Lister v Hesley Hall Ltd - Background

The decision of the House of Lords in Lister and others v Hesley Hall Ltd [2002] 1 AC 215 restated the test to determine when an employer will be held vicariously liable for the wrongs of an employee. The case arose out of the sexual assault of a resident of a care home by the manager of the home.

Prior to Lister the test to determine the vicarious liability of an employer had been that set out by Salmond (1907, p. 83). It was said that an employer would only be liable for the employee’s wrongful conduct if it occurred in the course of employment. A wrongful act was within the course of employment if it was either (a) a wrongful act authorised by the employer, or (b) a wrongful and unauthorised mode of doing an authorised act. The first of these situations was unproblematic, primarily because it rarely arose, but over the years, situation (b) gave rise to a number of cases where its application resulted in apparent injustice (see for example Trotman v North Yorkshire CC [1999] LGR 584). The perceived problem with the test as it stood was that intentional wrongdoing, and particularly criminal acts, would rarely render an employer vicariously liable (although an early example is to be found in Lloyd v Grace, Smith & Co [1912] AC 716). As had been pointed out in the Court of Appeal, it was difficult to see sexual abuse of a child as an unauthorised mode of doing an authorised act (Trotman, p. 591 per Butler-Sloss LJ). In the context of Lister, sexual abuse had nothing to do with the employee’s job as a care home manager, even if the employment created the conditions for the abuse.

However, Salmond, in the same section then went on to say that ‘a master… is liable even for acts which he has not authorised provided they are so connected with acts which he has authorised that they may rightly be regarded as modes – although improper modes – of doing them’ (Lister, p. 775 per Lord Steyn). This is the germ of what might be termed the ‘ sufficient connection’ or ‘close connection’ test and can properly be regarded as the ratio of Lister. Provided there is a sufficiently close connection between a wrongful act and the acts the employee was employed to perform, the employer will be vicariously liable.

One can see at a glance that the emphasis of the test has shifted dramatically. The original Salmond test was more limited in scope with its emphasis on authorisation. Only if the conduct were a means (albeit wrongful) of doing an act authorised by the employer would liability arise. The modern approach following Lister is unconcerned with authorisation. Rather, the emphasis is on a ‘close connection’ with employment. Thus the wrongful act need not be a means of performing an authorised act, provided its commission is sufficiently connected to employment (Glofcheski, 2004). One can demonstrate this distinction using the facts of an old Scottish case. In Power v Central SMT 1949 SLT 302 a bus conductor assaulted a passenger following a dispute about payment for a ticket. As the passenger was stepping down from the bus the conductor pressed a bell, thereby signalling to the driver that the exit was clear and that it was safe to pull away. The passenger had not cleared the exit and was thrown to the ground. The Inner House of the Court of Session rejected the passenger’s claim that the conductor’s employers were vicariously liable for her actions. Her actions were motivated by personal spite and could not be viewed a means of performing an authorised act. However, following Lister and the move away from analyses of authorisation, it is almost certain that vicarious liability would be established now. Pushing the bell, which fell within one of the conductor’s usual tasks no matter the motivation for doing so, would be sufficiently closely connected to her employment to establish vicarious liability. Only if she were truly engaged on a ‘frolic of her own’ might vicarious liability be avoided. For example if she left the bus and followed the passenger down the street before assaulting her, one might expect the law to view this as unconnected to employment. However, following one of the cases to be considered below, post-Lister vicarious liability may well be established even in those circumstances.

Lister and the Policy Basis of Vicarious Liability

Stevens makes the point that ‘[t]he central problem in ascertaining the boundaries of…vicarious liability is that it has no settled basis. Unless we know the purpose a rule is intended to serve, it has no natural limits’ (Stevens, 2006, p. 201). To that end, a number of their Lordships in Lister proposed various background reasons why an employer ought to be liable for the wrongdoing of his employees. However, as Lord Clyde observed, ‘A variety of theories have been put forward to explain the rule. The expression “respondeat superior” and the maxim “qui facit per alium per se”, while they may be convenient, do not assist in any analysis’ (Lister, p. 232). That did not stop his Lordship suggesting that there are relevant policy bases for the liability and these may vary depending on the facts of a given case (Lister, p. 235). On the facts of Lister Lord Clyde suggested that the imposition of liability was justified because the employer had been entrusted with the safekeeping of the children and had delegated the performance of that duty to an employee. This is identical to the ‘assumption of responsibility’ argument used by Lord Hobhouse to justify the imposition of vicarious liability. He said: ‘the liability of the employers derives from their voluntary assumption of the relationship towards the plaintiff and the duties that arise from that relationship and their choosing to entrust the performance of those duties to their servant’ (Lister, p. 239). Lord Millet on the other hand argued that vicarious liability is best viewed as a loss distribution device which imposes liability on an employer to protect an innocent party where the employer’s business enterprise created the risk of injury. His Lordship accepted that no liability should arise where employment merely created the opportunity for wrongdoing. However, he also went on to say that an employer ought to be liable for those risks which experience shows are inherent in the nature of the business (Lister, p. 243).

Thus, risk creation, loss distribution, assumption of responsibility and incidental risk are the underlying policy bases for the imposition of liability in Lister and, according to the court, for the imposition of vicarious liability generally. However, as Lord Millett conceded, none of the tests that have been developed are either ‘intellectually satisfying’ or effective to enable the outcome of a particular case to be predicted. Likewise, as Atiyah observed some years ago, the policy bases for vicarious liability cannot be viewed in isolation. Rather, they have to be taken cumulatively with the result that some but not all will be relevant in a particular factual situation. This is to be contrasted with the earlier views of Baty (1916), who believed that the policy bases should be viewed in isolation and could therefore be attacked on an individual basis (Atiyah, 1967, p. 15). Stevens has, however, argued more recently that if no one policy basis is ‘intellectually satisfying’ in the words of Lord Millett, then it is difficult to see how some combination of them is any more satisfactory. (Stevens, 2006, p. 202 (for a very brief history of the development of the rule and its policy justifications see Williams, 1957, p. 228 onwards)). Glofcheski also considers it important that any rule imposing vicarious liability be pitched ‘at the right level of generality’ meaning that it should have ‘a clear policy rationale and justification for a finding of close connection.’ He also suggests that if recent judicial developments are to have the effect of upsetting a century-old understanding of the law then the new test should be the subject of careful consideration. Employers, he says, who are being asked to ‘foot the bill for employees’ unauthorised and insubordinate conduct are entitled to as much’ (Glofcheski, 2004, p. 19). These concerns do not seem to be reflected in the decisions following upon Lister, and in the context of the licensed trade, it will be seen that these developments have had a potentially far-reaching impact.

Lister and Door Stewards

It has been unfortunate that a number of the more prominent recent cases concerning vicarious liability have involved door stewards, particularly at a time when the Government, through the Private Security Industry Act 2001, has attempted to more closely regulate that industry. Probably the most remarkable decision is that of the Court of Appeal in Mattis v Pollock [2003] 1 WLR 2158. A door steward, who was involved in a fight with customers on the premises, stabbed one of those customers in a revenge attack off the premises. The claimant sued the owner of the club vicariously for the assault of the door steward. That claim was dismissed at first instance but was allowed on appeal.

The court began with a review of Lister and the House of Lords decision that followed upon it in Dubai Aluminium Co Ltd v Salaam [2001] 3 WLR 1913. In Dubai a partner in a law firm drafted a series of fraudulent agreements that resulted in the aluminium company making unnecessary payments. His fellow partners were held vicariously liable under the Partnership Act 1890, s. 10 (that case is used as the basis for a proposed classification of vicarious liability cases by Loubser and Reid (2004)). The relevance of that case for the purposes of the present discussion is that the court in Mattis drew heavily upon the speech of Lord Millett. The court in Mattis viewed Dubai as clarifying the central principles of Lister. In particular the court adopted Lord Mustill’s comments where he said that its is ‘no answer to a claim against the employer to say that the employee was guilty of intentional wrongdoing, or that his act was not merely tortuous but criminal, or that he was acting exclusively for his own benefit, or that he was acting contrary to express instructions, or that his conduct was the very negation of his employer’s duty…vicarious liability is not necessarily defeated if the employee acted for his own benefit’ (Dubai, pp. 1941-1942). Further, Lord Mustill opined that vicarious liability may even arise where the act of the employee is an independent act in itself (Dubai, p. 1943). However, as the court in Mattis impliedly realised, the effect of Lister and Dubai has been to render many of the earlier authorities of doubtful validity (Mattis, p. 2166 per Judge LJ). As already noted, it appears that the Scots case of Power is one of them (this concerned is echoed also by Glofcheski (2004), in particular in relation to the extensive body of ‘frolic’ cases). That said, the court emphasised the fact-specific nature of the enquiry. As Judge LJ said, ‘even where an employee behaves violently towards a fellow employee while at work, that is, at his employer’s premises and during working hours, the claim against the employer for vicarious liability may nevertheless fail’ (Mattis, p. 2166). What Judge LJ seems to be contemplating are circumstances so blatantly unconnected with employment that it would be wholly unreasonable to hold otherwise. Using a door steward as an example, this might arise where a steward assaults a customer, not because they are a customer or because of anything done on the premises, but because the customer had earlier crashed into the door steward’s car. To some extent, what Judge LJ said there is somewhat at odds with the decision in Mattis. At its most basic, a door steward who had left work assaulted someone who was no longer a customer off the premises. This was certainly how the court at first instance interpreted the facts. However, the finding of vicarious liability on appeal is less surprising when one considers what Lord Clyde had said in Lister. Since the focus is now on the closeness of the connection, and not on the extent of an employee’s authority, Lord Clyde observed that ‘while consideration of the time at which and the place at which actings occurred will always be relevant, they may not be conclusive’ (Lister, p. 235). This highlights the inherently factual nature of the enquiry. It also supports the concern that vicarious liability may be even more difficult to predict than before.

The court in Mattis took the view that the relevant events should not be regarded as a series of separate incidents. Rather, the close connection between employment and wrongdoing arose because the initial incident started at the employee’s place of work during normal working hours (Judge LJ, p. 2169). This raises some interesting questions, especially when taken with the court’s finding that ‘where an employee is expected to use violence while carrying out his duties, the likelihood of establishing that an act of violence fell within the broad scope of his employment is greater than it would be if he were not’ (Judge LJ, p. 2167). Note the use of the term ‘broad scope of employment’ rather than merely ‘scope of employment.’ This seems to suggest that acts should only be excluded from the scope of employment if totally disassociated, but if they are tenuously connected, then they may give rise to vicarious liability.

Mattis is clearly an alarming development for businesses employing door stewards since prior to Lister the employers would not have been liable. The question would have been: was assaulting a customer outside the premises in a revenge attack an unauthorised mode of controlling access to the premises and maintaining order on the premises? The answer would undoubtedly have been, no. The scope of a door steward’ s employment is generally narrow and restricted to certain basic functions such as controlling entry, searching, ejection and crowd control. That being the case, any test that measured vicarious liability against those functions would necessarily be narrow in range. The Mattis formulation, following Lister and Dubai, is much broader. A practical example may serve to illustrate how far Mattis can be taken.

As a former door steward and trainer of door stewards the writer can attest to the following as a familiar scenario: a door steward manning the door of licensed premises is approached by a drunk seeking entry. Because of his condition he is turned away. The drunk, unhappy at this, threatens to return at closing time to confront the steward when he is off duty. In most cases this is an empty threat and the drunk has long since departed by the time the door steward ends his shift. However, what happens if the drunk is waiting for the steward and attacks him? No doubt the drunk will be no match for a sober door steward and the steward will come off the better of the exchange. But that is exactly the problem. What if the steward strikes the drunk and injures him? Has the steward committed an assault in the course of his employment, or more properly, is there a sufficiently close connection between his employment and his wrongdoing to render his employer vicariously liable? The answer would appear to be, yes. As Lord Clyde said in Lister the fact that events occur off work premises and outside working hours will not be determinative, leaving open the opportunity for liability to arise in circumstances such as these. The connection between employment and wrongdoing arises because the initial incident is work related, even though the drunk never actually gained entry to the premises and even though, technically, he never actually became a customer either. Seen as a loss distribution device this scenario sits reasonably comfortably within the modern conception of vicarious liability. However, if seen as arising on the basis of assumption of responsibility, it seems much less satisfactory. The employer cannot be argued to have assumed a responsibility to someone who never even made it on to the premises. Indeed, the employer would have committed a licensing offence if he had been admitted. But even seen as a loss distribution device, it seems unjust that the employer ought to compensate in these circumstances. His employee was off duty and off the premises. The person assaulted never gained entry to the premises and in fact was refused entry in compliance with licensing law. How then is it fair and reasonable to impose liability? The problem with vicarious liability as a loss distribution device is that it presupposes the existence of public liability insurance in the absence of any express legal requirement to have such cover in place. It further assumes that, even if a policy were in place, its terms are broad enough to include incidents occurring off duty and off the premises. That is perhaps a rather rash assumption. Even if it were not, one can imagine insurers might seek to rely on technical defences to avoid liability since the imposition of liability may by no means be obvious to anyone other than an appellate judge in these circumstances. However, as has already been noted, it is somewhat futile to try and justify the imposition of vicarious liability on one policy ground at a time. It has to be recognised that a range of policies may be at play, unsatisfactory as they all may be. Nonetheless it is to be hoped that Mattis represents the outer limits of the post-Lister test. It seems even Lord Millet, whose speech was so heavily relied upon in Mattis, expressed the hope that this decision was not based on anything he had said in Lister (see the comment in Saggerson, 2004, p. 7). The effect of Mattis seems therefore to be that an employer will be liable for the acts of an employee, even when engaged on a so-call frolic of his own (Vekria, 2004, p. 9).

Vicarious Liability and the Agency Door Steward

In light of the decisions considered above licensees and others employing door stewards have sought to find a means of ensuring that the wrongdoing of staff does not give rise to liability. Applying basic principles, if only an employee can render an employer vicariously liable, the simple solution is to ensure that the member of staff is employed, not as an employee, but as an independent contractor. It is well established that only in exceptional circumstances will an employer be held liable for the torts or delicts of an independent contractor. However, liability has not been so easily avoided. The following cases illustrate how the courts have dealt with attempts to do so.

Temporary Deemed Transfer of Employment and Door Stewards

In Hawley v Luminar Leisure Plc [2006] EWCA Civ 18 a door steward employed by an agency company (‘ ASE’) was supplied to a licensed premises owned by Luminar. The steward seriously assaulted a member of the public outside the premises. It was a term of the contract between ASE and Luminar that no temporary transfer would ever take place. Nonetheless, at first instance, Luminar were held vicariously liable for the door steward’s assault under a temporary deemed (or pro hac vice) transfer of employment. Luminar appealed.

The Court of Appeal essentially affirmed the approach taken by the judge at first instance, which was to say that due to the degree of control exerted by the management of the licensed premises over the door steward, he had become their temporary deemed or pro hac vice employee. Such transfers are extremely rare because the facts rarely support such a finding. They are also unusual because the impact of a transfer is to shift vicarious responsibility from the general employer to the temporary employer. Vicarious liability in such circumstances is therefore doubly strict in that the employer is fixed with a ‘no-fault’ liability for a person he did not directly employ. Thus, the House of Lords in Mersey Docks and Harbour Board v Coggins & Griffith (Liverpool) Ltd [1947] AC 1, the primary authority in this field, laid down a set of stringent criteria to be satisfied before a transfer will be held to have taken place. From the speech of Lord Porter, a number of considerations can be identified. In any case concerning a temporary deemed transfer of an employee, the decision will depend upon the individual facts. However, in assessing those facts, a number of considerations may be relevant. First, the burden of showing that a transfer has taken place is a heavy one. Second, the questions of who engaged the negligent employee, who has the power to dismiss and who pays him must also be considered. Third, the question of who exercises control over the employee is particularly relevant. The proper test is essentially to ask which of the two employers is entitled to control the manner in which the particular act was to be executed. Control therefore relates not just to control of what an employee does, but also how he is to do it. Once it is established that such control exists, it is irrelevant that on the particular occasion where a negligent act occurred, control was not in fact exercised. Fourth, the inquiry should concentrate on the relevant negligent act, and one should then ask whose responsibility it was to prevent it (or put another way, who had power to give orders in relation to how that particular piece of work was done?). Fifth, a transfer can only occur if an employee consents, which effectively means that the two employers cannot contract out of a transfer without the employee’s express permission. And finally, responsibility should lie with the employer in whose act some degree of fault, though remote, may be found (this summary is gleaned from the opinion of May LJ in Viasystems (Tyneside) Ltd v Thermal Transfer (Northern) Limited [2005] IRLR 983, p. 984, discussed later).

Applying this to the particular facts of Hawley a transfer was deemed to have taken place and Luminar were found vicariously liable for ASE’s employee. What was significant was that the door steward, who worked as part of a small team supervised by a head door steward, was accustomed to following instructions from Luminar management. Management instructed the door staff in relation to where to stand, whom to remove and admit and in relation to the general control of the premises. The management effectively regarded ASE’s door stewards as part of its own team of staff. It was also significant that the stewards wore Luminar uniforms and that the head steward was regarded as little more than a figurehead rather than a supervisor or manager in his own right. Notwithstanding the parties’ attempts to contract out of a potential deemed transfer, the facts disclosed that in reality Luminar exercised a sufficient degree of control over ASE’s staff for them to be fixed with liability for their wrongful acts.

There is little doubt that on the facts of this case this analysis is correct. However, one can see that it may well have a significant impact on working practices generally. The circumstances in which the door steward was transferred are by no means unusual. It is not uncommon for a licensee employing agency stewards to give them instruction as to how they are to operate on his premises. Individual premises operate individual dress codes and age restrictions and have individual tolerance levels as regards the behaviour of customers. The core question will ultimately relate to the degree of instruction and control exercised over the agency staff before a deemed transfer will take place. Given the inherent difficulties in predicting its application, the safest means to avoid liability would seem to be to give no instructions whatsoever. That, however, is rather unrealistic. Few licensees would likely allow door stewards a free reign within their premises. One way around this would be to give instructions directly to the management of the agency supplying the door stewards in the expectation that they will relay instructions to staff on the ground. This does, however, seem rather laborious.

The judge at first instance in Hawley considered another argument based on the question of temporary deemed transfer (Wilkie J, reported at [2005] EWHC 5 (QB)). This arose out of the advice of the Privy Council in Kauppan Bhoomidas and Port of Singapore Authority [1978] 1 WLR 189. Here it was suggested that where a member of a team of agency staff acted negligently, there could be no question of a temporary transfer taking place where that person was subject to the control of a foreman or supervisor. Even though the supervisor received detailed instruction from the putative temporary employer, his presence broke the connection between the temporary employer and the negligent employee (per Lord Salmon at 193). Applying this to the facts of Hawley one can see that if the head door steward had truly been acting as a supervisor rather than as a bare employee like his colleagues, no transfer would have occurred. Useful as this approach may be, it is only of relevance where multiple stewards are employed and will not assist where a sole self-employed steward is involved. It will also afford no protection where the supervisor is in fact the negligent employee or is complicit in the negligence of his subordinates. The fact that he has been receiving detailed instructions to relay to other staff may of itself establish that a temporary transfer has occurred. Thus, employers might be tempted to use the contract of employment to avoid the liability. As was established in Mersey Docks employers cannot bi-laterally agree a term of the contract preventing a temporary transfer from taking place. The reason for this is that the party whose conduct creates the vicarious liability is not a party to the agreement. Thus, if the employee can be persuaded to enter a tri-partite agreement to the effect that he will always remain the employee of the general employer, the matter is resolved and a temporary transfer is defeated. The potential difficulties here concern obtaining the agreement of the employee and the general employer to such an arrangement. As regards the employee, if this were a condition of employment or a pre-condition to employment one can hardly see this being a particular objection since it is the employer or their insurers that will ultimately be sued. Notwithstanding the employer’ s supposed right of relief against the negligent employee considered in Lister v Romford Ice and Cold Storage Co [1957] 1 All ER 125, very few if any employers take advantage of it in practice (see, by way of contrast, Williams, 1957, p. 220). The employee has nothing to lose and can have little difficulty with such an arrangement.

From the point of view of the general employer, one could imagine that if the creation of such a relationship was all that stood between a large service contract and no contract at all, they are likely to sign. If anything, their position is more certain in that they know the outer limit of their liabilities and can more easily insure against them. At least, that is what one could have been forgiven for thinking. Seen as a means of circumventing the modern authorities on vicarious liability it might be questioned whether a future court faced with such a contractual arrangement might strike it down on policy grounds. In any case, since the decision in Viasystems (Tyneside) Ltd v Thermal Transfer (Northern) Limited [2005] IRLR 983, there is another aspect to vicarious liability that will also have to be considered (Stevens, 2006, p. 201).

Dual Vicarious Liability and Door Stewards

Sir Patrick Atiyah (1967) considered in his seminal work on vicarious liability that there may be circumstances in which both a general and a temporary employer could be found liable for an employee’s wrongdoing. Indeed, Atiyah considered it strange that the courts had never considered the idea of ‘dual vicarious liability’ to be the obvious solution to the temporary deemed transfer problem. This way, the claimant receives his compensation and the employers fight it out between themselves as to whether one is entitled to an indemnity or contribution from the other (p. 156). However, Atiyah considered it unlikely at the time of writing that the courts would go down this route unless they considered the general and temporary employers to be engaged in a joint enterprise. He thought this unlikely because, notwithstanding the contractual nexus between the parties, they are in effect pursuing independent objectives. That was not to say that two employer could not be liable on common facts. Rather, they would not both be vicariously liable. For example, one might be liable in contract, and the other in negligence. Or one might be vicariously liable, whereas the other might be in breach of statutory duty (pp. 157-158). Notwithstanding Atiyah’s somewhat gloomy prognosis for the development of dual vicarious liability, the Court of Appeal in Viasystems was persuaded that there are indeed circumstances in which such dual vicarious liability will arise.

The facts were that an independent contractor was employed to install air conditioning in a factory. The contractor sub-contracted the work to an installation company. The installation company then sub-contracted with a third company for the provision of fitters and mates on a labour-only basis. A fitter’s mate, under the supervision of a fitter from his company and an employee of the installation company negligently caused a flood. The legal question was: were his employers, the installation company, or both, vicariously liable for his negligence? In other words, was liability confined to either the general employer or the temporary employer, or could both be held liable concurrently? The answer was, both.

At first instance it had been decided that no transfer had taken place on an application of the principles set out in Mersey Docks. The Court of Appeal however was persuaded to take the analysis a stage further. May LJ considered the early authorities and although these had been taken to mean that dual liability was impossible, he could find no real support for that approach. The cases that seemed to suggest that only one employer could be vicariously liable never really considered the point directly and dealt with it on the basis of an assumption. The one case that did deal with the point directly, Laugher v Pointer (1826) 5 B & C 547, was decided at a time when the courts were extremely concerned to avoid multiplicity of actions, ‘ an objection which modern procedure does not find unduly troublesome’ (Viasystems, p. 986). The case was effectively decided on that basis rather than on some principled opposition to the idea of dual vicarious liability. May LJ was therefore able to conclude that there was no authority binding on the court and thus preventing it from developing a principal of dual liability. That being the case, his Lordship framed the test as follows: ‘if, on the facts of a particular case, the core question is who was entitled, and in theory obliged, to control the employee’s relevant negligent act so as to prevent it, there will be some cases in which the sensible answer would be each of two ‘ employers’ (p. 990). Applying this test to the facts, May LJ held that dual liability was established in this instance. May LJ therefore viewed control as the crucial aspect, and where control is dual, so is liability. This is to be contrasted with the approach of Rix LJ who was not persuaded that control could be determinative. He said:

I am a little sceptical that the doctrine of dual vicarious liability is to be wholly equated with the question of control….Even in the establishment of a formal employer/employee relationship, the right of control has not retained the critical significance it once did….I would hazard, however, the view that what one is looking for is a situation where the employee in question, at any rate for relevant purposes, is so much a part of the work, business or organisation of both employers that it is just to make both employers answerable for his negligence (p. 993).

This conception looks at control together with the degree to which an employee has been integrated into the temporary employer’s business etc.

The existence of two approaches to dual vicarious liability raises the question whether only one of them is to be preferred. In that regard it might be argued that May LJ formulated the correct test. Control is of the essence. Rix LJ, who suggested that integration is also core, seems to be confusing the so-called ‘integration test’ which is used to determine whether an employment relationship exists at all. In a vicarious liability situation it is accepted that a person is employed, but the question is by whom. That aspect is decided by determining who can tell the employee what to do and how to do it. In other words, control is the key feature since an employee could be subject to the control of two employers but not be integrated into the temporary employer’s organisation.

One can use the facts of Hawley to illustrate this. The appeal in Hawley was decided after Viasystems and counsel were given permission to introduce arguments based on dual liability. However, these were rejected on the facts. Control was determinative. Control had so far been transferred to the temporary employer that the door stewards had been transferred and there was no question of dual liability. The only role ASE had reserved to itself was to pay the steward’s wages and provide replacements if Luminar objected to a particular individual. However, it is submitted that dual liability might have arisen had the stewards been given instructions by both their own employers and the management of the premises. Even though they wore uniforms supplied by the general employer, or even wore their own clothes, the question of dual control and therefore of dual liability is thus at issue. However, there is no question of integration here. Simply because a person employed by one party works on the premises of another does not mean they have been integrated, but they could be controlled by both in terms of the instructions received. The general employer might give instruction on the circumstances in which to remove a customer whereas the temporary employer might instruct on initial admissions. The general employer might instruct how and when to search whereas the temporary employer could instruct on what to do with drugs or weapons if found. The door steward as an example seems therefore to expose the weakness in Rix LJ’s approach. The suggestion is thus that May LJ’s test based on control alone is to be preferred. Indeed, control has been taken to be the central issue in Canada where the question of dual liability also recently arose (Blackwater v Plint (2005) 258 DLR (4^th) 275). Notwithstanding, one Canadian commentator has criticised the control approach. It has been suggested that if control is the test, ‘then parents should be vicariously liable for the torts of their children, and superior servants should be liable for the torts of their subordinates – positions which are clearly not now the law’ (Neyers, 2006, p. 198). This does, of course, ignore the fact that parents and superior employees do not employ their children or subordinates, a subtle but crucial distinction. The author therefore views ‘control’ with too great a degree of generality, when really it is specific to an employment relationship for these purposes.

Nonetheless, the Viasystems decision represents a dramatic addition to vicarious liabilities and employers will now have to factor this into any provision they have in place to limit or avoid such liabilities. It is still suggested however that the practical approaches mooted above are relevant. That is, in order to ensure that only the general employer will be liable, the temporary employer ought to filter instructions through the general employer itself or through a supervisor who acts as a conduit for instructions. Alternatively, the matter might be dealt with contractually.

This then leads to the question of contribution. If two employers are vicariously liable, are they financially responsible to the same extent? May LJ decided that:

[f]or dual vicarious liability, equal contribution may, depending on the facts be a logical necessity. This is because vicarious liability is a policy device of the law to redistribute the incidence of loss from a supposedly impecunious employee, who is personally at fault, to one or more supposedly solvent and insured employers who are not personally at fault. The court is not, therefore, concerned to look for personal responsibility in the employer based on what might otherwise have been direct responsibility (p. 990).

Thus, because vicarious liability is a ‘no fault’ or ‘strict’ liability, the question of apportioning blame between general and temporary employer is broadly irrelevant.

Although May LJ suggested that equal contribution is likely, it is by no means inevitable. In the Canadian case of Blackwater, for instance, the court said that it was ‘just and equitable’ to apportion liability on an unequal basis because, on the facts, one party had more control than the other. That is not to say that the same approach will develop in the UK, but it seems at least possible. Neyers is critical of this approach because of the degree of latitude inherent in a ‘just and equitable’ approach and it may be that the equal apportionment approach will ultimately be preferred (Neyers, 2006, p. 199). Additional questions also arise here. What is the situation where one employer is insolvent? Could a sympathetic court assess that employer’s contribution as nil? If this is possible, the supposed loss distribution basis for such an approach is eroded. The idea behind loss distribution is effectively to spread the cost of negligence claims as thinly as possible through insurance, the costs of which are then passed on to the consumer through higher prices. This argument also presupposes that any given market will withstand price hikes. Using the licensed trade as an example and the financial pressures it faces in light of the additional costs of compliance imposed by recent legislation, this argument may be rather flimsy. Relevant legislation includes the Licensing Act 2003 and the Licensing (Scotland) Act 2005 which comes into force in 2009. Currently, licensing in Scotland is regulated by the Licensing (Scotland) Act 1976. As regards security personnel, the Private Security Industry Act 2001 is also of relevance here. Further, in Scotland, a smoking ban was introduced on 26 March 2006 by the Smoking, Heath and Social Care (Scotland) Act 2005. A further ban will extend to England and Wales should the Health Bill currently before Parliament be passed. Cumulatively, this legislation arguably places financial pressure on the licensed trade, a factor apparently ignored by courts when extending vicarious liabilities.

One can also level the criticism that the apportionment of liability is predicated upon both the solvency and insurance of both employers. This approach is undermined as soon as it is found that one employer is either insolvent, or uninsured. The reasons for the absence of insurance may be two-fold. First, that in the absence of an express statutory requirement, employers may not see the need to insure. And second, even if insurance is in place, it may be inadequate to cover the loss claimed depending on the extent of the cover and the exclusions in the policy. It is worth considering the insurance aspect in isolation as it has been a feature in some of the door steward cases.

Insurance and Vicarious Liability

Insurance is fundamental to tort liability generally, not just to vicarious liability. Insurers pay out 94% of compensation and in nine out of ten cases the real defendants are insurance companies (Lewis, 2005, p. 86). That being so, one might argue that employers need have no concerns about potential liabilities since they will always be covered. However, as one might imagine, an insurer on the wrong end of a sizeable claim from another insurer may well look for a means to avoid payment. This is effectively what happened in Hawley. One aspect of that case decided by the Court of Appeal turned on the wording of the insurance policy held by the door steward agency. It stipulated that it would indemnify the agency against legal liability for damages and reasonable costs and expenses arising from accidental bodily injury to any person. The term ‘accidental’ was defined to mean ‘sudden, unforeseen, fortuitous and identifiable’. The question to be decided by the court was from whose perspective was the nature of the injury to be viewed? If it were that of the victim then the injuries would be accidental, but intentional if viewed from the perspective of the attacker. Ultimately, it was held that because the policy was once of public liability, injury was to be viewed from the perspective of the public, or injured party. Whilst this decision is consistent with prior authority (see for example Gray v Barr [1971] 2 QB 897) it tends to strain the meaning of what is ordinarily understood to be ‘ accidental’. An assault is not accidental simply because one did not wish it to happen or see it coming. Nonetheless, that is the decision, but one can see that in light of it insurers might amend their policies to stipulate that liability does not arise in respect of damage intentionally caused. Additionally, where as in Hawley, injuries are caused off the premises, geographical limitations might well become a feature of licensed trade polices generally. Thus, only injuries sustained within a closely defined definition of the ‘premises’ will be on-risk. As to whether this is the result remains to be seen but it may highlight the need for employers to check the fine print of policies in future.

This leads on to the concern that it is perhaps implicit in the reasoning of those courts that have expanded the scope of vicarious liability that liability insurance is behind every claim. Thus, it is no real detriment to the employer to pay for an employee’s wrongs as his insurer ultimately pays. And, whilst the cost of a claim under a policy may result in higher premiums, those additional costs can be spread widely and passed on to customers through higher prices. This is all very well, and may indeed be true in many cases. However, as a generalisation, it is somewhat dangerous. This is because at present there is no legislative requirement for businesses in the United Kingdom to carry public indemnity insurance. This is to be contrasted with the requirement to have employers’ liability cover which, at its most basic, is concerned with injuries to employees. It is fair to say however that many such insurance products are ‘ dual’ in the sense that they cover bother employee and public liabilities. This was certainly the case in relation to the policy in Hawley. It should be noted, though, that the defendants in that case were one of the country’s largest pub and club operators and it is no surprise that a large-scale sophisticated enterprise should have such ‘dual’ cover. However, there is nothing to say that such a responsible approach is reflected in the policies carried by smaller-scale enterprises. Nonetheless, that seems to be the tacit assumption of courts where vicarious liabilities have been extended, and especially where the policy consideration of loss distribution has been used as the justification.

Further, whilst a full consideration of the philosophical foundations of tort and the role of insurance is beyond the scope of this article, it is worth noting that the broadening scope of vicarious liability raises issues much wider than those mentioned expressly by any of the courts. Readers may well be familiar with the ongoing debate between tort law scholars about the true function of the law of tort and the role of insurance. Atiyah (1996) in particular has been one of the most prominent actors in this debate, proposing that the action of damages be abolished entirely with its replacement being left to the free market. (See also Atiyah, 1997 and Cane, 1996). In this conception the law of negligence would be virtually abolished as regards personal injuries. Claims would instead arise on the basis of first party insurance to cover medical expenses and lost income, claims against the extant social security system and a non-insurable tort exercisable against the wrongdoer, but limited in size financially. This is obviously a radical suggestion and is not without its critics. Keeler, for example, is critical of Atiyah’ s supposition that first party insurance can adequately supplement the social security system and thus provide adequate redress to claimants, although his paper is more specifically concerned with the role of tort in the difference between corrective and distributive justice and Atiyah’s comments in that regard (Keeler, 2001, p. 30; Burrows, 1998).

Stapleton (1995) on the other hand has warned of the dangers of confusing insurance with tort, to the extent that first party liability insurance can be seen as an alternative to an action in tort. This idea was developed by what she loosely termed the ‘Yale lawyers’ who suggested that there was historical evidence to support such a view. They further suggested that this was justified on a normative basis because the parties in a tort action could be linked by a bargaining relationship, meaning that tort liabilities could be insured against and the costs passed down the line (Stapleton included George Priest and Alan Shwartz of Yale and Epstein, Danzon and Huber within the term ‘Yale lawyers’). However, Stapleton convincingly demonstrated the flaws in this argument. First, she observed that the relational view of tort law parties is misconceived since it is at odds with the imposition of obligations on unrelated parties, the paradigm example occurring in Donoghue v Stevenson [1932] AC 562. And second, the tort-as-insurance argument ignores the underlying policy of the law of tort to deter wrongful conduct, which operates in addition to the policy of ensuring that an injured party is justly compensated.

Stapleton’s concern was to refute the argument that tort is unnecessary and can be replaced by insurance and to deny the suggestion that insurance is central to the question whether liabilities in tort should be imposed at all. In her conclusion she warned that ‘[c]ommentators and judges should think twice before making off-hand comments that insurance should be relevant to the scope of tort liability, that judges should take into account the ‘realities of insurance’ or that they should address the comparative insurability of parties. Such statements are dangerous.’ However, she did go on to say that ‘[e]ven if we decide that the full restorative measure of tort is not justified generally or for a specific misfortune, this does not mean that we are content to abandon the victim to the mere possibility of first party insurance…. For instance, we may want to provide support via a socialised risk system in the funding of which we may choose to incorporate risk-related contributions from injurers’ (Stapleton, 1995, pp. 833-837 and pp. 843-844). This last point is of particular interest. For present purposes, it is not suggested that vicarious liabilities be abolished in favour of a first party insurance system. Rather, my argument is that the realities of the caselaw discussed in this paper are that it creates a cumulative pressure on employers to insure against public liabilities in the absence of an express obligation to do so. Such insurance may be desirable, but it is not compulsory. Would it not be better that such an obligation be overtly recognised and placed on statutory footing? It cannot be suggested that claimants in vicarious liability cases do no deserve to raise claims or indeed that they should carry their own first party insurance. On the other hand, it should be recognised that where a business is uninsured or the relevant incident is ‘off risk’ then the potential effects on the business could be substantial if not devastating. Further, the continued reliance by recent courts on the loss distribution justification for imposing vicarious liability seems predicated on the view that losses can be distributed by insuring against them. The higher costs of insurance are thus spread widely through nominal increases in prices to be paid by consumers. But if, as Stapleton claims, insurance is irrelevant to the policy aim of deterrence, where is the incentive to insure, other than as a matter of self-preservation or social responsibility? In order to clarify matters, it may seem more realistic to require those businesses that have contact with the public to carry public liability insurance. For obvious reasons, this seems particularly relevant to the licensed trade, and could, if required, be included in any amendments to the current licensing legislation. In other words, compulsory public liability insurance could be introduced on an industry-specific basis.

Insurance and the Duty to Employ a Competent Contractor

Insurance was also a significant feature in Naylor (t/a Mainstreet) v Payling (2004) PIQR 36. In this case a door steward seriously injured a customer when ejecting him from licensed premises. On this occasion the steward was an uninsured independent contractor. Since vicarious liability only rarely attaches for the behaviour of independent contractors and there is little sense in suing them directly if they are uninsured and devoid of assets, as was the case here, the claimant had little option but to pursue the employer. The ground upon which the claim was raised was that the employer had breached the primary duty to employ a competent contractor by failing to ensure that the contractor was insured. That argument was rejected. Notably, it was also rejected in the recent unreported Scottish decision of the Outer House of the Court of Session in Honeybourne v Burgess [2005] CSOH 151 which also involved a violent door steward who was an independent contractor.

In Hawley the claimant attempted to rely upon the Court of Appeal decisions in Gwilliam v West Hertfordshire Hospitals NHS Trust [2003] PIQR 7 and Bottomley v Todmorden Cricket Club [2003] PIQR 18. In Gwillian it was held that there had been a duty to check the insurance position of an independent contractor where such a duty was effectively imposed by statute, in this case the Occupiers Liability Act 1957, s. 2. The hospital had checked the insurance position of the contractor but their cover lapsed a few days before the relevant injury occurred. It was held that the hospital had discharged its duty having asked and been advised that cover existed. In Bottomley a duty to confirm insurance cover in order to discharge the duty to employ a competent contractor arose because of the extra-hazardous nature of the activity. It emerged that no confirmation had been obtained after a member of the club was seriously burned following an explosion caused by an inept pyrotechnic display team.

As a starting point it was observed by Neuberger LJ in Naylor that ‘[a]n employer will not be liable for the negligence of his independent contractor, unless it can be shown that (a) he negligently selected an incompetent contractor, or was in some way responsible for the negligent way in which the independent contractor carried out the task, or (b) the task involved was unlawful, extra-hazardous, or carried out on the highway, or (c) the duty in respect of which the employer is alleged to be negligent was statutory in origin, or on some other basis non-delegable in nature…Save in the absence of special circumstances, in my view the law does not cast a free standing duty on an employer to satisfy himself that his independent contractor has insurance cover or would otherwise be good for a claim’ (para. 34).

The reasons for this conclusion are three-fold. The first relates to the types of case recognised by the law as giving rise to liability. These fall into two categories. Those cases where the employer will not be liable for the torts of contractors, provided that reasonable care is taken in selecting him for the relevant task; and those cases falling into the established categories of exception (para. 38). The second reason was that the law develops on an incremental basis in the field of negligence, rather than by seeking to apply or construct general restitutionary or compensatory principles. In other words, there is not enough evidence to support the creation of a new, discrete duty (para. 39). The third reason, and perhaps the most interesting in this context, was that, except where an employer is himself under a duty to have insurance cover, it would be unfair to create an obligation to satisfy himself that his independent contractor has insurance cover (para. 40).

On this final point it was observed that, if an employer is himself required to have insurance cover, for example by statute or by professional requirement, then the duty might arise. However, where there is no such requirement the fact that the employer does in fact have insurance does not mean that he then has to ensure that his contractors are adequately covered (para. 41). The crucial point in this case was that there is no general duty placed on an employer to have public liability insurance.

Unfortunately, there are a number of questions left unanswered. No guidance was given as to when the duty will be activated or what amounts to a ‘professional requirement’ to carry liability insurance. Does this relate only to the rules of a body to which an employer must subscribe in order to trade (as in the case of the legal profession) or does it also relate to the rules of a body joined voluntarily? Would a requirement to insure imposed as a condition attached to a liquor license be sufficient to activate the duty? We are not told. In addition, where the duty is activated, is the contractor required to have equal or equivalent cover to that of the employer? What happens if the exclusions or definitions in the contractor’s policy are more restrictive? If the type of conduct in question is excluded under the policy, is this tantamount to having no insurance at all and therefore a breach of duty? What happens if the upper value of a claim is lower under the contractor’s policy? Does this amount to a partial or total breach of duty? None of these questions is addressed.

Until decided otherwise, the consequences of this decision for any trade subject to direct or indirect insurance requirements may be significant. The concern would be that, in time, a number of cases could arise on the basis of the liability suggested by the Court of Appeal in Naylor. In other words, we may indeed see a body of law developing, sufficient to establish a further discrete head of liability in relation to independent contractors. Thus far, however, the evidence is against such a development, as confirmed recently by the Outer House of the Court of Session in the unreported case of Honeybourne v Burgess [2005] CSOH 151. Lady Smith firmly rejected the argument that the employer of a door steward in his capacity as an independent contractor has a duty to enquire as to his insurance position as a matter of primary duty. Her Ladyship was against the imposition of the duty on the basis of the practical problems canvassed above. The one point of difference was that in Honeybourne the argument was also based partly on occupiers’ liability. However, the Occupiers’ Liability (Scotland) Act 1960 does not contain a corresponding duty to that in the 1957 Act applicable in England and Wales to employ a competent contractor. That being the case, the Gwilliam arm of the argument was doomed to failure from the outset. As noted, the common law arm of the argument based on Bottomley also failed for the same reasons as given by the court in Naylor.

However, the real concern here is that until the new licensing legislation is bedded-in it will be unclear whether insurance could potentially be attached to a licence as a condition and thus whether the duty will activate on a more general level.

Vicarious Liability and Breach of Statutory Duty

It should be noted that ss. 1 to 7 of the Act do not apply in Scotland, and thus there is no corresponding offence of ‘harassment’ in Scots law. Rather, such conduct is dealt with under the common law as a breach of the peace. Sections 8 and 9 of the Act do however apply in Scotland, and create a right of civil action in relation to harassment together with an offence of breaching a civil non-harassment order. As a result, whilst Majrowski might support the general principle that an employer can be vicariously liable for the breach of a statutory duty placed only on an employee, in Scotland it would not be an authority as regards breach of such a duty under s. 1 of the Protection from Harassment Act 1997.

In a review of the early authorities dealing with this issue, Auld LJ in the Court of Appeal was able to discern three discrete strains of thought. The first accepted that liability could arise, the second rejected such a contention, and the third left it undecided as to whether liability might arise in these circumstances. However, it was observed that Atiyah (1967) had argued in favour of such a liability and had concluded from the authorities that the question was almost entirely open (pp. 280-284).

Placing the problem in its modern context, Auld LJ was persuaded that the extensions to vicarious liability post-Lister indicated that the liability was not restricted to breach of common law duty. Dubai Aluminium made it clear that liability extended also to breach of equitable duty. Further, Lister and the later cases established that ‘it is immaterial whether the conduct in respect of which a claimant seeks to hold an employer to account is a breach of a common law or statutory duty, and whether or not it is a criminal offence as well as a civil breach’ (para. 38). Provided there is a sufficiently close connection between the breach of statutory duty and employment, vicarious liability will arise. Once this general proposition was agreed upon, the remainder of the decision concerned its application to the specific provisions of the Protection from Harassment Act which need not concern us here. Notably, there was no agreement on that aspect of the case, with Scott Baker LJ in the minority arguing that the Act did not give rise to vicarious liability on the part of the employer.

On appeal to the House of Lords ([2006] 3 WLR 125), it was similarly held that there may be vicarious liability for breach of statutory duty. The speeches concentrate almost exclusively on the provisions of the 1997 Act and add little to the analysis of the principle itself as stated by the Court of Appeal. Only Lord Nicholls addressed the issue. In summary, his view was that, if the policy basis of the liability is loss distribution, and that policy permits an employer to be held liable for the common law and equitable wrongs of an employee, then the ‘ rationale also holds good for wrongs comprising a breach of statutory duty or prohibition which gives rise to civil liability, provided always the statute does not expressly or impliedly indicate otherwise’ (p. 128).

This decision is clearly correct as there can be no logical means of distinguishing between the legal categories of an employee’s wrongdoing. As with the other cases considered above, the decision represents a further extension to vicarious liability generally, and it is impossible to say that it will or will not affect the licensed trade more than any other trade or profession. Certainly, as regards door stewards there is always scope for harassment in relation to customers. However, it is clear that whether a statutory provision will give rise to vicarious liability depends upon the wording of the provision in question. In consequence, it is difficult to predict the extent to which vicarious liability has been extended without having a specific factual situation and a particular statute to consider.

Defences to Vicarious Liability?

Conclusion

Although this has been a largely industry-specific analysis, it has also shown that the developments in the law of vicarious liability give rise to more general concerns. The underlying policy basis for the liability has yet to be clearly identified, supposing that it is in fact identifiable. There are no clear limits to the application of the Lister test and what will amount to a sufficiently close connection between wrongdoing and employment. We have seen the creation of an entirely new species of liability with the introduction of ‘dual’ vicarious liability, the limits of which are yet to be tested. Added to this are the industry-specific examples such as Mattis and Naylor, the latter being a rare instance of a temporary deemed transfer of employment (in what are reasonably common circumstances within the industry). Taken together, and then added to the recent legislative pressures placed on the licensed trade through licensing, smoking and private security legislation, the licensed trade out of all others is shown to have suffered worst as a result of these recent developments. Further, as has already been noted, the cumulative effect of the recent caselaw may be to create an indirect pressure to obtain public liability insurance in the absence of an express legal obligation to do so. That said, it is no doubt a relief to the industry as a whole that the special form of liability which requires employers to ensure that their independent contractors are insured has not been extended to include the factual situation where licensees employ independent door steward contractors.

What these developments seem to require is an ultra-cautious approach to public liabilities. To that end licensed businesses may see it as a matter of necessity to ensure that extensive public liability insurance cover is in place - cover which extends to incidents occurring off the premises and possibly outside working hours. This in turn leads to an argument that such cover could be introduced as an industry-specific mandatory requirement, since licensed businesses have featured so prominently in the recent caselaw. In any event, the costs will be passed down to the consumer, along with the increased costs associated with the licensing, smoking and private security legislation noted above. What then is the ultimate result? Although it is a long way from the circumstances in Lister, in real terms, it seems that the effect of these judicial developments will be to put a few pence on the price of a pint.