para 200; affirmed because no sufficiently close connection between wrongdoing and relationship between defenders, but appellant abuse survivor’s contention accepted that the judge misapplied to the facts her correct statement of the law relating to quasi-employment [2023] EWCA Civ 996 at paras 64-75. Recognition that counsel on both sides had fully argued the point may explain the brief mention of policy in Hughes v Rattan [2021] EWHC 2032 (QB), [2022] 1 WLR 194 at paras 101-102, 131 (unaddressed when the Court of Appeal disagreed with the judge on vicarious liability in a lengthy obiter dictum: (n 23)). 90 Adiukwu v Secretary of State for the Home Department [2020] CSIH 47, 2021 SC 38 at paras 7 (Lady Dorrian), 53 (Lord Glennie), 69 (Lord Woolman, concurring), citing and summarising authority. On the relationship between the approach set out here and assumption of responsibility, see Benyatov (n 89). For wider discussion, see J O’Sullivan, “Negligence, Policy and ‘Novel Exceptions’”, in J Gardner and others (eds), Politics, Policy and Private Law, vol I (2023) 105. 91 Robinson v Chief Constable of West Yorkshire Police [2018] UKSC 4, [2018] AC 736 at paras 26, 42 (Lord Reed, with whom Lady Hale and Lord Hodge agreed). 92 N v Poole Borough Council [2019] UKSC 25, [2020] AC 780 at para 64. 93 Royal Bank of Scotland International Ltd v JP SPC 4 [2022] UKPC 18, [2023] AC 461 at para 80. 94 Robinson (n 91) at paras 27, 29, 69(1)-(2) (Lord Reed). 95 Ibid at paras 85-87 (Lord Mance). For disagreement in an apex court, see Rankin (Rankin’s Garage & Sales) v JJ [2018] SCC 19, [2018] 1 SCR 587 at paras 18, 28 (Karakatsanis J, for the majority), 71-76 (Brown J, with whom Gascon J agreed). 96 A question of degree: Reeman v Department of Transport [1997] PNLR 618 (CA) at 625. Page 10 of 25 entirely new ones, involves policy reasoning.97 The closer one is to facts on which a duty of care has been recognised, the more easily the necessary policy assessment tips in favour of further development.98 The parallel is plain between the “authority then policy” approach in the English akin to employment inquiry, and that applicable to the Anglo-Scots duty of care question. But the analyses are distinguishable. In negligence, the courts consider it possible for a case to be incapable of decision “in accordance with an established line of authority”, so that policy reasoning is necessary.99 The necessity arises, it is said,100 because a judge can be unable to construct from the cases a normative statement applicable to a fact situation in such a way as suffices to say whether a duty of care exists as a matter of law.101 By contrast, it is always possible to dispose of a quasi-employment case based on ample existing authority. There may be no precedent on all fours.102 But at the very least, there is available a suite of indicia of employment against which to test the details of the relationship between a wrongdoer and a person sought to be made vicariously liable.103 A pertinent distinction between duty of care and quasi-employment situations thus lies in what (it is thought) the authorities can do. In novel negligence cases, the headline question is whether persons generally should owe a duty of care on a given fact pattern.104 There is no conceptual limit to the considerations which may bear on this. Liability “for negligence may be argued in an almost unlimited range of circumstances”, and “[p]articular cases can always raise questions which cannot be covered by principles or policies of relatively general application”.105 So one can well understand how authority may run out of steam, rendering policy necessary.106 In quasi-employment cases, the headline question is more granular: do comparatively fact-focussed indicia of employment, and relevant individual decisions, support the conclusion that this relationship is akin to one in which a wrongdoer serves a business in subordinated, 97 Darnley v Croydon Health Services NHS Trust [2018] UKSC 50, [2019] AC 831 at paras 15-16, noting that it is unnecessary for “the precise factual situation” to have been “been the subject of a reported judicial decision” to exclude policy: “[i]t is sufficient that the case falls within an established category in which the law imposes a duty of care”; Robinson (n 91) at para 84 (Lord Mance). 98 James-Bowen v Commissioner of Police of the Metropolis [2018] UKSC 40, [2018] 1 WLR 4021 at para 23. 99 Commodity Solution Services Ltd v First Scottish Searching Services Ltd [2019] SAC (Civ) 4, 2019 SC (SAC) 41 at paras 31, 33, amid an exemplary application of the approach described in the preceding paragraph; Glasgow City Council v First Glasgow (No 1) Ltd [2019] CSOH 101, 2020 SLT 75 at paras 39-40 (but mistakenly implying that policy is inaccessible when analogising from established categories, despite the statement quoted ibid at para 37 from Robinson (n 91) at para 29); Hughes v Turning Point Scotland [2019] CSOH 42, 2019 SLT 651 at paras 81, 93-94; compare A v B Ltd [2022] CSOH 34, 2022 SLT 577 at paras 17-21. 100 We leave aside to what point judges handle precedents and decide cases as they purport to. For two discussions, of two angles, among many, see D Kennedy, “Freedom and Constraint in Adjudication: a Critical Phenomenology” (1986) 36 J Leg Ed 518; B Leiter, “Legal Realism and Legal Doctrine” (2015) 163 U Penn LR 1975. 101 L Duarte D’Almeida, “What is it to Apply the Law?” (2021) 40 Law and Philosophy 361 at 371-372, 383-385. 102 C Michelon, “The Uses of Precedent and Legal Argument”, in T Endicott and others (eds), Philosophical Foundations of Precedent (2023) 117 at 118-123. 103 See those set out in, and in the text to, n 66, above. The courts regularly direct themselves according to generalised guidance in other contexts, such as contractual interpretation disputes: Paterson v Angelline (Scotland) Ltd [2022] CSIH 33, 2022 SC 240 at para 32; Ashtead Plant Hire Co Ltd v Granton Central Developments Ltd [2020] CSIH 2, 2020 SC 244 at paras 9-17. 104 Winfield & Jolowicz (n 4) paras 5.012, 5.016; Clerk & Lindsell (n 51) paras 7.08, 7.15-7.30, stating that a duty of care “applies to a general class of relationship and damage” before discussing how to determine its existence. 105 M Armitage (gen ed), Charlesworth & Percy on Negligence (15th edn, 2022) paras 2.02 (“the potential scope of negligence as a basis for legal liability is virtually unlimited”), 2.48, 2.58. 106 S Deakin and Z Adams, Markesinis and Deakin’s Tort Law (8th edn, 2018) 94-95. Page 11 of 25 dependent fashion under another’s control? 107 Perhaps, then, because what bears on the existence of quasi-employment is ordinarily more concrete,108 the authorities cannot run out of steam. So not only is policy redundant, there is no clear start-stop point for its use. Yet it has not been suggested that it is relevant in every case of this kind. It will be clear that we would exclude policy reasoning from the akin to employment analysis.109 But we are not fantasists. One cannot physically prevent pleaders and judges from mentioning policy. And like it or not, they will employ policy.110 So, based on the foregoing, here is how they should – if they must. First, the Supreme Court’s five factors should be discarded. Their replacement should be the fair, just and reasonable construct, borrowed from the law of negligence. It leaves room to identify and argue based on policy considerations worthy of serious discussion.111 Secondly, and especially since policy reasoning is strictly unnecessary, it should be discussed openly, and it should influence exceptionally. It should feature alongside as full an account of relevant authority as possible. And at most, it should serve to prevent absurd or grotesque outcomes.112 Against the above backdrop, the few Scottish cases addressing quasi-employment do not disquiet. The earliest, and the only one finally to determine liability, is 2018’s Grubb v Shannon.113 Shannon leased premises which she fitted out and ran as a beauty salon with two other therapists and a hairdresser under a brand which she devised. One therapist, Higgins, worked from the room otherwise used by Shannon. She contributed modestly to rent, but retained all income from customers. She was generally present on the same three days each week, with keys and authority to open and lock up. She had access to the salon’s Facebook account to solicit custom for the overall concern, which she further aided by appearing in publicity imagery and attending a third party wedding fair in branded clothing with the other therapist, whose arrangement with Shannon for another room was similar. All three beauty therapists offered a set range of treatments at agreed prices; promotions or discounts were in the end up to Shannon. Higgins’ beauty therapy was a central part of Shannon’s main business activity, carried on over an extended period, subject to substantial restrictions on pricing, and a more than minimal degree of control as to what she did as a self-employed person. She effectively stood in Shannon’s shoes when working, going about assigned activities which carried a risk of injury through carelessness. The pursuer reacted allergically to the eyebrow tint which Higgins applied without conducting an initial skin test, or inquiring about allergies or experiences with similar treatments. Though not Higgins’s employer – and though, for example, she paid Higgins no money, provided no equipment, and collected no tax or national 107 See Uber (n 15) esp at paras 73-75 (referring to employees and workers). The notion of dependent labour runs through the discussion in Z Adams and others, Deakin and Morris’ Labour Law (7th edn, 2021) ch 2. 108 Compare the indicia set out in, and in the text to, n 66, above, with the more abstract discussion of policy and typical concerns in Clerk & Lindsell (n 51) paras 7.26-7.27; C Witting (ed), Street on Torts (16th edn, 2021) 44-46. 109 This would be to go further than the court in BXB (n 55) at paras 44, 58(iv), 82, which observed that policy is unnecessary “in the vast majority of cases”. 110 J Thomson, “Principle or Policy? The Judicial Development of the Law of Delict” [2003] CLP 123 at 136-138, 150-151; J Thomson, Scots Private Law (2006) 5-8. 111 On the tacit knowledge lawyers have of what arguments are likely proper, see, e.g. J Bell, “The Acceptability of Legal Arguments”, in N MacCormick and P Birks (eds), The Legal Mind: Essays for Tony Honoré (1986) 45; S Swaminathan “Analogy Reversed” [2021] CLJ 366 at 380-385; S Swaminathan, “Common Law” in M Siems and PJ Yap (eds), The Cambridge Handbook of Comparative Law (2024) 235 at 245-247. 112 We thus disagree with the court in BXB (n 55) at paras 44, 58(iv), 82 that it is even a “useful final check” in cases which are merely “difficult”. 113 (n 10). Page 12 of 25 insurance – Shannon was vicariously liable for Higgins’s negligence based on a relationship between them akin to employment. Drawing on English learning,114 Sheriff Reid held that: “Ms Higgins did indeed carry on activities, entrusted to her by the defender [Shannon], as an integral part of the defender’s business activities, and for the benefit of the defender; that her activities were not attributable to the conduct of a recognisably independent business of her own, or of a third party; and that the negligence which forms the basis of the action was a risk created by the defender by assigning those activities to Ms Higgins.”115 This followed discussion using the Supreme Court’s five factors as convenient headings.116 The second, third, and fifth were taken as reflecting longer-standing guidance on the identification of employment. 117 Along with the “recognisably independent business” admonition118 – a different,119 but clear,120 way to state that vicarious liability does not arise from the relation of principal and true independent contractor – the second to fourth factors are echoed in the statement of principle which Sheriff Reid repeatedly emphasised, and closely tracked in his just-quoted conclusion.121 He understood himself to be examining whether Shannon’s relationship with Higgins exhibited “characteristics similar to those found in employment […] such as to make it fair, just and reasonable” for it “to give rise to vicarious liability”,122 without also standing back and considering policy as a separate question.123 So it is unsurprising that the complete quintet of factors merely provided a framework within which to assess many factual details of the defenders’ relationship124 by extensive reference to case law. An overt consideration of risk-creation would likely not feature in Sheriff Reid’s note were he writing it today.125 He might further adopt a different structure to render unambiguous the conformity of his reasoning with more recent authority.126 Nevertheless,127 though Sheriff Reid used the Supreme Court’s five factors to give form to his thinking, this did not produce a 114 See Campbell and Lindsay (n 9) 203-206, noting that only one Scottish case, Marshall (n 49), was cited for the pursuer on the akin to employment point, but was not drawn upon by, or before, the court as it consciously (Grubb (n 10) at paras 66, 73) extended Scottish doctrine. 115 Grubb (n 10) at para 112. 116 Ibid at paras 86-111. 117 Ibid at paras 72, 89, 96-97, 106; citing E (n 7) at para 72. 118 We doubt Sheriff Reid’s understanding of this phrase: see n 61, above, and Grubb (n 10) at para 100 (original emphasis): “recognisable to a reasonable person or objective bystander standing in the shoes of the injured claimant”. But his analysis went beyond features of the defenders’ relationship knowable to third parties. 119 J Lee, “The Supreme Court, Vicarious Liability and the Grand Old Duke of York” (2020) 136 LQR 553 at 555. 120 Barclays (n 15) at paras 20-22, 24; Ng Huat Seng (n 62) at para 64; D Nolan, “Reining in Vicarious Liability” (2020) 49 Industrial LJ 609 at 612, 613-614. 121 Grubb (n 9) at paras 68-71 (“backbone” characteristics), 94, 96, 100, 112; Cox (n 41) at para 24. 122 Grubb (n 9) at para 63; mis-citing Cox (n 41) at para 24 for Armes (n 7) at para 54. See also Grubb (n 10) at para 67: “What, then, are these ‘characteristics’ of an employment relationship?” 123 See his reference to “a number of important ground rules to be observed before one embarks upon an analysis of the [defenders’] relationship”, the fifth of which was that “it will not always be necessary to ask the broader question, namely whether it is ‘fair, just and reasonable’ to impose liability. The ‘whole point’ of seeking to align the five criteria with the various policy justifications for its imposition [is] to procure a result that [is] inherently fair, just and reasonable”: ibid at para 70; quoting Cox (n 41) at para 41. 124 Details which were decisive in the judgment originating the list of policy reasons: CCWS (n 41) at paras 56-60; Blackpool (n 65) at para 62. 125 See the relatively brief discussion in Grubb (n 10) at paras 103-104. Based on our previous work, we assume that the relevant warning on the English appeal in BXB (n 55) at para 69 will come to be treated as authoritative: Campbell and Lindsay (n 9) 209-210. 126 See especially Barclays (n 15) at paras 27-28; BXB (n 55) at para 58. 127 With the rest of this paragraph, compare Reid (n 52) para 3.29. Page 13 of 25 particularly errant analysis in substance. Rather, it assisted with examining the defenders’ communings in detail. And despite certain infelicitous turns of phrase,128 there is more textual support in his note for the view that he did not – rather than that he did – unduly downplay the principle that vicarious liability does not ordinarily arise from a relationship of principal and true independent contractor.129 Developments subsequent to Grubb v Shannon pose no difficulty. In Kennedy v Bonnici,130 the court avoided confusion into which the pleadings might have led it and concluded that no question of quasi-employment arose. The pursuer offered to prove employment proper, and the trustee’s obligation to meet liabilities incurred by a predecessor was unrelated to vicarious liability. The Supreme Court’s decision in Barclays Bank Plc v Various Claimants131 was relied upon by counsel and noticed by the court. So, too, in X v Y.132 That case concerned the liability of the Crown for assaults and harassment by a Sheriff, and directly engaged the relevant principles. The opinion shows real effort at detailed examination of the defenders’ relationship, including statutory underpinnings; the position as to payment, appointment, removal, and guidance – if not supervision – by the Crown; and the significance of the principle of judicial independence. Cases said to be relevant were also addressed. It may be noted that policy was not brought to bear in this novel case. The main reason for this is likely that the central question – answered negatively – was whether the pursuer was bound to fail to establish a relationship akin to employment. However, policy’s being left out of account lends at least a little support to our view that one can very well decide quasi-employment cases without it. Grubb v Shannon may require to be read a little charitably in light of subsequent developments, which have continued the smooth embrace in Scotland of English learning as it has appeared.133 However, if pleaders and judges are receptive, the approach which we advocate to identifying quasi-employment and policy reasoning can readily bed down in due course. C. LINKING WRONGFUL CONDUCT TO THE DEFENDERS’ RELATIONSHIP If a relationship giving rise to vicarious liability is established, focus turns from the first stage (“Stage One”) to the second stage (“Stage Two”) of the basic inquiry: the link between that relationship and the wrongdoing defender’s conduct. The analysis is complicated by the sheer variety of fact patterns on which it must be conducted.134 We start with the generally accepted test, before considering whether any modification to that test is merited in specific contexts. (1) Close connection: the standard test 128 Grubb (n 10) at para 101, where Sheriff Reid said that the therapists were “self[-]employed” and “independent contractors”, but independent contractors “to whom activities forming core parts of the defender’s enterprise were assigned or entrusted”. This language is less problematic when read with Scottish appellate authority for the proposition that one may be vicariously liable for the negligence of an independent contractor who is so integrated into one’s business as effectively to be part of one’s workforce: Marshall (n 49) at 121, 125-126; discussed Campbell and Lindsay (n 9) 203-206. 129 See the immediately preceding footnote, and nn 61, 118-120, above, on the expression “recognisably independent business”. 130 (n 50) at paras 28, 42-44, 52-56; citing Barclays (n 15) for general propositions but not applying it as the court was concerned with a purported employee-wrongdoer. 131 (n 15). 132 (n 50) at paras 13, 18, 22-32; obviously applying Barclays (n 15), cited in argument. (We say detailed in the main text, bearing in mind that that the question was simply whether to allow proof.) 133 Campbell and Lindsay (n 9) 208 fn 95. 134 Dubai Aluminium Company Ltd v Saalam [2002] UKHL 4, [2003] 2 AC 366 at paras 24-26; BXB (n 55) at para 33. Page 14 of 25 In assessing whether a legally significant connection between wrongdoing and relationship exists, Scots law has followed English law in adopting the “close connection” test. Recent Inner135 and Outer House authority,136 taking its lead from the UK Supreme Court,137 has treated as authoritative the formulation of Lord Nicholls of Birkenhead: the wrongful conduct must be so closely connected with the acts the partner or employee [or quasi-employee] was authorised to do that, for the purpose of the liability of the firm or the employer [or the quasi-employer] to third parties, the wrongful conduct may fairly and properly be regarded as done by the partner while acting in the […] course of the firm’s business or the employee’s employment [or the quasi-employee’s quasi-employment].138 As well as being used to give content to this test, English case law has been cited freely by the Scottish courts in its application.139 There also is evidence of the inverse:140 the two most significant Inner House decisions since that of the House of Lords in Lister v Hesley Hall Ltd,141 Wilson v Exel UK Ltd142 and Vaickuviene v J Sainsbury Plc,143 have been cited approvingly by English courts in factually similar inter-employee violence cases.144 Few are the Scottish applications of the close connection test which surprise.145 135 Shaw IH (n 3) at paras 8, 5, 17 (counsel’s submission taking BXB (n 55) as authoritative), 18, 22 (court accepting counsel’s submissions in law and citing BXB (n 55)). 136 Shaw OH (n 3) at para 26; X v Y (n 50) at para 33. 137 Wm Morrison Supermarkets plc v Various Claimants [2020] UKSC 12, [2020] AC 989 at para 25 (Lord Reed PSC) (“Morrison II”); BXB (n 55) at para 70 (Lord Burrows). Before this, Lord Nicholls’ formulation only was referred to in passing in earlier Scottish decisions: Wilson v Exel UK Ltd [2010] CSIH 35, 2010 SLT 671 at para 8; Vaickuviene v J Sainsbury Plc [2013] CSIH 67, 2014 SC 147 at para 23. 138 Dubai Aluminium (n 134) at para 23. The bracketed additions and omissions reflect the “minor adjustments” proposed in BXB (n 55) at para 58(iii) (Lord Burrows): first, the inclusion of quasi-employment situations; secondly, the removal of “ordinary” before “…course of the firm’s business…”. 139 In addition to the cases discussed below, see e.g. Kelly v Gilmartin’s Executrix 2002 SC 602 (OH) at para 46, comparing the facts with Lister v Hesley Hall Ltd [2001] UKHL 22, [2002] 1 AC 215; Somerville (n 21) at para 30, comparing the case to Graham v Commercial Bodyworks Ltd [2015] EWCA Civ 47, [2015] ICR 665; Oil States Industries (UK) Ltd v S [2022] CSOH 52, 2023 SC 209 at paras 91-93, referring to Petrotrade Inc v Smith [2000] Lloyd’s Rep 486, noting that it pre-dated Lister (n 139) and Wilson (n 137). Pre-Lister, see eg Bell v Blackwood Morton & Sons Ltd 1960 SC 11 (IH) at 24-25 (following Tunney v Midland Railway Co (1865-66) LR 1 CP 291); at 12 (following Nottingham v Aldridge [1971] 2 QB 739); Thomson v British Steel Corporation 1977 SLT 26 at 30 (distinguishing Nottingham v Aldridge); Ward v Scotrail Railways Ltd 1999 SC 255 at 263 (drawing comparison with Keppel Bus Co Ltd v Ahmad [1974] 1 WLR 1082 (PC, Singapore)). 140 See additionally Trafalgar Multi Asset Trading Co Ltd v Hadley [2023] EWHC 1184 (Ch) at para 584, referring to Oil States Industries (n 139); JXH v The Vicar, Parochial Church Council and Churchwardens of the Parish Church of Holcombe Rogus [2023] EWHC 3221 (KB) esp at paras 150-153, 165-167, 200(i), discussing Shaw IH (n 3). 141 (n 139). 142 (n 137). 143 (n 137). 144 Both Wilson (n 137) and Vaickuviene (n 137) are cited approvingly in Graham (n 139) at paras 18-20. Wilson is cited in Weddall v Barchester Healthcare Ltd [2012] EWCA Civ 25 at paras 36-40, 47, 50 and in Chell v Tarmac Cement and Lime Ltd [2020] EWHC 2613 (QB) at para 17 (affirmed on appeal, with no reference to Wilson (n 137), but relying heavily on Graham (n 139), at [2022] EWCA Civ 7). 145 See, arguably, Sharp v Highlands and Islands Fire Board [2005] CSOH 111, 2005 SLT 855, obiter: tackle on firefighting instructor by recruit during “traditional” end of course football match said to be closely connected to recruit’s employment; Robertson v Scottish Ministers [2007] CSOH 186: harassment of co-worker by shop steward arguably closely connected with shop steward’s employment by prison service; no discussion of potential liability of trade union. Page 15 of 25 However, it is noteworthy that a good deal of judicial consideration of Stage Two has taken the form of obiter comments following conclusions as to prescription or limitation, or after findings of a lack of a wrongdoing,146 or of no sufficient relationship at Stage One.147 Other judicial consideration has been prompted by arguments that pleadings lack specificity or relevancy. Such consideration proceeds on the basis of facts asserted fairly generally within written pleadings. While it has been noted that the close connection test requires “a judgment that is critically dependent on enquiry”, 148 there clearly are cases where stated facts so obviously disclose no sufficiently close connection that a case may disposed of without a proof. The result of all this is an understandable lack of depth in discussion of the close connection test, which has in Scotland received relatively little appellate attention.149 Scots lawyers routinely mine richer seams of case law south of the Border where doctrinally appropriate.150 At first glance, there is much to be gained from consulting discussion of Stage Two by English judges, who decide more cases and must apply an identically formulated test.151 However, dangers may attend the tracking of developments in a larger neighbouring jurisdiction. Of these, we highlight two, in the hope that future development here of the close connection test will be untrammelled by them. First, reception of the English approach may be incomplete. Without more, this would not necessarily be bad. But blending learning from older Scots authorities with the newer close connection test could produce a defective mix. Such a thing could happen slowly and non-deliberately. Common law regimes most often change incrementally. Instead, this danger has manifested itself in attempts by the Inner House actively to graft the so-called “Salmond test” onto the close connection test. In the twentieth century, the Salmond test was an appealing formula for linking wrongful conduct to the relationship between defenders giving rise to vicarious liability.152 An employer would be liable if it directly authorised an employee’s wrongdoing, or if the employee’s conduct constituted a “wrongful and unauthorised mode of doing” some authorised act.153 This second limb extended to unauthorised acts, provided these were “so connected with acts which [the employer had] authorised, that they [could] rightly be regarded as modes – although improper modes – of doing them”.154 146 See eg Campbell v The University of Edinburgh, unreported, 14th May 2004 (Temporary Judge RF MacDonald QC); Sharp (n 145); Shields v Crossroads Orkney [2013] CSOH 144, 2014 SLT 190; Scottish Water Business Stream v Automatic Retailing (Scotland) [2014] CSOH 57; Taylor v Quigley [2016] CSOH 178. 147 eg The Royal Bank of Scotland v Bannerman Johnstone Maclay 2003 SC 125 (OH). 148 Royal Bank of Scotland Plc v Bannerman Johnstone Maclay [2005] CSIH 39, 2005 SC 437 at para 68 149 In addition to Wilson (n 137) and Vaickuviene (n 137), see Bannerman (n 148); Philip v Highland Council [2018] CSIH 53 at para 38, and, most recently, Shaw IH (n 3) at paras 18-23. 150 We discuss this in the context of Stage Two in the twenty-first century in Campbell and Lindsay (n 9) at 206-214. 151 See BXB (n 55) at para 58(iii). 152 P Giliker, “Comparative Conclusions”, in VLCLW (n 9) 221 at 229-230; Vaickuviene (n 137) at paras 18-21, noting “particular enthusiasm” by reference to older cases. 153 J Salmond, The Law of Torts: a Treatise on the English Law of Liability for Civil Injuries (1st edn, 1907) 83 (original emphasis); Schokman (n 2) at para 63 (Edelman and Steward JJ). 154 Ibid at 84; referred to as “arguably a third limb” of the test by D Tan, “Enterprise Risk and Vicarious Liability in Singapore”, in VLCLW (n 9) 121 at 126; compare Vaickuviene (n 137) at para 18 in fine: “a gloss on the extent of the liability for a wrongful mode of doing an authorised act and not a separate third category of liability”. Page 16 of 25 Though thought to license a fairly broad approach,155 the Salmond test dealt poorly with intentional acts by employees.156 In Lister,157 the Appellate Committee drew upon the “so connected” gloss to the second limb of the Salmond test to finesse the applicable test into one of close connection.158 Notwithstanding the adoption of the Lister test into Scots law, the Inner House – through Lord Carloway – has evinced a desire to maintain the Salmond test when assessing closeness of connection. In Wilson, Lord Carloway stated that the Salmond test applied “[w]ithin the context of the broad [close connection] test”, and that the need for the employee’s acts to be within the “scope of his employment” was a “fundamental principle”.159 In Vaickuviene, it was said that, despite Lister, “the test remained one of whether the actings of the employee were ‘within the scope of his employment’”.160 The close connection test was considered a simple “variant of the Salmond formulation”. 161 The idea, then, is that the Salmond test assists in finding or denying the necessary close connection in more run-of-the-mill cases.162 There are signs that this synthesised approach may still find favour at first instance. For example, in Oil States Industries (UK) Ltd v S,163 the Lord Ordinary stated in applying the close connection test that the “sufficiency of the connection may be gauged by asking whether the wrongful actings can be seen as ways of carrying out the work which the employer had authorised”.164 But for the most part, the synthesised approach has not been followed or mentioned in Scottish cases, whether pre-Wilson,165 or post-Vaickuviene.166 This suggests that it is neither easily extracted from Lister, nor intuitive generally to pleaders and judges having to apply legal principles to concrete facts. In England, it has been said rhetorically that the 155 See, e.g. Ilkiw v Samuels [1963] 1 WLR 991 (CA) at 1004 (Diplock LJ, citing Salmond’s The Law of Torts (13th edn, 1961) 122 and Canadian Pacific Railway Co v Lockhart [1942] AC 591 (PC, Can) at 599; see also the judgment of Willmer LJ at 997, to which Diplock LJ further cross-referred). 156 P Giliker, “An Introduction”, in VLCLW (n 9) 1 at 8. 157 (n 139) para 15-16 per Lord Steyn (seeing Salmond’s gloss as the “germ” of the close connection test, but cautioning against the mechanical application of the former); para 20 (Salmond test involves a “simplistic” question, and a broad test considering the nature of the wrongdoer’s employment avoids “pitfalls of terminology”). Lord Hobhouse noted that (at para 60) that the “Salmond test is a convenient rule of thumb which provides the answer in very many cases but does not represent the fundamental criterion…”. Lord Millett, at para 70, stated that the Salmond test was “not a statutory definition” and that “precise terminology [was] not critical. What [was] critical [was] that attention should be drawn to the closeness of the connection between the employee’s duties and his wrongdoing…”. cf Lord Clyde, for whom, at para 37, the “unauthorised mode/authorised act” test was one way of establishing the necessary connection, but who, like Lord Millett, noted that the “critical element” is the “necessary connection between the act and the employment”. 158 Mohamud (n 5) at paras 39, 56; Shaw OH (n 3) at para 24, calling the Salmond test “the previous test”, and suggesting the close connection test’s origins lie in Bazley v Curry [1999] 2 SCR 534 (SCC) and Jacobi v Griffiths [1999] 2 SCR 570 (SCC). See also BXB (n 55) at para 30; citing Lister (n 139) and Bazley. 159 Wilson (n 137) at para 25 (Lord Reed agreeing). 160 Vaickuviene (n 137) at para 19 (Lord Brodie and Lord McGhie generally agreeing). 161 Ibid at para 32. 162 This seemingly accords with what Lord Clyde said in Lister (see (n 157)), but Lord Clyde still thought the the core test to be one of close connection. No other Law Lord in Lister conceived of a retained role for the “unauthorised mode/authorised act” formulation. See e.g. para 69 per Lord Millett (close connection test has “advantage of dispensing with the awkward reference” to unauthorised modes). 163 (n 139). 164 Ibid at para 90. Sufficiency should not be taken as a term of art. Otherwise, this formulation falls into the trap, discussed below, of following Mohamud despite its subsequent “explanation” in Morrison II (n 137). 165 See, e.g. Bannerman Johnstone Maclay (OH) (n 147) at para 76 (Lord Macfadyen): “I approach the matter on the basis that the test of whether an unauthorised wrongful act committed by an employee is nevertheless within the scope of his employment is whether the wrongful act is so closely connected with his employment, that is with what he is authorised by his employer to do, that it would be fair and just to hold his employers vicariously liable despite their being innocent of the wrongdoing.” 166 See Campbell and Lindsay (n 9) at 209 fn 104. Page 17 of 25 Salmond test is the starting point in employers’ vicarious liability cases.167 But it is more often mentioned only in framing the close connection test, 168 or in the course of historical observations.169 We have only traced one English post-Lister case in which the “unauthorised mode” / “authorised act” formula was applied dispositively.170 The judges’ central focus is ordinarily on the close connection test, without any such overlay.171 Appellate courts have substituted first instance analyses hewing closely to the Salmond test with their own, conducted with the broader close connection test.172 We already have explained that Lord Carloway’s synthesised test is unfortunate.173 Placed as it is within the close connection test, the Salmond element neither adds to the latter nor alters the outer limits of liability. It is thus redundant. The test’s recent re-emergence in Oil States,174 albeit brief, prompts further comment. The employee in that case accepted a bribe and in return determined a procurement process in a particular way which favoured the payer. The employee was said to be “carrying out work he was authorised to do, but in an unauthorised way”, so that the close connection test was satisfied on the facts.175 To conceive of the acceptance and receipt of a bribe as carrying out work the recipient was authorised to do stretches the concept of “authorised act” so wide as to render it valueless as a touchstone of vicarious liability. If undiscovered, and principles of count, reckoning and payment were not brought to bear, the bribe would remain in the employee’s pocket as a result of his independent diversion of the proper process. Taking the bribe seems less like the performance of an authorised act in an unauthorised manner than a frolic of the employee’s own.176 The result in Oil States is more intuitively explicable on the basis of the close connection test. The Salmond “unauthorised mode” / “authorised act” analysis also obscures matters in cases of employee theft. Admittedly, the Scottish courts previously have analysed such a fact pattern as involving “an ill way of executing the work which has been assigned to [the employee]”.177 However, as was noted by Lord Millett in Lister: “[s]tealing a client’s property cannot sensibly be described as an unauthorised mode of dealing with it on her behalf”.178 Misappropriation of property by an employee is so inimical to her responsibilities of custody and care that it is ridiculous to speak of such a deviation’s amounting to an unauthorised mode of discharging 167 Graham (n 139) at paras 8-9, immediately acknowledging that the close connection test was applicable by reference to Lister (n 139) and Dubai Aluminium (n 134). 168 See, e.g. BXB (n 55) at paras 1-7 esp 2. 169 See, e.g. those by Lord Toulson in Mohamud (n 5) at paras 25-39. 170 See, semble, Pattihis v Jackson [2002] EWHC 2480 (QB) at paras 34-35 (relying on a pre-Lister (n 139) edition of Clerk & Lindsell on Torts (18th edn, 2000)). 171 See, e.g. Mattis v Pollock [2003] EWCA 887, [2003] 1 WLR 2158 at paras 18-19, 30-32; Attorney General for the British Virgin Islands v Hartwell [2004] UKPC 12, [2004] 1 WLR 1273 at para 16; Gravil v Carroll [2008] EWCA Civ 689 at para 21. 172 See eg Bernard v Attorney General of Jamaica [2004] UKPC 47 at para 22 (Lord Steyn); Brown v Robinson [2004] UKPC 56 at para 12 (Lord Carswell). 173 Campbell and Lindsay (n 9) at 209. 174 (n 139). 175 (n 139) at para 92. 176 See BXB (n 55) at paras 74-76; Morrison II (n 137) especially at para 47 (Dubai Aluminium (n 134) at para 32); Wilson (n 137) at para 30, 32-34; Vaickuviene (n 137) at para 28, 37. The exercise of separating frolics – the pursuit by the wrongdoer of her own interests – from misguided attempts by the wrongdoer to further the interests of a person sought to be made vicariously liable, is often useful in straightforward situations (less simple are ones involving, e.g. sexual abuse or the delict of fraud). 177 Central Motors (Glasgow) v Cessnock Garage and Motors 1925 SC 796 (IH) 802 (Lord Cullen); followed in Walker v Scottish & Newcastle Breweries 1970 SLT (Sh Ct) 21. 178 Lister (n 139) para 74. Page 18 of 25 authorised responsibilities. And where a third party steals with the assistance of an employee, it is the more absurd to think in such terms.179 All this underlines the point that the Salmond test “was not effective for determining the circumstances in which it was just to hold an employer vicariously liable for [its employee’s] committing an act not authorised by the employer.”180 Upon it, the close connection test may be considered an improvement,181 because it lays bare the “evaluative judgment”182 obviously in play when analysing the association of wrongdoing with what a wrongdoer is supposed to be doing. Retaining some form of “unauthorised mode” / “authorised act” yardstick within the close connection assessment risks undermining the justifications183 for adopting such an open-ended test: effectiveness born of transparency. In addition, the evaluative “fairly and properly”184 formula renders otiose any separate place for policy reasoning at Stage 2, even as a “final check on the justice of [an] outcome”.185 However, as discussed in Section B(2), above, we acknowledge that some use of policy reasoning is likely inevitable, undesirable though this may be. A second danger, related to the first, is that developments south of the Border could go untracked, with the result that English cases refining or applying the close connection test in a now-discredited manner are still employed uncritically in Scotland. It is, for example, regrettable that straightforward reliance upon Mohamud v Wm Morrison Supermarkets Plc (Mohamud)186 should continue, despite its subsequent treatment in Various Claimants v Wm Morrison Supermarkets Plc (Morrison II). 187 Mention in the former of a “sufficient” connection was explained in the latter as an understandable compression following a fuller exposition of the relevant law; the test remained one of “close” connection.188 The statement that the employee’s motive was “irrelevant” was not a generally applicable legal principle; it was a factual observation made after the legal test had already been satisfied.189 This amounts to quite a substantial redrawing of Mohamud, without, of course, explicitly190 overruling the decision. Therefore, reliance on Mohamud without – at least – cross-checking against Morrison II is inadvisable. English judges have grounded their more recent analyses in the latter 179 cf The Huntsworth Wine Company v London City Bond [2021] EWHC 2831 (Comm) at para 175 (close connection test potentially satisfied in such circumstances). 180 Mohamud (n 5) at para 56 (Lord Dyson MR). See similarly Schokman (n 2) at para 93 (Gleeson J). 181 Mohamud (n 5) at para 56. 182 Dubai Aluminium (n 134) at paras 24-26; BXB (n 55) at para 33; Shaw IH (n 3) at para 18. Though Lord Nicholls in Dubai Aluminium (n 134) at para 24 used the phrase “value judgment”, we have preferred the expression “evaluative judgment” from his Lordship’s para 26, as the former phrase is omitted from Lord Reed’s lengthy quotation in Morrison II (n 137) at para 22; and Lord Burrows’s discussion in BXB (n 55) at para 33. 183 Having argued here that the law should at least retain transparency, space constraints oblige us to leave for another day whether fact patterns such as those centring on employee theft are more profitably seen as involving primary liability for the breach of non-delegable duties. See further Schokman (n 2) at paras 53, 70-81 (Edelman and Steward JJ). 184 BXB (n 55) at para 58(iii). 185 The latest way in which the use of policy has been countenanced by the Supreme Court in BXB ibid at para 58(iv) and 82. 186 (n 5). 187 (n 137) 188 Morrison II (n 137) at para 25, explaining Mohamud (n 5) at para 45, and cited in BXB (n 55) at para 56. See also Blackpool (n 65) at para 85, stating that this amounted to a “reversal”. 189 Morrison II (n 137) at para 29-30, explaining Mohamud (n 5) at para 48. See also BXB (n 55) at paras 45, 56. 190 Mohamud (n 5)’s Westlaw status still flags it as having had only “Positive/Neutral Judicial Consideration”. See, however, how the case has been interpreted in light of Morrison II (n 137) by eg Giliker (n 88) at 71; Lee (n 119) at 558-559. Page 19 of 25 decision. 191 But there are signs that Scottish judges are treating Mohamud less circumspectly.192 Scottish commentary from 2022 suggests that certain Scottish decisions may now be analysed, or decided differently, in the light of the UK Supreme Court’s 2016 decision in Mohamud.193 The matter of real interest is the extent of the shadow cast by the 2020 Supreme Court decision in Morrison II, as refined in BXB.194 The Scottish courts having adopted the close connection test, its most recent, authoritative formulation, must ground the analysis at Stage Two unless or until it is superseded by another inquiry.195 (2) A “tailored” test? A further pressing question is whether the close connection test is in any way modified to deal with the context of abuse in its various forms.196 Until relatively recently, the rules of limitation and prescription meant that historic sexual abuse cases were almost always cut down at a preliminary stage of proceedings, before detailed consideration of either stage of the vicarious liability inquiry.197 Our scepticism about the courts’ restrictive interpretation of the Limitation (Childhood Abuse) (Scotland) Act 2017 has abated following more recent appellate consideration of the relevant provisions.198 A more permissive limitation regime has, in turn, led to consideration of the approach at Stage Two in a small number of abuse cases. Kennedy v Bonnici199 concerned a challenge to the specificity and relevancy of a pursuer’s pleadings, which alleged that various parties involved in the running of a Benedictine Community School were vicariously liable for abuse perpetrated by a headmaster and lay teachers. Lady Wolffe made reference to the “affirmation” of a tailored test for such cases in Morrison II,200 and denigrated the defender’s submissions for failing to “take into account the more tailored test (as explained by Lord Reed)”.201 Applying that test, Lady Wolffe held, albeit on a summary basis, that the appointment of the teaching staff conferred authority upon them, including to favour or punish students. Subsequent abuse of that authority by staff was therefore closely collected with their relationship with the school’s authorities.202 191 TVZ (n 63) at para 238; Huntsworth Wine Company (n 179) at para 107. 192 X v Y (n 50) at paras 37 (reference to the discredited “sufficient connection” formulation) and 38 (explicitly following Mohamud in taking a “broad view” of the wrongdoer’s duties). See also, possibly, Shaw OH (n 3) at paras 18-19, 21 (counsels’ submissions mentioning “sufficient connection” without direct reference to Mohamud), cf para 26 (Lord Brailsford referring properly to the need for a “sufficiently close connection”). 193 Reid (n 52) para 3.52 fn 191 and 3.63, citing Shields v Crossroads Orkney [2013] CSOH 144, 2014 SLT 190 and Vance v Bough [2008] CSOH 70. 194 Shields (n 139) appears correctly decided in light of BXB (n 55) which, along with Morrison II (n 137), would entail that Vance (n 139) be decided in the same way. 195 BXB (n 55) may of course be expected to become a, if not the, primary point of reference for lower courts in future. The case was treated as authoritative by counsel and the court in Shaw IH (n 3): see n 135, above. 196 See Campbell and Lindsay (n 9) at 210-211. 197 For the limited consideration that exists, see eg Kelly (n 139) at para 46; M v Hendron 2005 SLT 1122 at para 95-96. 198 See Campbell and Lindsay (n 9) at 215-217; and, now, B v Congregation of the Sisters of Nazareth [2022] CSIH 52, 2023 SCLR 116: the death of an abuser will not usually be held to prevent a fair hearing, at least in the context of allegations of a “culture” of abuse. See also GD v Sisters of Nazareth [2023] SC EDIN 27 (All-Scotland Personal Injury Court). 199 (n 50) 200 (n 137) at para 57-58. 201 (n 50) at para 59. 202 Ibid. Page 20 of 25 The pursuer in X v Y203 was a solicitor who allegedly was sexually harassed by a sheriff. She invited the court to bear in mind the different application of the “close connection” test in sexual abuse cases.204 This again was prompted by Lord Reed’s discussion in Morrison II. Like Lady Wolffe in Kennedy, Lord Clark focussed on conferral of authority, this time that which was conferred upon the sheriff over legal practitioners, and which did not end at the courtroom door.205 By the sheriff’s appointment, there was thus more than a mere opportunity to harass the pursuer. Control and authority were, for Lord Clark, key to assessing the sufficiency of the relevant connection in abuse cases.206 His Lordship allowed allegations regarding an assault in a court reception area, and assaults in the sheriff’s chambers and surrounding secured corridors, to go to proof.207 He dismissed as irrelevant averments relating to vicarious liability for an indecent assault on a train, and a video call made to the pursuer after a formal complaint about the sheriff. These “did not fall within even a broad view of [the sheriff’s] functions or field of activities” and did not “show abuse of the special position” conferred upon him by the Crown.208 We query whether this is an apt analysis of the harassing video call.209 This was made following complaint about the sheriff’s conduct on court property and while – as accepted by Lord Clark – exercising authority as a judicial officer. Given that the call was prompted by regulatory oversight of the sheriff’s official conduct, it is not too strained to say that it was closely connected to his assigned functions. The most detailed consideration of the approach at Stage Two in abuse cases – certainly post-Morrison II – came in C and S v Shaw.210 The first defender, Shaw, was employed as a caretaker by the second defender at a sports centre. Prior to appointment, Shaw had befriended the mother of two brothers. Shaw pleaded guilty to a criminal charge in relation to abuse of the brothers at their own home and at his tied caretaker’s house. Additionally, it was alleged that he abused one brother on the premises of the sports centre. The pursuers sought to render the second defender vicariously liable for abuse perpetrated by Shaw within the caretaker’s house, and the alleged abuse at the sports centre itself. It was conceded that the test at Stage 1 was met.211 At first instance, the Outer House found the Stage Two test unsatisfied, and that conclusion was upheld by the Inner House. Between the two courts’ opinions, however, the Supreme Court gave judgment in BXB.212 To that decision we therefore briefly divert, before considering the reasoning at both levels in Shaw. In BXB, the Supreme Court scotched any notion that a different version of the close connection test applies in the sexual abuse context, or that the test operates differently in such cases.213 We discuss below our general agreement with the proposition that no different test should apply in cases of this kind. But we respectfully query the precedential basis on which the Court rejected any so-called tailored test.214 Lord Burrows noted that “special rules for sexual abuse” were 203 (n 50). 204 Ibid at para 19. 205 Ibid at para 38. 206 Ibid at para 39. 207 Ibid at para 40. 208 Ibid at para 41. 209 We also share the view that the judge perhaps took too individuated an approach to the various incidents forming the course of conduct: S Di Rollo KC, Case Note, 2023 SLT (News) 80, 81-82; D Brodie, “Psychiatric harm” (2023) 171 Reparation Bulletin 3. 210 Shaw OH and IH (n 3); Campbell and Lindsay (n 9) 214 and fn 148. 211 Shaw OH (n 3) at para 25. 212 Lord Brailsford recorded counsel’s observation that permission to appeal had been given ibid at para 22. 213 BXB (n 55) at para 58(v). 214 For the terminology, see Morrison II (n 137) at para 36; BXB (n 55) at paras 37 and 58(v). Page 21 of 25 “rejected by Lord Reed in Cox”.215 It appears216 that the reference is to a passage in Cox where Lord Reed is said to have “indicated that the sexual abuse of children is not a special category of case and that the general approach to vicarious liability applies to such cases”.217 Yet Lord Reed was there focussed only on something particular about the analysis at Stage One: It is important, however, to understand that the general approach which Lord Phillips [in CCWS218] described is not confined to some special category of cases, such as the sexual abuse of children. It is intended to provide a basis for identifying the circumstances in which vicarious liability may in principle be imposed outside relationships of employment. […] It results in an extension of the scope of vicarious liability beyond the responsibility of an employer for the acts and omissions of its employees in the course of their employment, but not to the extent of imposing such liability where a tortfeasor’s activities are entirely attributable to the conduct of a recognisably independent business of his own or of a third party.219 It is unsurprising that little of the judgment in Cox speaks to Stage Two: the case involved the other Stage.220 The courts have clearly indicated that each Stage of the vicarious liability inquiry is separate.221 Reference could further be made to another judgment of Lord Reed which did concern Stage Two – Various Claimants v Morrison.222 There, Lord Reed observed that “the close connection test has been applied differently in cases concerned with the sexual abuse of children”,223 referring to “a more tailored version of the close connection test”.224 Of course, as Morrison II did not concern sexual abuse, these comments were themselves obiter. But it was unfortunate for the Supreme Court in BXB not to discuss them, and instead to focus solely on general comments in Cox pertaining to Stage One. Lord Reed concurred with Lord Burrows in BXB. So he, too, must be taken as having rejected a tailored test for, or the tailored application of the general test in, abuse cases. Returning to Shaw, Lord Brailsford in the Outer House explicitly tailored his application of the close connection test, focussing on the authority which had been conferred on Shaw as part of 215 BXB (n 55) at para 58(v). His Lordship here acknowledged that “one can reasonably interpret some judicial comments as supporting” those special rules, but did not elaborate. 216 Ibid at para 44, citing Cox (n 41) at para 29. 217 BXB (n 55) at para 44. 218 (n 41) at paras 35, 47; cited in Cox (n 41) at para 28. 219 Cox (n 41) at para 29 (emphasis added). 220 There is some brief reference to stage two by cross-reference to the “complementary” judgment, handed down contemporaneously in Mohamud (n 5): Cox (n 41) paras 1-2, 15, the court observing, nevertheless, that the appeal before it was “concerned with the first question”: “what sort of relationship has to exist” for vicarious liability to arise? 221 See eg Morrison II (n 137) at para 31 (rejecting the relevance of the five CCWS policy factors to stage 2); Personnel Contracting (n 16) at para 191 (Gordon J: “[t]o the extent that a fact or matter may be considered relevant to both questions, not only is that fact or matter considered for a different purpose in answering each question but the weight to be attached to that consideration is likely to be different”); Bird (n 50) at para 78. 222 Morrison II (n 137). 223 Ibid at para 23. He went on to explain, referring to Lord Phillips’s judgment for a seven-member court in CCWS (n 41) at paras 83, 85, that the different approach emphasised “the importance of criteria that are particularly relevant to that form of wrongdoing, such as the employer’s conferral of authority on the employee over the victims, which he has abused”. Given these statements, we sympathise with the acceptance of a tailored test by the Court of Appeal in BXB (n 58) at paras 64, 84 (Nicola Davies LJ), 92-106 (Males LJ). 224 Morrison II (n 137) at para 36. Page 22 of 25 his caretaking duties.225 As these duties extended only to taking care of the sports centre, its equipment, and ensuring compliance with its rules, they did not confer any authority to have “isolated contact” with users, or be accompanied by someone on his rounds, or to befriend or associate with children. 226 Moreover, Shaw’s association with the brothers pre-dated his employment. No close connection was found. Subsequent to the decision in BXB, the Inner House in Shaw commenced its analysis by affirming that authority was not a “touchstone”.227 However, it emphasised the fact-sensitive nature of the “evaluative judgment” at Stage Two.228 This meant that, in certain cases, authority could be “an important element”.229 Here, the authority abused by Shaw had been conferred upon him by the children’s mother, not by his employer, and the abuse did not arise from the exercise of his duties or occupation of the caretaker’s house.230 The fact that the most intense abuse occurred in the caretaker’s house did not affect that analysis: the “progressive stages of intimacy” had “commenced and progressed substantially” before Shaw’s employment.231 That employment neither “created, nor significantly enhanced” the risk of abuse.232 Our first worry is that the prominence accorded to conferral of authority in this case is perhaps greater than legitimate post-BXB, which cautions against authority-based reasoning, and suggests that steady focus on the general close connection test is preferable.233 Secondly, the Inner House’s reliance on risk-based reasoning at Stage Two234 sits ill with the close connection test as affirmed in Morrison II235 and BXB.236 It may be queried, then, whether the Inner House successfully applied to the facts before it the law which it evidently considered to be authoritative in principle.237 If the post-Morrison II and BXB dispensation is to be taken seriously, it is plain that although Lord Brailsford in the Outer House considered his analysis to be tied to the idea of conferral of authority, his approach is preferable to that of the Inner House. Much of his Lordship’s discussion focusses on the connection between the wrongdoer’s duties and the abuse. This seems an orthodox application of an untailored close connection test, not dependent on the abuse of conferred authority. It remains to note that the other two first instance judgments considered above, following Morrison II but pre-dating BXB, were given on the assumption that the close connection test applied differently in the abuse context. Their outcomes are, however, unobjectionable on this basis. Given its factual similarity to Lister, Kennedy v Bonnici could have been decided in the same way by application of the close connection test. X v Y can be re-analysed in similar fashion, even if conferral of authority figured prominently in Lord Clark’s discussion. 225 Ibid at para 27. 226 His Lordship drew attention to Jacobi v Griffiths (n 158) where, on stronger facts, the close connection test still was not satisfied. 227 Shaw IH (n 3) at para 18. 228 Ibid at paras 18 to 19. 229 Ibid at para 20. 230 Ibid. 231 Ibid at para 21. 232 Ibid. 233 BXB (n 58) at para 71. 234 Shaw IH (n 3) at para 21. 235 Morrison II (n 137) esp at paras 22-23, 25, 30-31. 236 BXB (n 55) at esp at para 58(iii). (The court in BXB ibid at para 69 also expressly disapproved the overt use of risk-based reasoning at Stage One.) 237 See n 135, above. We also agree with the diplomatic assessment of Master Dagnall in JXH (n 140) at para 168, as “slightly puzzling”, the counterintuitive observations by Inner House in Shaw IH (n 3) at para 22 about the relation of the facts in BXB (n 55) to those before it. Page 23 of 25 It therefore seems that the Scottish courts’ use of the so-called tailored test for abuse cases evidences a third danger – related to the second identified in Section C(1), above – in following the case law of a larger neighbouring jurisdiction. Incipient or experimental developments in English law may be adopted prematurely,238 before they are fully fleshed out or, ultimately, knocked on the head. This is not to dissuade the Scottish courts from being independently creative, or formulating distinctive analyses.239 And the Supreme Court in BXB offered no principled reason for rejecting a separate test for abuse. However, that rejection is welcome.240 The facts of X v Y bring to mind the potentially invidious distinctions which might fall to be drawn in demarcating the respective domains of a general test and a tailored one. Does, for instance, a tailored test apply to adults,241 or only children? Does it apply to wrongdoing occurring in the workplace? Does it apply only to sexual abuse, or would it extend to any wrongdoing motivated by sexual gratification? And does it apply outwith situations where responsibility for a pursuer is conferred on a wrongdoer by an employer or quasi-employer? These boundary questions could obscure the analysis which can be conducted simply by focussing on the connection between abuse and the wrongdoer’s duties. D. A SCOTTISH LAW OF VICARIOUS LIABILITY? We are primarily concerned to examine core aspects of the basic vicarious liability inquiry. Finally to be discussed is the relevance to our subject of Scots law’s distinctiveness. We here understand distinctiveness to refer to Scotland’s separate legal system.242 (1) Significant convergence The “doctrine of vicarious liability as it exists in Scotland … today is predominantly English in derivation”.243 At both the relationship and connection stages of the basic inquiry, English learning is frequently cited and influential.244 It has informed Scottish case law on employment and employment pro hac vice.245 It is poised to catalyse and shape the embrace in Scotland of 238 Although, given the comments of Lord Reed in Morrison II (n 137), we sympathise with the temptation of the Scottish courts to adopt a tailored application of the close connection test for abuse cases. 239 See, e.g. in Australia, Schokman (n 2) at para 36, referring to the inquiry, in certain unspecified classes of case, as to whether the wrongdoer’s role “placed him in such a position of power and intimacy that the performance of his role could be said to give rise to the occasion for his wrongful acts such that they could be said to have been committed in the course or scope of the employment”. 240 See also A Silink and D Ryan, “Twenty Years on from Lister v Hesley Hall Ltd – is there now a ‘Tailored Close Connection Test’ for Vicarious Liability in Cases of Sexual Abuse, or Not?” (2022) 38(1) Journal of Professional Negligence 15, particularly at 30-32. 241 As held by the Court of Appeal in BXB (n 58). 242 See especially Scotland Act 1998 ss 28-30; Treaty of Union 1707, Arts XVIII and XIX; DM Walker, “The Union and the Law” (2007) 52(6) JLSS 14; NW Barber, The United Kingdom Constitution: an Introduction (2021) 22-24, ch 15. 243 E Reid and M Loubser, “Strict Liability” in R Zimmermann, K Reid and D Visser (eds), Mixed Legal Systems in Comparative Perspective: Property and Obligations in Scotland and South Africa (2004) 605 at 626-627. 244 Campbell and Lindsay (n 9) at 199-211. See also the apparent approval of party agreement that the Scots and English doctrines of vicarious liability align in JXJ (n 87) at para 120; and the heavy reliance on English sources in the Scottish literature cited in n 52, above. 245 See ibid at 199-200; and on the latter, recognised initially on the basis of English sources, P Simpson, “Vicarious Liability” in K Reid and R Zimmermann (eds), A History of Private Law in Scotland, vol II (2000) 584 at 604-605, also detailing at 588-600, 605-608, other instances in which the legal position north of the Border aligned with that to the south: Scottish judges not all opposed to common employment; following in Scotland of English rejection of Scottish attempts to expand common employment, then of later English cases restricting it; manner of dealing with skilled wrongdoers moving in step. Page 24 of 25 dual vicarious liability.246 It did catalyse the definitive recognition of quasi-employment as a relationship giving rise to vicarious liability.247 And more Scottish authority replicates the English approach to the close connection test than mixes in the Salmond test.248 There may exist in this field an inclination among Scots lawyers to keep pace with their larger neighbouring jurisdiction. Regardless, practitioners seem to envisage that developments south of the Border will be emulated in due course.249 This latter sentiment is discernible in recent sources, amid the BXB litigation250 and subsequent to it.251 The tracking in Scotland of English law looks set to continue as occasions arise, especially so far as concerns moves made by the Supreme Court. Our view remains that after English cases, decisions from other common law countries are most likely to be employed by Scots lawyers looking furth of their own jurisdiction.252 (2) Insignificant divergence Where Scots and English law do not entirely align in this area, conscious effort to achieve variance is unobservable. Three examples; one pre-emptive rejoinder. First, dual vicarious liability remains theoretical in Scotland not because the English approach has been rejected.253 Rather, the opportunity has not arisen to adopt it.254 Lack of authority is not an inherently Scottish phenomenon.255 Secondly, the bastardised amalgamation of the Salmond and close connection tests did not occur because English cases were found wanting. It stems from an honest but strained reading of case law.256 That could quite easily happen elsewhere than in a Scottish court. Thirdly, any omissions north of the Border to track gradual English modifications of the close connection test are unconcerted. If not entirely random,257 they are perhaps simply the product of advocacy which exploits the intricacies of different verbal formulae. That constitutes legitimate and effective pleading in many jurisdictions.258 Finally, it might be thought that we forget our earlier suggestion in Section B(2), above, that Scots law 246 Campbell and Lindsay (n 9) at 200-203. 247 See ibid at 204-206; and Grubb (n 10). 248 See discussion in Section C(1), above; Campbell and Lindsay (n 9) at 209 fn 104. 249 Ibid at 209-210. 250 BXB (nn 58 (High Court and Court of Appeal) and 55 (Supreme Court)). See, e.g. Shaw OH (n 3) at para 22, where Lord Brailsford, summarising argument, mentions the appeal to the Supreme Court; and notes from both a firm with a specialist abuse practice and a senior Advocate, discussing and citing the Supreme Court’s decision without hinting that it does not, strictly, determine Scots law: K Wade, “The Case that limits Vicarious Liability?” Brechin Tindal Oatts Blog, 1 June 2023, available at https://tinyurl.com/caselimitsvicariousliability1; Di Rollo KC (n 209) at 81-82. 251 BXB (n 55) was treated as authoritative by counsel and the court in Shaw IH (n 3) at paras 8, 15, 17 (counsel), 18, 22 (court accepting submissions in law and citing BXB (n 55)). 252 Campbell and Lindsay (n 9) at 211-214. Subsequently, see, e.g. mention of Canadian authority and its discussion by Lord Brailsford in Shaw OH (n 3) at paras 22, 24, 29-30; and counsel in Shaw IH (n 3) at paras 13(4), 16. 253 Compare Clancy v A, B, C and D [2022] NSWCA 119 at paras 197-198. 254 Campbell and Lindsay (n 9) at 200-203, noting that low-level Scots authority against the notion has been sidelined. 255 See D Ryan, “Connecting Cautiously: Vicarious Liability in Ireland in Comparative Perspective”, S Todd, “Vicarious Liability in New Zealand”, and P Giliker, “Comparative Conclusions”, all in VLCLW (n 9) at 103, 171-172, 238-239 (also noting less deference in certain smaller jurisdictions than in Scotland to English learning), 243-244. 256 See M Campbell, “Somerville v Harsco Infrastructure Ltd: Transferred Intent and the Scope of Vicarious Liability” (2016) 20 Edin LR 211 at 213-16, questioning how the synthesised test appears from Lord Steyn and Lord Clyde’s speeches in Lister (n 139), as suggested in, e.g. Wilson (n 137) at paras 26-27. 257 Counsel and solicitors were all different in Kennedy (n 50); Shaw OH (n 3) and IH (n 3); and X v Y (n 50). 258 Including England, and including in vicarious liability cases: Giliker (n 88) at 61, discussing the phrase “doubtful case”. Page 25 of 25 should and can take a different approach to policy than the one discernible from an examination of English quasi-employment cases. But this is not because we think English law should remain as it is. Indeed, our argument would be unaffected – and some might think this article mercifully shorter – if we simply omitted discussion above of whether our preferred position could be adopted in Scotland. We suggest in the previous Sub-section that Scots law’s distinctiveness is not really being asserted in the realm of vicarious liability. We suggest in this Sub-section that it is irrelevant to the only noteworthy divergences from English law which exist. No unthinking angliciser,259 Lord Cooper stated in 1954 that there “is nothing peculiar to Scots law in the rules relating to the vicarious liability of a ‘superior’ for an ‘inferior’”.260 In 2024, this appears true, not only of the points addressed here where Scots and English law do align, but also of those where they do not. E. CONCLUSIONS To our central discussion, Scots law’s distinctiveness is tangential (Section D). At the relationship stage of the basic inquiry (Section B), general questions usefully mediate the articulation between facts and the indicia of relationships giving rise to vicarious liability. Desirable refinements are identifiable and should be effected, as judges past have encouraged (B(1)). The proper approach to quasi-employment cases is prescribed above. At best, policy reasoning belongs at the outermost margin (B(2)). At the connection stage of the basic inquiry (Section C), Scottish courts should not muddy the waters by relying on outdated manifestations of the close connection test, which they generally have applied satisfactorily (C(1)). They should no longer apply a tailored test in abuse cases, for the decision of which the usual test is adequate (C(2)). 259 For a summary, context, and references, see D Buchan, Diminished responsibility and the Scottish criminal law tradition (LLM(R) thesis, Glasgow, 2022) at 26-32, 35-36 and n 62; available at https://theses.gla.ac.uk/82862/. 260 Macdonald and Hayward (n 27) at 465 (in Macdonald’s case). |
|---|