The liability of an employer for a tort committed by an employee is an area of significant controversy. The problem is particularly acute where the employee’s tort is an intentional one. Various tests and guides have been formulated as Judges struggle to decide whether and why vicarious liability should be imposed in the cases before them. Perhaps the most recent and important of these is the “close connection” test. Skandinaviska, which involved fraudulent schemes by a finance manager to meet his casino debts, provided the Singapore Court of Appeal the opportunity to expound on this challenging area of law. This note explains and comments on this seminal decision.
Vicarious Liability: The Close Connection Test and the Skandinaviska Decision
Introduction
It is trite law that an employer is vicariously liable for a tort committed by its employee in the course of employment. Essentially, vicarious liability turns on two elements: “employee” and “in the course of employment”. The first element asks if the tortfeasor is an employee or an independent contractor; there is no vicarious liability for a tort by the latter.1In the legal development in this area, several tests and concepts have emerged, the foremost of which are the control test, the business integration test, the economic reality test and the multi-factorial approach.2
The second element – “in the course of employment” – is said to be a “most vexed requirement”3and has often resulted in decisions which are inconsistent or irreconciliable, leading observers to suspect that many decisions were really based on the Court’s perception of what justice required.4An early yardstick used is the Salmond test5of authorised and unauthorised act – if the act was authorised but the mode was unauthorised, the act was nevertheless done “in the course of employment”. However, the Salmond test, though generally helpful, had its limitations and various other tests or factors were introduced, including:
1.The degree of excessiveness of the employee’s conduct;
2.The employee’s motive;
3.The scope of the employee’s authority; and
4.Whether the conduct benefited the employer.
Even with all these tests, Courts had difficulty in dealing satisfactorily with cases of intentional torts, such as instances of sexual abuse in welfare homes. In Lister v Hesley Hall,6(“Lister”) the House of Lords held a boarding house vicariously liable for the sexual assaults committed by the warden on children in his care on the basis of the close connection between the employee’s nature of employment and the tort he committed. In so doing, it overturned the Court of Appeal decision in Trotman v North Yorkshire CC,7which had applied the Salmond test and held a school not vicariously liable for the sexual assault committed by its deputy headmaster. In Lister, the House of Lords chose instead to follow the lead of the Canadian Supreme Court in Bazley v Curry.8The advantage of the close connection test, as Lord Millett observed, is that “it would accord with the underlying rationale of the doctrine [of vicarious liability] and be applicable without straining the language [of the Salmond test] to accommodate cases of intentional wrongdoing”.9
Until the Singapore Court of appeal in Skandinaviska Enskilda Banken v Asia Pacific Breweries[2011] SGCA 22 (“Skandinaviska”), it was unclear whether the close connection test applied in Singapore.
Facts of Skandinaviska
In Skandinaviska, Chia, a finance manager in Asia Pacific Breweries (“APB”) obtained credit facilities from several banks through false pretences and forged board resolutions. The banks claimed repayment of the loans on grounds of agency, vicarious liability, negligence and restitution. In the High Court,10the action failed on all four grounds. (This note focuses on vicarious liability.)
In dealing with vicarious liability,11Belinda Ang J relied on the principle in Lloyd v Grace Smith12that the question whether an employee was acting in the course of employment was the same as the question whether he was acting within the scope of his actual or ostensible authority. As she had earlier in her judgment found that Chia had no actual or ostensible authority to bind APB, it followed that Chia’s fraudulent acts were not done in the course of his employment and APB was not vicariously liable. There was thus no need, in her view, to consider the Lister test of close connection.
On appeal, the appellants argued, inter alia, that the principle laid down in Lloyd v Grace Smith was superseded by the close connection test, which if applied to the case would result in APB being vicariously liable.
Judgment of Court of Appeal
Chief Justice Chan Sek Keong delivered the judgment of the Court of Appeal. On vicarious liability, Chan CJ began with a short account of the development of the “close connection” test tracing it from its supposed origin in the Salmond test, to the rejection of the Trotman approach by the Canadian Supreme Court in Bazley v Curry and finally to the House of Lords decision of Lister. His honour noted that the close connection test was subsequently applied in several cases in different jurisdictions.
Singapore’s Position on Close Connection
In the Chief Justice’s view, the close connection test is an “intellectually satisfying and practical criterion” for determining whether vicarious liability should be imposed on an employer for torts committed by an employee “during an unauthorised course of conduct”, whether the wrong was intentional or inadvertent.13He found the test satisfying and practical as it imposes liability only when it is “fair and just” to do so and provides a workable concept.
Also, the test requires a Court to openly confront the question of whether liability should lie against the employer, rather than “obscuring the decision beneath semantic discussions”14of scope of employment and mode of conduct.
As regards what is fair and just, Chan CJ adverted to the two policy considerations of victim compensation and deterrence highlighted by McLachlin CJ in John Doe v Bennett.15
Policy of Victim Compensation
On victim compensation, Chan CJ explained:16
… an innocent victim of an employee’s tort should, under ordinary circumstances, be compensated. In this regard, the employer is usually the person best placed and most able to provide effective compensation to the victim. In our view, making the employer vicariously liable is not only a practical solution, but also fair and just. After all, a person who employs another to advance his own interests and thereby creates a risk of his employee committing a tort should bear responsibility for any adverse consequences resulting therefrom.
His honour pointed out that “the objective of victim compensation rests on the implicit premise that the victim of the tort is not at fault for the tort, or at least bears less fault for the tort than the party who is morally responsible for the tort”.17He further explained why victim innocence is important:18
The main touchstone of legal liability at common law is usually, and justifiably, moral culpability … Vicarious liability, in contrast, operates regardless of whether there is any fault on the part of the person who is ultimately made to provide compensation for the blameworthy party’s wrongdoing … As a form of “strict” liability … vicarious liability is an anomalyin the common law. For this reason, vicarious liability can only be justifiedif the victim of the tort is himself not at fault, or is less at fault than the blameworthy party and/or the ultimate defendant. (Emphasis added).
In short, since vicarious liability does not require fault on the part of the employer, the imposition of such liability can only be justified if the victim was innocent or at least more innocent as compared with the employer and/or the employee.19Otherwise, the policy of victim compensation as a justification for imposing liability, in the words of the Chief Justice, “loses much of its moral force”.20
Policy of Deterrence
On the policy consideration of deterrence, Chan CJ thought the consideration was a legitimate one. The policy consideration, he explained:
… rests on the fundamental premisethat the employer is best placed, relative to everybody else, to manage the risks of his business enterprise and prevent wrongdoing from occurring. (Emphasis added).
Employers are generally in the best or better position to prevent wrongdoing by their employees. He noted that there could, however, be two situations where this may not be so. The first is where the person best placed to prevent the tort may not be the employer but the victim himself or some third party.21The second is where the employee’s tort is “uncontrollable and, therefore, not amenable to deterrence”, as in cases of excessively risky enterprises, spur-of-the-moment torts and intentional torts. Chan CJ opined:22
In such situations, it may well be possible to find that the employer has done all that is reasonable to deter the tort and yet has failed to prevent the commission of the tort. In such situations, deterrence as a justification for imposing vicarious liability loses much of its force.
The sentiment expressed here is that if imposing vicarious liability does not serve the purpose of deterrence, Courts should be reluctant to impose liability.
Other Policy Considerations as Counterweight
Chan CJ also pointed out that the applicability of the policy considerations of victim compensation and deterrence does not inevitably lead to the conclusion that the employer is vicariously liable, and vice versa. There could be other policy considerations which could act either as a “counterweight” to victim compensation and deterrence or, it is implicit from the judgment, as a reason or additional reason to impose liability.23
Chan CJ was cognizant of the discomfort some Judges had with making decisions based on policy considerations of what is fair and just, but pointed out that English Courts had thus far been unable to forge a legal rule without reference to policy considerations.24
Relation with Spandeck Test of Duty of Care
He also observed that, conceptually, the close connection test is closely related to the Spandeck test for duty of care, and added that a claim (based on vicarious liability) which fails the close connection test is not likely to pass the Spandeck test in a claim on negligence, and vice versa, as both tests take into the account the criterion of foreseeability of harm occurring to the victim. Indeed, in Skandinaviska itself, the appellants’ claim failed on vicarious liability as well as negligence.25However, Chan CJ cautioned that this does not mean that a finding of negligence can never be made where a claim in vicarious liability fails, and vice versa, as the two doctrines, though closely related, are not identical.
He was also careful to add that the close connection test is a guide and not a rigid definition of circumstances of liability.26
Factors in Applying the Close Connection Test
Chan CJ referred approvingly to the factors laid down in Bazley v Curry for ascertaining whether there was, in the circumstances of the case, the requisite closeness of connection:
1.The opportunity the enterprise afforded to the employee to abuse his (or her) power;
2.The extent to which the wrongful act may have furthered the employer’s aims;
3.The extent to which the wrongful act was related to inherent risks of friction, confrontation or intimacy;
4.The extent of power conferred on the employee in relation to the victim; and
5.The vulnerability of potential victims to wrongful exercise of the employee’s power.
Application of Law to Skandinaviska Facts
Applying the legal framework that he had outlined, Chan CJ gave five reasons why APB was not vicariously liable for Chia’s acts of deceit:27
1.The functional connection between Chia’s employment as finance manager and Chia’s fraudulent scheme was “illusory” since his position gave him very limited financial authority. Chia had no power to source for credit facility or to borrow money on the company’s behalf;
2.APB could not have reasonably contemplated that Chia might defraud a third party28which he had not authority to deal with as finance manager; Chia’s fraud was “entirely unforeseeable”;
3.The policy consideration of victim compensation did not avail as Skandinaviska was in a much stronger financial position than APB and was not a vulnerable victim. Neither was APB an innocent victim as much of the blame for the successful perpetration of Chia’s fraud lay with APB;
4.As for deterrence, it cut both ways in the present case but there was a greater need for banks as compared with trading companies (such as APB) to adopt prudential measures. Chia’s fraud could have been prevented by the bank taking elementary measures, such as contacting APB’s directors to verify that APB accepted the credit facilities. Imposing vicarious liability on APB may create “an unacceptable moral hazard” in that banks may be encouraged to take only minimal precautions against fraud; and
5.Finally, Chan CJ found that none of the five Bazley v Curry factors were present except for the first (opportunity afforded by enterprise to employee to abuse his power) but even then the connection was a tenuous one.
In view of the above, it was not fair and just to impose vicarious liability on APB. In short, there was no close connection between Chia’s fraud and his employment, and the policy justifications for imposing vicarious liability were absent in the present case.
Comments
The Skandinaviska is significant on several counts.
First, and most obviously, it declares that the close connection test applies in Singapore and to all torts, whether intentional or unintentional.
Second, it makes it clear that apart from close connection there is a need to consider the policy considerations of victim compensation and deterrence. Victim compensation is premised upon victim innocence whilst deterrence is based upon the comparative ability to deter wrongdoing. If the premises upon which the considerations are based do not obtain in the case at hand, then it would not be fair and just to impose vicarious liability.
Third, Chan CJ’s judgment introduced a new and important factor or consideration of the employer foresight of the employee’s tort.
However, each of these in turn raises questions or doubts.
First, what is the status of all other tests previously used? Are the Salmond test, the “frolic” test and other tests to be totally discarded, even though each of them may be useful in certain contexts? Or are they to be assimilated into or subsumed under the general concept or rubric of “close connection”? Chan CJ’s judgment does not give a clear indication, although his honour’s caution29that the close connection test is a guide and not a rigid definition may suggest that pre-existing tests may usefully be resorted to in appropriate situations.
Second, do both considerations of victim compensation and deterrence have to be satisfied before vicarious liability can be imposed or is it sufficient that one is satisfied? Here, again, there is no clear indication from the judgment. It should be observed, however, that victim compensation and deterrence are two primary goals of tort law and there is no discernible judicial approach or philosophy of requiring both these goals to be satisfied in order for tort liability to be imposed. Logically, so long as one of the two goals is satisfied, it should be permissible, policy-wise, to impose liability.
Third, the role of foreseeability in vicarious liability poses the greatest difficulty. It is observed that in Chan CJ’s excursus of the law, there was no mention of foreseeability as a factor for deciding vicarious liability and it was only when he came to the application of the legal framework to the facts of the case that foreseeability emerged. If we examine Chan CJ’s reasoning as to why APB was not vicariously liable, we find that four reasons were given:
1.There was no close connection between Chia’s fraud and his employment: paras [90], [95];
2.APB could not foresee that Chia would commit, and commit successfully, the fraud on the banks: para [91];
3.Victim innocence, the premise upon which victim compensation is based, was not satisfied: para [92]; and
4.The deterrence justification did not obtain as there was a need to hold banks to a higher standard of prudence and responsibility: para [95].
Reading the judgment, one is drawn to the inevitable conclusion that Chan CJ was introducing foreseeability as an additional factor in the determination of vicarious liability, either under the broad concept of “close connection” or as a third requirement30(in addition to “close connection” and “fair and just”).
Assuming this is so, there is a further question of whether foreseeability is a necessary requirement for finding a close connection or whether it is simply an additional factor to be appended to the five factors of Bazley v Curry. If the former interpretation is correct, then vicarious liability is gravitating closer to the realm of fault-based liability. A strong argument may be made that foreseeability should have, at most, a minor role to play in vicarious liability (as one of many factors in ascertaining close connection) and that its true or primary role is in the tort of negligence. It is a matter of conjecture which interpretation is intended by the Court of Appeal.
Concluding Remarks
The Court of Appeal decision in Skandinaviska is to be welcomed. It has endorsed the close connection test and mandated that its application be carried out alongside policy considerations of victim compensation and deterrence. However, as with most legal developments, Skandinaviska brings with it questions and doubts, perhaps the most difficult of which is the role of foreseeability in the determination of vicarious liability. It is indeed a landmark decision, although its full implications will only be appreciated and worked out over time.
► Low Kee Yang *
* LL.B (NUS), LL.M, PhD (King’s College London); Associate Professor, School of Law, Singapore Management University. The author benefitted from a discussion on the point of foreseeability with colleague Gary Chan.
Notes
1 The employer, however, may have primary liability if he is in breach of his own non-delegable duty or if he had authorised the independent contractor’s tort and thus be liable as joint tortfeasor.
2 See for example, Deakin, Johnston & Markesinis,Markesinis & Deakin’s Tort Law (OUP, 6th ed) pp 666-677.
3 Markesinis & Deakin’s, ibidat 678.
4 See Lunney & Oliphant,Tort Law: Text & Materials(OUP 4th ed) p 853.
5 The test first enunciated by J W Salmond in The Law of Torts(Stevens & Haynes, 1907).
6 [2002] 1 AC 215.
7 [1999] LGR 584, CA.
8 (1999)174 DLR (4th) 45. The actual name of the case is a lengthy one - The Children’s Foundation, the Superintendent of Family and Child Services in the Province of British Columbia and Her Majesty The Queen in Right of the Province of British Columbia as represented by the Ministry of Social Services and Housing v Patrick Allan Bazley.
9 [2002] 1 AC 215,245.
10 [2009] SGHC 197.
11 Ibidat [193] – [199].
12 [1912] AC 716; confirmed by the House of Lords in Armagas v Mundogas [1986] AC 717.
13 [2011] SGCA 22 at [75], citing Bokhary PJ in Ming An Insurance v Ritz Carlton [2002] 3 HKLRD 844 at [19].
14 [2011] SGCA 22 at [75], citing McLachlin J in Bazley v Curry.
15 [2004] 1 SCR 436 at [20].
16 [2011] SGCA 22 at [77].
17 [2011] SGCA 22 at [78].
18 Ibid.
19 The requirement of innocence or relative innocence on the part of the victim helps to mitigate the apparent harshness of making an “innocent” employer vicariously liable.
20 Put another way, if a blameworthy or more blameworthy victim is able to sue an employer, then the principle of vicarious liability lacks any moral force.
21 Chan CJ, at [80] gave the example of where the equipment required to perform a job is supplied by an independent contractor.
22 [2011] SGCA 22 at [81].
23 His Honour did not give examples of such other considerations but emphasized the “ultimate goals of fairness and justice must be paramount”: ibid.
24 Ibidat [82] – [83].
25 Ibidat [97] – [105].
26 Ibidat [85].
27 Ibidat [90] – [95].
28 Chan CJ remarked, at [91], that the connection would have been much closer if Chia had defrauded a bank which APB had existing dealings with.
29 [2011] SGCA 22 at [85].
30 If so, then a comparison of vicarious liability and duty of care under negligence yields an approximate mirroring of the elements of foreseeability, proximity and policy.