Canterbury Law Review
There are many recent cases where victims of sexual abuse have brought civil actions for damages. These can raise various significant legal issues, and one of these, so far not much explored in New Zealand, is whether and the extent to which there can be vicarious liability for sexual abuse. Victims of sexual abuse frequently wish to sue the employer of the perpetrator rather than the perpetrator personally. A practical reason is the employer is more likely to be able to satisfy any award of damages. The focus of this paper is on the problems posed by claims of this kind in the context of New Zealand law.
Part I considers the nature of vicarious liability and why it is imposed. Particular attention will be given to the justifications for vicarious liability when applied in a sexual abuse context. Part II then examines the law as to when an employer may be held vicariously liable for such acts. Three recent leading overseas cases, Bazley v Curry, Jacobi v Griffiths, and Lister v Hesley Hall have considered this question. It is clear from these decisions that it is possible to impose vicarious liability for sexual abuse perpetrated by employees. What is required is a significant connection between the creation of a risk (the employment) and the wrong (the sexual abuse) that accrues therefrom. The implications of these developments in the light of the accident compensation scheme operating in New Zealand is then examined in Part III. This involves consideration of cover under the scheme and whether a sexual abuse victim can still maintain an action for exemplary damages. The discussion leads on to Part IV, which gives particular attention to the problem of vicarious liability for exemplary damages.
Vicarious liability is a common law doctrine that imposes liability upon one party fortorts committed by another party. In one of the leading treatise on vicarious liability, Atiyah has defined vicarious liability as a liability imposed by the law upon one person as a result of
a tortious act or omission by another,
some relationship between the actual tortfeasor and the defendant whom it is sought to make liable, and
some connection between the tortious act or omission and that relationship.
The most common form of liability is where an employer is held liable for the torts of an employee committed in the employee's course of employment. The employment relationship satisfies the second requirement, and the connection between the tort and the employment is expressed in the formula that the tort must have been committed in the course of employment. An employer's vicarious liability must be distinguished from any direct or primary liability the employer may also face. An employer is primarily liable for his or her own torts, including any failure to take all proper care in its employment procedures. Further, an employer who authorises an employee to commit a tort commits a tort in its own right and may be held independently liable. An employer who acquiesces intortious misconduct by an employee may also be independently liable in negligence. This is in contrast to imposing vicarious liability on the employer for the acts of his or her employees. Thus, it can be seen that vicarious liability is a doctrine of strict liability, where the employer is liable despite there being no finding of fault or negligence on the employer's part. The employer's vicarious liability is based on his or her servant's tort being imputed to him or her, whereas an employer's primary liability is linked to breach of his or her personal duty. The focus of this paper is on the vicarious liability of an employer, independently of any potential primary liability.
Vicarious liability is one of the most firmly established legal principles throughout the common law world. This may seem peculiar considering the doctrine is at odds with the fundamental principles of tort law that a person should only be liable for loss or damages caused by his or her own acts or omissions, and that a person should only be liable where he or she has been at fault. Although there is no one accepted view as to possible justifications, it is generally accepted the doctrine of vicarious liability continues to survive and be applied because it serves public policy. By ascertaining the policies at the heart of the doctrine, it should be possible to identify the circumstances where the doctrine will or should apply. If some common understanding can be reached as to the supposed justification of the doctrine, this may result in greater uniformity in its application. Numerous reasons have been advanced as to the rationale for vicarious liability. Given the general acceptance that the doctrine of vicarious liability has its basis in policy considerations, the question is what exactly those policy considerations are. Fleming has given an overview:
Most important... is the belief that a person who employs others to advance his own economic interest should in fairness be placed under a corresponding liability for losses incurred in the course of the enterprise; that the master is a more promising source of recompense than his servant who is apt to be a man of straw without insurance...; and that the rule promotes wide distribution of tort losses, the employer being a most suitable channel for passing them on through liability insurance and higher prices... The principle gains additional support for its admonitory value in accident prevention... [because] deterrent pressures are most effectively brought to bear on larger units like employers who are in a strategic position to reduce accidents by efficient organisation and supervision of their staff.
Fleming's words distil four key ideas to be found in the modern case law and academic writings — fairness, the deep pocket principle, distributive justice and deterrence. We will examine these ideas in turn.
Ideas of underlying rightness are often referred to in the cases. In Bazley v Curry, the Supreme Court of Canada identified a primary consideration behind vicarious liability as being the concern to provide a just and practical remedy to people who suffer as a consequence of wrongs perpetrated by an employee. This is based on Fleming's first justification. Employers employ servants to advance their economic interests and so it is thought that, in all fairness, they should bear the burden of providing a just and practical remedy for wrongs perpetrated by their employees. Put in another way, it is seen as right and just that the person who creates a risk should bearthe loss when the risk ripens into harm. In this sense, vicarious liability can be recognised as regulating the employer's risk-taking. It is a deliberate allocation of risk. This public policy consideration has been found in early judgments as well as those in the present day. In the old case of Duncan v Finlater, Lord Brougham made the well-known statement that "the reason that I am liable is this, that by employing him I set the whole thing in motion; and what he does, being done for my benefit and under my direction, I am responsible for the consequences of doing it". So, this fairness justification really derives from a concept of social responsibility. It is just that employers bear the costs that their enterprises generate, rather than innocent victims.
The "deepest pocket" principle justifies vicarious liability because the employer is considered to be more likely to be able to compensate the injured party than the employee who caused the damage. This is related to the concept of corrective justice, which emphasises the importance of providing adequate compensation for the victims oftortious misconduct. Corrective justice is concerned with righting wrongs. It seeks to reflect generally accepted principles of morality and personal responsibility. It has been suggested that one of the most important social goals served by vicarious liability is victim compensation. Preferably an innocent tort victim should be able to pursue damages from a solvent defendant, and as many employees have insufficient resources to defend or honour a tort judgment, vicarious liability improves the chances the victim can recover from the solvent employer. So, it is more likely that the wrong will in fact be put right.
Vicarious liability serves a loss-spreading function because employers are potentially better able to distribute tort losses to a wider community. First, loss distribution comes about because the employer can pass on the cost of his or hertortious vicarious liability, through the price of the goods or services, to the community which consumes the product or service. So this distribution is one of the expenses of doing business. Ultimately the expense will be borne by the consumer of the product, who should pay the costs which the hazards of the business have incurred. Secondly, an employer can insure against such tort losses and loss is then distributed to the community which pays premiums in the same category as the employer's insurer. Or, the cost of insuring against the employer's legal liabilities can also be included in the cost of the goods or services that are produced. In brief, if the costs of torts committed by employees in the course of employment are allocated to the employer, then the employer can take them into account rationally in pricing and output decisions. Further, this spreading of losses can take place over a period of time, rather than one individual having to pay out a large sum at a particular time.
Very arguably it is more socially expedient to spread losses from injuries to third parties among a large group of the community rather than among only a few. The employer is responsible for the losses not merely because he or she is better able to pay, but because he or she is more able to spread and distribute the loss. It has been said, then, that loss distribution is not a justification for imposing liability on the employer but is merely a description of where the loss ends up being borne. It has also been argued that because insurance follows liability, it cannot be used to justify or create liability.
So loss-spreading as a justification for vicarious liability is not an entirely convincing argument.
The other major policy lying at the heart of vicarious liability identified in Bazley was deterrence of future harm. The argument is that fixing the employer with responsibility for the employee's wrongful act, even where the employer is not negligent, may have a deterrent effect. Holding the employer vicariously liable for employees' torts will encourage the employer to put in place efficient administration, and supervision to reduce accidents and intentional wrongs. Accordingly, taking such steps will reduce the risk that such wrongs occur. The fear of vicarious liability gives employers great incentive to take every precaution to see that the enterprise is conducted safely. An employer who then implements a cost-effective deterrence strategy would potentially avoid facing a future large judgment in tort. Of course, evidence as to the extent to which the risk of liability in civil cases actually deters is not easy to acquire. Many people know little or nothing about the civil consequences of tortious conduct, which lessens the deterrence argument. Furthremore, in cases of negligence, it is difficult to see how a person can be deterred from making a mistake. Moreover, a defendant is less likely to be deterred when he or she is insured. Arguably, deterrence is most likely to operate in institutional settings where market considerations are likely to influence behaviour. The circumstances giving rise to vicarious liability, where an employer is in quite a good position to control or prevent tortious misconduct by an employee, fall into this category.
It can be argued that imposing liability on the employee alone would provide adequate deterrence. However, there are a number of reasons why this would not be so. First, if victims are unable to identify the particular employee who committed the wrong then employees are unlikely to be deterred. Vicarious liability, however, will achieve deterrence so long as employers are able to influence the class of employees. Secondly, employee liability may not provide adequate deterrence because employees are often not financially responsible, which means that suing the employee may not be worthwhile for the victim. The employee then has little incentive to avoid causing harm. Thirdly, it may be that individuals do not respond rationally to the prospect of liability, whereas employers do. By imposing vicarious liability, it might be anticipated that employers will induce their employees to behave reasonably. Fourthly, it has been suggested that 'under a regime of personal liability many employees would be able to force their employers to bear the cost of precautions in cases where employees can use employers' resources to take precautions and employers have difficulty observing such actions'.
Where the employer is sued the employee, of course, remains personally liable. Indeed, the employer may seek an indemnity or contribution from the employee, although this right is rarely, if ever, enforced. Seemingly, this weakens the impact of deterrence in relation to the employee. But the employee may be deterred by the possibility of dismissal from employment and/or a criminal charge.
In the main, the arguments justifying vicarious liability are quite convincing. Accepting them, they may have special resonance in sexual abuse cases. The problem of victims actually recovering compensation tends to be exacerbated in tortious sexual abuse cases. This exacerbation occurs because an employee who commits a sexual tort will, most probably, have been dismissed from work and may possibly be imprisoned for the criminal form of that conduct. The chances of recovering judgment from such an employee must be very low. This low likelihood of recovery makes the case for vicarious liability in a sexual abuse context that much stronger. If we accept that tort law is concerned to distribute losses, vicarious liability can assist in this function. Of course, it is true that if all persons carried their own accident insurance, then all losses from accidental injury would be spread throughout the community. However, most people do not possess accident insurance, and vulnerable victims (especially sexual abuse victims) are particularly unlikely to carry such insurance. So, again, vicarious liability is a particularly effective way of distributing tort losses in these circumstances. There may be a special need for deterrence in this type of case. Sexual abuse involves serious criminality and appropriate means of deterring that sort of behaviour should be utilised. It seems clearly desirable that employers should be encouraged to institute procedures to prevent offending where there is a risk of sexual abuse.
Therefore, there are reasonably convincing justifications forthe imposition of vicarious liability in sexual abuse cases. We will now look at the law as to when such vicarious liability will be imposed.
As we have seen, an employer is not vicariously responsible for all tortious acts of its employees. The employer is liable only for those acts which fall within the employee's course of employment. However, the point at which an employee steps outside his or her employment may be difficult to determine. In deciding whether an employee's tort has been committed in the course of his or her employment, usually the 'Salmond' test is applied. Salmond maintained that a wrongful act done by a servant is deemed to be in the course of employment if it is either (1) a wrongful act authorised by the master, or (2) a wrongful and unauthorised mode of doing some act authorised by the master. Furthermore, a master is liable even for acts which he or she has not authorised, provided they are so connected with acts which he or she has authorised, that they may rightly be regarded as modes - although improper modes - of doing them. So if the unauthorised and wrongful act of the servant is not so connected with the authorised act as to be a mode of doing it, but is an independent act, the master is not responsible: for in such a case the servant is not acting in the course of his or her employment, but has gone outside it.
It is relatively straightforward to apply the Salmond formulation to situations of negligent conduct by employees. However, a situation involving intentional torts by employees does not fit neatly into the framework of 'improper modes' of performing authorised tasks. This is because the employee may be seen as more intent on satisfying the employee's own interests than those of the employer. For example, a serious assault is antithetical to any concept of an act done in the course of employment. The question is whether or how Salmond's formulation can apply in the case of an employee's wilful wrongdoing, which is our present concern.
It is apparent from the case law that courts are much more reluctant to impose vicarious liability where an employee has committed a deliberate tort rather than where the employee has merely been negligent. But it is perfectly clear that employers can be held vicariously liable for intentional acts. For example, in some cases the use of force may naturally be expected of certain employees. Stewards at dance halls and persons authorised to repossess their master's property who assault third parties have been held to be acting within the course of their employment, regardless of the motives of the employee or whether or not the force was in excess of what was authorised. In this category, the employee's conduct can be seen as reasonably incidental to his or her allotted duties. When that is so, vicarious liability may be imposed.
In the leading decision in Lloyd v Grace, Smith & Co, the House of Lords established that an employer could be liable for a servant's fraud and other dishonest conduct even if the employee acted for his own benefit. In this case solicitors were liable for the dishonesty of their managing clerk, who had induced one of their clients to transfer property to him and then disposed of it for his own advantage. The principle was applied in Morris v CW Martin & Sons Ltd, where a client took her mink stole to the defendants for cleaning and the fur was entrusted to an employee who then stole it. It was held that the defendants were liable, as the employee converted the fur in the course of his employment. These and similar cases concerning dishonest employees have been explained on the basis that the employer owed a duty of care to the plaintiff and had chosen to perform it by selecting that particular employee to carry it out. As that duty cannot effectively be delegated, the employer cannot escape liability by reference to the theft or fraud of the employee. But it is not at all clear what is so special about these cases or when a 'non-delegable' duty will be imposed. Their underlying rationale is obscure.
The various decisions show the concept of vicarious liability for intentional wrongdoing is well established. Yet they have been explained in a range of ways and do not provide clear principles or guidelines for application in future cases. They do at least provide the background to the recent sexual abuse cases, to which we now turn. As will be seen, these cases may in fact throw some light on how the earlier cases can best be understood.
There are no New Zealand cases concerning vicarious liability for sexual abuse. The leading decisions in this area are now without doubt those of the Supreme Court of Canada in Bazley v Curry, Jacobi v Griffiths, and the House of Lords in Lister v Hesley Hall. However, before these decisions, a number of cases from England and Commonwealth jurisdictions imposed vicarious liability for sexual assaults and some denied it. We will look at what they say before examining the recent reviews of the whole question.
Prior to Bazley and Jacobi, the lower Canadian courts had examined a variety of claims where it had been sought to make employers liable for sexual assaults committed by employees. In the 'janitor' cases there was no vicarious liability on school boards for sexual assaults committed by janitors. The employment provided no more than the opportunity to commit the acts. Nor was the Catholic Church vicariously liable for sexual assaults committed by a priest against young parishioners. The fact the priest was placed in a position of trust and authority and was provided the opportunity to do wrong did not make the employer liable, especially in light of the fact that he acted criminally and against his vows. But in another decision the Catholic Church was vicariously liable for sexual assaults committed by a priest, as he was involved in the care or nurturing of children in the course of his employment. The Salmond 'unauthorised mode' test did not feature in the courts' determinations on liability. What did feature was the strength of the connection between the employment and the sexual assaults. In most of the cases, there was no sufficient connection to justify the imposition of liability. In particular, mere opportunity to commit the assaults was not enough.
The US cases initially demonstrated a general reluctance to hold employers vicariously liable for employees' sexual torts, but this may be changing. There are many cases on the issue and only a few will be mentioned. Claims against education authorities are common. In one, the employer was not liable where a young blind, deaf and mute student was sexually assaulted by the coordinator of her educational programme. The employee's actions were done solely for his independent, malicious, mischievous and selfish purposes. A similarly narrow view was taken in another case involving a young boy who was sexually assaulted by a school teacher during an off-campus extracurricular activity. The Supreme Court of California declined to impose vicarious liability, notwithstanding the policy concern that holding the school district liable might deter such behaviour, because this would be at the cost of significantly discouraging extracurricular activities and one-on-one exchanges between students and teachers. It considered also that there was not a sufficient degree of authority in school-related sexual assault cases to impose vicarious liability. However, the same court reached a different result where a police officer stopped the victim in her car, ordered her into his car, drove her to her home and sexually assaulted her. The court held that the city was vicariously liable for the police officer's rape. This decision to impose vicarious liability was a deliberate allocation of risk, based on the notion that it would be 'unjust for an enterprise to disclaim responsibility for injuries occurring in the course of its characteristic activities'.
Before the House of Lords' decision inLister, the English Court of Appeal considered the issue of vicarious liability for sexual assault in Trotman v North Yorkshire County Council. A young epileptic and mentally handicapped boy was allegedly abused by the deputy headmaster of the special school which he attended. This occurred during a school trip overseas, where, due to his condition, the boy was required to share a room with the headmaster. It was held (unanimously) that the employers were not vicariously liable for these actions. The court took a narrow interpretation of the Salmond test, with Butler-Sloss LJ saying:
[The headmaster's] position of caring for the plaintiff by sharing a bedroom with him gave him the opportunity to carry out the sexual assaults. But availing himself of that opportunity seems to me to be far removed from an unauthorised mode of carrying out a teacher's duties on behalf of his employer. Rather it is a negation of the duty of the council to look after children for whom it was responsible... But in the field of serious sexual misconduct, I find it difficult to visualise circumstances in which an act of the teacher can be an unauthorised mode of carrying out an authorised act, although I would not wish to close the door on the possibility.
This approach applies the Salmond test indifferently to torts involving negligence as well as to torts involving intention. Indeed, on a strict interpretation of the test, it is unlikely that a sexual assault can be said to be an unauthorised mode of performing an authorised task. But a different outcome might have resulted if the courts had considered the purpose of vicarious liability, and what the most socially beneficial outcome might be, and had taken into account developments in other jurisdictions. To these questions or matters we now turn.
In June 1999, the law surrounding vicarious liability for torts involving sexual misconduct was examined by the Supreme Court of Canada in the companion cases of Bazley v Curry and Jacobi v Griffiths. These decisions provide us with penetrating analyses both of underlying policy and of legal principle, helping point the way for the future.
The Children's Foundation, anon-profit organisation, provided residential care for emotionally disturbed children aged between 6 and 12 years. The aim of the Foundation was the alteration and replacement of deviant behavioural patterns with more socially acceptable behavioural patterns. Mr Curry was employed by the Foundation as a child-care counsellor to fulfil the role of parent to the children. Over a period of several months, Curry began a seduction of the plaintiff. He turned bathing and tucking-in routines into sexual explorations and sexual abuse. He was dismissed from employment and ultimately prosecuted and convicted for this (and other) criminal misconduct. The issue for the Supreme Court was whether the Foundation was vicariously liable for the tortious conduct of Curry. The court was unanimous in holding that it was indeed vicariously liable.
McLachlin J, delivering the judgment of the court, used the 'Salmond' test as a starting point: an employer is vicariously liable for (1) employee acts authorised by the employer; or (2) unauthorised acts so connected with authorised acts that they may be regarded as modes (albeit improper modes) of doing an authorised act. Her Honour said that the concern in cases involving vicarious liability for sexual assaults is obviously with the second limb of the test. The problem is how to distinguish between an unauthorised mode of performing an authorised act that attracts liability, and an entirely independent act that does not. She suggested the second branch of the Salmond test should be approached in two stages. First, prior precedents should be examined to decide if they determine on which side of the line the case falls. Secondly, if the earlier cases do not suggest an answer, the broader policy rationales behind vicarious liability should be reviewed to determine whether liability should be imposed. McLachlin J also recognised that the court had a duty to provide guidance for lower tribunals. Three general categories of case in which employers had been held vicariously liable for 'unauthorised torts' were examined. First, there were cases based on the rationale of 'furtherance of the employer's aims'. These cases were based on the agency rationale implicit in the Salmond test, whereby because the employee was acting in furtherance of the employer's aims, he or she is said to have 'ostensible' or 'implied' authority to do the unauthorised act. Secondly, there were cases based on the employer's creation of a situation of friction. If such a situation is created, then an employee's intentional misconduct is said to fall within the scope of employment and the employer is vicariously liable for the harm. This corresponds with the category of cases analysed above where vicarious liability has been imposed where there is authorised force in the course of employment. Thirdly, there were the dishonest employee cases, where courts have increasingly held employers vicariously liable. This was on the basis of fairness and policy. Although McLachlin J thought that these cases were not factually similar enough to the issue of vicarious liability for sexual abuse to be of much help as precedents, there was a common feature that could be discerned from them. They suggested that where the employee' s conduct was closely tied to a risk that the employer's enterprise had placed in the community, the employer may justly be held vicariously liable for the employee's wrong. The one case that was factually similar was Trotman. Her Honour criticised the decision in this case because it failed to successfully integrate the dishonest employee cases, and because the sexual act was not described in terms of the employee's duties of caring for and supervising vulnerable students, but in terms unrelated to those duties. The decision was unsatisfactory and did not resolve the issue before them.
The next step was then to consider the policy reasons for vicarious liability. McLachlin J acknowledged that any decision about whether to impose vicarious liability is one involving policy choices. As already noted, she identified the two main policy considerations behind vicarious liability as being a concern to provide a just and practical remedy for the harm and deterrence of future harm. McLachlin J thought these policy purposes are served only where the wrong is so connected with the employment that it can be said the employer introduced the risk of the wrong:
The question in each case is whether there is a connection or nexus between the employment enterprise and that wrong that justifies imposition of vicarious Accident Compensation Schemeliability on the employer for the wrong, in terms of fair allocation of the consequences of the risk and/or deterrence.
By itself the 'close connection' test does not in fact give any more guidance than the unauthorised mode test. But McLachlin J went on to provide guidelines for lower courts to follow when determining whether an employer is vicariously liable for an employee's unauthorised, intentional wrong. First, the courts should openly confront the question of whether liability should lie against the employer, rather than obscuring the decision beneath semantic discussions of 'scope of employment' and 'mode of conduct'. Secondly, the fundamental question is whether the wrongful act is sufficiently related to conduct authorised by the employer to justify the imposition of vicarious liability. It is not enough that the employment merely provided the opportunity for the tort. There must be a significant connection between the creation or enhancement of a risk and the wrong that accrues therefrom. In other words, the enterprise and the employment must materially enhance the risk of the tort occurring. Thirdly, in deciding whether this is so, anumber of subsidiary factors may be considered. When related to intentional torts, these include:
the opportunity that the enterprise afforded the employee to abuse his or her power;
the extent to which the wrongful act may have furthered the employer's aims (and hence be more likely to have been committed by the employee);
the extent to which the wrongful act was related to friction, confrontation or intimacy inherent in the employer's enterprise;
the extent of power conferred on the employee in relation to the victim;
the vulnerability of potential victims to wrongful exercise of the employee's power.
In summary, then, an employer is vicariously liable for an employee's sexual abuse of a client where the employer's enterprise and empowerment of the employee materially increased the risk of the sexual assault and hence the harm. Applying this test to the case at hand, McLachlin J found the Foundation vicariously liable for Curry's sexual misconduct. The opportunity for intimate private control and the quasi-parental relationship and power required by the terms of employment created the special environment that nurtured and brought to fruition Curry's sexual abuse. The employer's enterprise created and fostered the risk that led to the ultimate harm. The abuse was not a mere accident of time and place, but the product of the special relationship of intimacy and respect the employer fostered, as well as the special opportunities for exploitation of that relationship it furnished.
Although McLachlin J thought that the prior cases holding employers vicariously liable for the unauthorised acts of employees were of little help in resolving the issue in this case, they do accord with the 'close connection' test that was then propounded. That principle just needed to be drawn out. In doing this, Bazley acknowledges that it is the relationship between the employee and the victim as structured by the employer that allows the abuse to occur. It does seem more useful to examine the context of the relationship as opposed to the individual actions of the employee when deciding whether vicarious liability is appropriate. The focus should be on the employer, who structures the relationship between the employee and the victim and effectively creates the power dynamics between them. By entrusting the employee to be in close physical proximity to the victim and by fostering a power imbalance between the two (whereby the employee is empowered in respect of his or her duties towards the victim), the risk of sexual abuse is heightened. The test thus helps to explain when and why an employer might be held vicariously liable for conduct which is adverse to the goals of the employer. It also allows the question to be determined on a coherent basis of principle.
The Bazley test requires an open consideration of whether vicarious liability shouldlie, considering the broader policy rationales underlying and justifying the doctrine. A policy based analysis thus provides flexibility and avoids inconsistent outcomes which may result where vicarious liability is determined by a more mechanistic formula. The close connection test emerged in the light of the court's analysis of the policy rationales behind vicarious liability. It seeks to achieve a fair outcome, by balancing the interests of the innocent victim on the one hand and the 'innocent' employer on the other, while taking into account the wider social concerns of vicarious liability. One limit of the Bazley approach is that it is stated to apply only where the case law is unclear. But if this is a desirable way of looking at the question, then why limit it in this way? Arguably this approach should be applied in the interpretation of older cases too. Earlier explanations have not always been satisfactory and re-evaluation would seem appropriate. Indeed the Bazley analysis has value for the law of vicarious liability generally, not just the law in relation to sexual or other criminal wrongdoing.
As laudable as the Bazley test may be, it may not be entirely straightforward in its application. This is demonstrated by the different results reached in Bazley and Jacobi, and by the split of opinion in Jacobi. The reasoning in this second case adds little to that in Bazley, as both the majority and minority agreed the two-stage test/policy analysis propounded by McLachlin J in Bazley ought to apply. However, it does show the Bazley approach only sets out a series of questions, to which there may well be disagreement about the answers.
In Jacobi, the employer, the Vernon Boys and Girls Club (a non-profit organisation) operated a centre designed 'to provide behavioural guidance and to promote the health, social, educational, vocational and character development of boys and girls'. The Club provided a recreational, not a residential, facility for children. It did not purport to offer parenting of any description to its members. Staff were encouraged to develop a 'rapport' with members. One Griffiths was employed as Program Director to organise recreational activities and the occasional outing. Griffiths cultivated his two victims through the club by gaining their friendship. He eventually sexually assaulted both victims (who were brother and sister) on a number of occasions at his house, and once during a bus drive to a club-related sporting event. The Supreme Court held by a majority of 4 to 3 that the employer was not vicariously liable. The different outcomes reflect different views about how to apply the principles outlined in Bazley to the particular facts.
Binnie J, giving the majority decision, found the precedents useful in showing the lack of support for vicarious liability in this context. On the particular facts he considered there was not a sufficiently strong connection between the enterprise and the sexual assault to justify imposing vicarious liability. The club merely afforded Griffiths the opportunity to commit the sexual assaults. The enterprise did not confer job-created power on Griffiths and did not encourage intimacy. It offered recreational activities for children, whereby the employee was encouraged to develop a positive rapport with the children, rather than develop an intimate relationship. His Honour was especially concerned that the connection should be applied with 'appropriate firmness' where it was sought to make a non-profit organisation vicariously liable.
McLachlin J, in the minority, considered there was a material connection between the employment and the tort. She thought the club's function of providing guidance and moral direction to youths necessarily caused the club to authorise, and the staff to develop, a mentoring relationship of trust and intimacy between the staff and the children. Having assumed this special mentoring responsibility, the Club ought also to be viewed as having assumed responsibility for the heightened risks it introduced.
The Bazley approach requires the judge to decide in any given case whether the connection between the employer's enterprise and the employee's tort is close enough to justify imposing vicarious liability. Although this is a useful way of approaching the question of vicarious liability, it is still a line-drawing exercise and is sure to create problems in its application. Judges will differ on whether the employment led to a material enhancement of the risk of abuse, as evidenced by the contrasting decisions in Jacobi. But this ultimately is a factual question. There is nothing wrong with the test itself.
Looking at Jacobi itself, very arguably it is better to take the broader view of McLachlin J. The narrow approach taken by the majority fails to appreciate how sexual abuse operates from the victims' perspective. The employment did afford more than a mere coincidental opportunity to commit the tort. It placed Griffiths in a position of power over the children, and it is this sort of employment that materially increases the risk that sexual abuse will occur. Although the assaults took place outside work hours and off work premises, the position of power still increased the risk of the sexual abuse occurring. Thus, vicarious liability should have been imposed. This view has also found favour with many academics. It would seem from the majority decision the cases most likely to attract vicarious liability for sexual assault are those that involve a parental or similar relationship between the employee and the victim. Such cases will usually involve employers who provide the victim with some type of residential care. Although the majority stated that vicarious liability should not be limited to parental authority situations, it may be that outside this context it will be very difficult to establish the necessary strong connection between the employment and the tort.
The question of whether different considerations should apply when seeking to hold a non-profit organisation vicariously liable was discussed in both Bazley and Jacobi. Although the court made it clear in Bazley there was no exemption from the ordinary rules of vicarious liability for non-profit organisations, the majority in Jacobi thought the objectives of effective compensation and deterrence were not advanced to the same extent when applied to non-profit organisations. Hence there was a need for the strong connection test to be applied with 'appropriate firmness'. But it is not clear what this means and, if there is to be a different approach, wherein lies the difference. Tort liability does not normally depend upon some aspect of the identity of the defendant. Intractable problems would be likely if it did. It has yet to be seen whether the Bazley framework will be adopted in New Zealand and, even if it is, whether it will alter the established approach. First, there have been cases imposing vicarious liability for employees' criminal conduct in New Zealand, and these have been accommodated within the established framework. Perhaps, then, liability can be imposed using established principles and case law. Secondly, there is a question whether the Bazley approach has purely replaced one semantic analysis — the 'modes of conduct' test — with another — a test of whether there is a significant connection between the enhancement of a risk and the wrong committed. It has been suggested it may be better simply for the courts to use broader fact based assessments when approaching the 'modes of conduct' test. But very arguably the established framework is not adequate. It is not easy to discern clear principles from the previous cases that may help in future cases. The Bazley approach, on the other hand, looks at the scope of risk created by the employment. This not only focuses attention on the right question but also provides a clear set of guidelines to follow when determining whether vicarious liability should be imposed.
The House of Lords has also recently considered the question of an employer's vicarious liability for an employee's sexual assault. In Lister v Hesley Hall, the defendants ran a school for children with emotional and behavioural difficulties and a boarding annex to the school. The boarding house was intended to be a home for the boys and not an extension of the school environment. Grain was employed as the warden of the boarding house. He lived at the boarding house and was responsible, inter alia, for making sure the boys got up, went to school and to bed, and organised evening activities. Over a period of time, he systematically sexually abused the appellants in the boarding house and was convicted of multiple offences involving sexual abuse. The appellants brought an action against the employers, on the grounds they were vicariously liable for the torts of the warden. The trial judge and the Court of Appeal were bound by the Court of Appeal's decision in Trotman, holding that sexual abuse by a teacher could not be regarded as 'an unauthorised mode of carrying out an authorised act' within the Salmond test for vicarious liability. The House of Lords, however, overruled Trotman and unanimously imposed vicarious liability. Unfortunately, their Lordships did not speak as one but instead delivered four separate judgments. Each requires separate examination.
Lord Steyn, with whom Lord Hutton agreed, confirmed the authority of the Salmond test but thought the reason for imposing liability where the employee's act is an unauthorised mode of doing some authorised act (the second limb of the test) had been overlooked. Salmond had in fact explained that when imposing vicarious liability under this limb, it was necessary to focus on the connection between the employment tasks and the tortious act. Salmond said 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. This, Lord Steyn said, was a broad and practical test serving as a dividing line between cases where it is or is not just to impose vicarious liability. His Lordship thought Salmond's explanation was the germ of the close connection test put forward by the Canadian Supreme Court in Bazley and Jacobi. The correct approach, as enunciated by Salmond, was to concentrate on the relative closeness of the connection between the nature of the employment and the tort. This approach simply employed the traditional methodology of English law.
Applying it, his Lordship was satisfied that Trotman ought to be overruled, on the basis that the connection between the employment and the torts was very close and justified the imposition of vicarious liability. So also, as regards the case at hand, the warden's torts were closely connected with his employment and inextricably interwoven with the carrying out by the warden of his duties. It was thus fair and just to hold the employers vicariously liable.
Lord Clyde adopted a similar approach to Lord Steyn. He also considered the Salmond test to be the starting point and concluded that:
what has essentially to be considered is the connection, if any, between the act in question and the employment. If there is a connection, then the closeness of that connection has to be considered. 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.
He thought three matters were relevant when considering whether the connection was close enough for the tort to be included within the scope of employment. First, when considering the scope of the employment a broad approach should be adopted, so that not too much focus is placed on the particular act complained of. The purpose and the nature of the act, and the context and circumstances in which it occurred, have to be taken into account. Secondly, while consideration of the time and place at which the acts occurred will always be relevant, they may not be conclusive. Thirdly, the opportunity of being present at particular premises whereby the employee has been able to perform the act in question does not mean the act is necessarily within the scope of employment. Lord Clyde thought the essence of Bazley similarly lay in the recognition of a sufficient connection between the acts of the employee and the employment. This was consistent with the traditional English approach outlined above. In the instant case, Grain's employment as warden gave him opportunity and access to the premises and the close contact with the boys which that work involved created a sufficient connection between the acts of abuse and the employment.
Lord Millett also considered the Salmond test to hold the answer to the question of when vicarious liability is imposed for intentional wrongdoing. He thought that vicarious liability should be imposed where the unauthorised acts of the employee are so connected with acts which the employer has authorised that they may properly be regarded as being within the scope of employment. This formulation had the advantage of dispensing with any reference to 'improper modes' of carrying out the employee's duties, and also directed attention to the connection between the employee's duties and his wrongdoing. He recognised that what is critical is the closeness of the connection between the employee's duties and his wrongdoing, not precise terminology. He acknowledged that this was the principle derived from Bazley and Jacobi. He was also satisfied that a sufficient connection was shown on the facts to justify liability.
Lord Hobhouse did not think that it was appropriate to follow the lead given in Bazley, as this dwelt more on policy than on defining the criteria for a legal principle to apply. He imposed vicarious liability on the basis that the employers voluntarily assumed a relationship towards the plaintiff that carried corresponding duties, the performance of which were entrusted to the employee. The classes of persons or institutions that are in this type of special relationship to another human being include schools, prisons, hospitals and even, in relation to their visitors, occupiers of land. Where this is so, the motive of the employee and the fact that he is doing something expressly forbidden and is serving only his own ends does not negate the employer's vicarious liability. The duties which the employer owed to the plaintiffs were more extensive than those owed by the public at large and included a duty to take all reasonable steps to safeguard the plaintiffs in their physical, moral and educational development whilst at the school. The employers entrusted the performance of that duty to their employee and that employee failed to perform that duty, rendering the employers vicariously liable for that breach of duty.
It is clear from Lister that it is now possible to impose vicarious liability for sexual abuse perpetrated by employees in institutional settings. The House of Lords has effectively revised the test of vicarious liability and endorsed a 'close connection' test. Their Lordships thought that they were clarifying the second limb of the Salmond test but in fact they seemed to refine and reinterpret it. Indeed, by focusing on the subsequent sentence made by Salmond (referring to the connection between the tortious act and the employment tasks), they may have extended it beyond what Salmond originally intended. They did not want the Salmond test to be read as narrowly as it was in Trotman, which focused too much on whether the wrongful acts could be described as unauthorised modes of performing the employee's duty. The correct approach was to focus attention on the closeness of the connection between the employee's act and the duties of his employment. Their Lordships all noted that the employment must do more than merely provide the opportunity to carry out the wrongful act. The question which will now tax the courts is where will the line be drawn. As seen from Jacobi, the answer may not always be clear.
Four of the five Lords in Lister approved of Bazley, although they did not themselves set out and (allegedly) did not take into account the social and policy considerations underlying the doctrine of vicarious liability. It has been questioned whether a court can really interpret and apply a pure common law doctrine without referring to the principles and policy behind the doctrine. Perhaps questions of policy did in fact creep into the issue of whether there was a close connection. Lord Steyn noted the two conflicting policies recognised by Fleming that were at stake: the social interest in furnishing an innocent tort victim with recourse against a financially responsible defendant, and a hesitation to foist any undue burden on business enterprise. Lord Clyde spoke of whether the employers should be vicariously liable, which, it has been suggested, may imply some social responsibility of the employer. Lord Millett recognised vicarious liability to be a loss-distribution device and considered the close connection test to be consistent with the policy behind vicarious liability. Lord Hobhouse, however, thought that an exposition of the policy reasons for a rule is not the same as defining the criteria for its application. But there is little doubt that policy did in fact play a part in their decisions. In practical effect, the boundaries of vicarious liability were redrawn in Lister. However, the House of Lords denied any redrawing, and said they were simply indicating where the boundaries had always been.
Despite arguably following different paths, one founded on policy and one on legal principle, the Canadian and the English courts have come to the same conclusion. There must be a significant connection between the creation of a risk (the employment) and the wrong (the sexual abuse) that accrues therefrom. Even though the House of Lords and the Supreme Court of Canada based their conclusions upon the second branch of the Salmond test, in fact a new test has emerged in both countries to govern vicarious liability for intentional wrongdoing. As we have seen, it has been called the 'close connection' test.
This might be better expressed as the 'scope of the risk' test. The idea of risk provides a way of looking at the cases which is both principled and helpful in practice. There may be problems in application, putting cases on one or other side of the line, but these are unavoidable. It will never be possible to achieve absolute certainty. Even so, when applying this test, it may be helpful to take into account the policy factors discussed in Bazley. These help to indicate whether there is a significant connection and inform the underlying idea.
The risk approach can also be seen as consistent with recent developments in the field of personal liability for harm done to others or by outside forces. The problem in that area is how to distinguish between causing a loss and providing the opportunity for its occurrence. For liability to be imposed, more is required than the mere fact that the defendant had contributed to the coincidence of time and space in which damage could happen. The concept of risk provided an answer to this problem. When analysing whether a defendant has caused a loss, it is helpful to ask what was the scope of the risk created by the defendant's conduct. This notion has received recent support in the New Zealand Court of Appeal and does seem to represent the current trend of thinking.
Where exactly do these developments leave us in New Zealand? The accident compensation scheme covers most claims for personal injuries. But in what circumstances can one still sue for damages at common law? A common law action can still be maintained where the claim is not barred by the accident compensation scheme. This bar is found in s 317(1) Injury Prevention, Rehabilitation, and Compensation Act 2001, which provides that:
No person may bring proceedings independently of this Act, whether under any rule of law or any enactment, in any court in NZ, for damages arising directly or indirectly out of -
(a) personal injury covered by this Act; or
(b) personal injury covered by the former Acts.
In spelling out the effect of this bar and the limits to the scheme, it is necessary to distinguish between claims for compensatory and claims for exemplary damages.
Cover for personal injury is determined by s 20 of the 2001 Act. We will deal first with physical injury. Section 20(1) gives the basic requirement:
A person has cover for a personal injury if-
(a) he or she suffers the personal injury in New Zealand on or after 1 April 2002; and
(b) the personal injury is any of the kinds or injuries described in s 26(1 )(a) or (b) or (c) or (e); and
(c) the personal injury is described in any of the paragraphs in subsection (2).
There are ten categories of personal injury which are covered in s 20(2). The only one which is of immediate relevance is para(a). This is
(a) personal injury caused by an accident to the person.
Personal injury is broadly defined in s 26(1) as meaning, inter alia, (a) the death of a person, or (b) physical injuries suffered by a person, including, for example, a strain or a sprain, or (c) mental injury suffered by a person because of physical injuries suffered by the person. However, by s 26(2), it does not include personal injury caused wholly or substantially by a gradual process, disease, or infection unless, broadly, it is work related or aconsequence of medical misadventure. As forthe meaning of'accident', this is defined at considerable length in s 25. The core requirement in paragraph (a) is that there be an application of a force or resistance external to the human body.
Claims for sexual abuse causing actual physical harm will be covered for accident compensation under s 20(2)(a), as amounting to personal injury caused by an accident. As is apparent from the definition of personal injury, this includes consequential mental injury. But where the victim contracts a disease caused by criminal sexual conduct it appears, somewhat bizarrely, that this is not covered by the Act, as the disease is not 'personal injury' falling within the categories specified.
A further category of cover is in s 21. Section 21 provides cover for mental injury caused by an act performed on, with or in relation to the person which is of a kind described in the listed criminal sexual offences in Schedule 3 to the Act. Section 26(1)(d) further provides that personal injury means mental injury suffered by a person in the circumstances described in s 21. Sexual abuse will generally involve one of these criminal offences. Accordingly, it is highly likely victims of sexual abuse will indeed have cover under the Act. There is no cover where mental injury is suffered as an outcome of non-criminal sexual misconduct. There is little scope for non-criminal conduct in the present context. Therefore, a victim of a criminal sexual assault who suffers physical or mental injury is able to claim compensation under the statutory scheme and cannot bring an action for compensatory damages. The victim can neither claim against the perpetrator him or herself, nor against the employer of the perpetrator on the basis of vicarious liability. This thus leaves little scope for such actions as were brought in Bazley, Jacobi and Lister, which were actions for damages for personal injury. However, where the accident compensation scheme does not apply, the victim of a sexual assault can sue for compensatory damages. Damage by sexually inflicted disease seems to be the only possible type of claim. The question of vicarious liability can still arise in this particular context.
Claims for exemplary damages are different. But determining whether an award of exemplary damages can be made is now a highly complicated business. First it is necessary to consider when there may be personal liability for exemplary damages. This will be done by a series of numbered points, which may assist in clarifying the position. Then we can consider the question of vicarious liability.
1. The question arises as to whether the statutory bar on bringing an action for damages where there is cover under the Act applies equally to claims for exemplary as well as compensatory damages. In Donselaar v Donselaar the Court of Appeal concluded actions for exemplary damages could still be maintained. The Court took a narrow interpretation of the barring provision and explained that proceedings for exemplary damages were not proceedings for damages arising directly or indirectly out of a plaintiff’s injury. They were awarded because of the outrageous character of the defendant's conduct. In a strict sense such damages did not arise at all, and in a looser sense they could be said to arise out of the acts of the defendant. This finding would not undermine the policy of the legislation, which was aimed at compensation as opposed to punishment and deterrence.
2. In Taylor v Beere the New Zealand Court of Appeal examined the basis for making an award of exemplary damages. It concluded that exemplary damages would only be awarded in an exceptional case to register condemnation of 'outrageous' conduct by the defendant, and to mark the 'contumelious disregard by the defendant of the plaintiff’s rights'.
The availability of exemplary damages was more recently considered by the Court of Appeal in Bottrill v A. The case concerned exemplary damages for medical negligence. Richardson P, delivering the judgment, thought the remedy of exemplary damages should be directed at advertent wrongdoing. He said the test for whether the defendant's conduct merited an award of exemplary damages is whether there was conscious risk-taking — where the defendant went ahead knowing of the risk or being indifferent to the consequences. Negligence alone, even gross negligence, is not enough.
3. So, in principle, claims for exemplary damages for sexual abuse can still be maintained. Although the threshold standard of conduct required to attract an exemplary award is high, a case of intentional sexual misconduct would easily qualify for such an award. It is deliberate wrongdoing and satisfies the Bottrill test.
4. However, there is an added problem. A case of this kind involves serious criminal offending, and this raises the question as to how that should affect the availability of exemplary damages. In Daniels v Thompson, J v Bell, W v W, H v P the Court of Appeal had to consider the effect of a prior criminal conviction on a subsequent claim for exemplary damages. Henry J, delivering the judgment of the majority, took the view the sole purpose of exemplary damages was to punish and deter the tortfeasor. If there has been a prior conviction and a penalty imposed for the same acts then punishment has already been exacted. Any award of exemplary damages would thus amount to double punishment, as the civil court would be imposing a second penalty for the same conduct. His Honour also held that where there has been a prior acquittal, it would offend the principle of finality and amount to an abuse of process to allow an action for exemplary damages to proceed. Finally, where a prosecution had not been concluded, he thought a civil proceeding should be stayed pending the outcome of the prosecution. If no prosecution had been commenced, then civil proceedings should only be permitted when it was clear that a prosecution was unlikely. Thomas J dissented from the decision. He considered a prior criminal conviction and penalty should not be an absolute bar to a claim for exemplary damages, although they should be taken into account by the civil court when determining whether an award is appropriate. His Honour thought the imposition of an absolute bar would impinge upon the basic right of all individuals in a free and democratic society to have access to the courts and would have a disproportionate impact on women, who are statistically much more likely to be victims of sexual abuse than men. His Honour also considered there to be essential differences between criminal and civil proceedings. They possess different objectives, address different wrongs, focus on different conduct, lead to different consequences, and the participants have a different status and different roles. Again, Thomas J thought that where the defendant had been acquitted of criminal charges or had not been prosecuted at all, there should be no bar on a claim for exemplary damages. However, where appropriate, the proceedings could be struck out or stayed as an abuse of process.
W v W and J v Bell were taken on further appeal to the Privy Council, which dismissed the appeals on the basis the question was one of public policy for the national court to decide. They declined to substitute their own views (if different) for those of the New Zealand court. The decision in Daniels was criticised on the basis it was unjust to victims of criminal misconduct, particularly victims of sexual abuse. But there are strong arguments in favour of the majority decision, at least where the defendant has been convicted and punished in a criminal proceeding. Even if on Thomas J's view there was no absolute rule barring a civil action but rather a discretion to award exemplary damages in appropriate cases, the discretion would only be exercised in exceptional cases. The plaintiff, then, having gone to the trouble and expense of bringing an action, is highly unlikely to succeed. But where the defendant has been acquitted on a prior criminal charge there is no question of double punishment. Any later civil litigation involves different rules of evidence and there is a different standard of proof. It would not seem to be an abuse of process to allow the plaintiff to prove his or her case on the basis of these different rules. Furthermore, the finding that a claim for exemplary damages should be struck out or stayed as an abuse of process, if a criminal prosecution is likely to commence, can be supported on the ground that the public interest in punishing serious criminal acts and protecting other potential victims from repeat offending must take priority over an individual's private interest in making a financial recovery.
5. There has since been legislative change in this area. Section 319 of the 2001 Act (formerly s 396 of the 1998 Act) provides that plaintiffs can still bring exemplary damages claims where they have suffered personal injury covered by the Act and whether or not the defendant has been charged, not charged, convicted, acquitted or discharged of an offence involving the conduct concerned in the claim for exemplary damages. Further, when determining whether to award exemplary damages and the amount of them, the court may have regard to any other penalty imposed for an offence involving the same conduct.
The objection to this is encapsulated in the majority decision in Daniels. By providing that the criminal punishment should be taken into account in deciding whether to make an award of exemplary damages and setting its amount, the civil court is allowed to second guess the criminal court's sentence. Furthermore, no guidance is given as to how the civil court is to go about making that assessment. Courts may be reluctant to 'top up' any criminal penalty, thus leaving the victim (who had been encouraged to bring an action by the removal of the Daniels bar) with nothing after the long litigation process.
6. Whatever the objections, in sexual abuse cases the victim generally can sue the wrongdoer for exemplary damages, as the victim is usually covered by the accident compensation scheme. The fact the wrongdoer has faced or will face criminal charges involving that conduct does not prevent an action. This seems something of a paradox. Normally, cover under the scheme means the victim cannot sue.
7. But the Daniels bar still applies in cases not involving personal injury claims covered by accident compensation. These cases include property damage, wrongful contacts which do not cause personal injury (such as spitting) or infecting another with a disease. A plaintiff cannot sue for exemplary damages involving conduct about which criminal charges have been or might be brought if he or she is not covered by the Act and so s 319 does not apply.
This suggests that s 319 does not represent a considered and well thought out conclusion of policy, because if Parliament decided, after a serious analysis of the issues, that the view taken in Daniels ought to be rejected, one might have expected that it would have been rejected across the board.
All of the above relate to personal liability for exemplary damages. Of course personal liability is a precondition to any possible vicarious liability. But there is a major issue as to whether vicarious liability for exemplary damages should exist at all. We turn now to consider this thorny question.
In McLaren Transport Ltd v Somerville it was simply assumed that an employer's vicarious liability can include liability for exemplary damages. The logic of this result was not questioned. The few cases in which the issue has actually been scrutinised, however, do support this view. In Carrington v Attorney-General, Henry J held that the Crown was vicariously liable for punitive (exemplary) damages awarded as a result of a police officer committing a wrongful arrest and detention. This was on the basis that what the police officer did was done in the course of employment. To hold that the employer was liable only for some of the damages which flowed from the relevant acts, namely compensatory damages, was to ignore the quality of what the police officer had done in the course of his duty. Although the awarding of exemplary damages might be anomalous, Henry J thought the instant case was within the anomaly:
what the superior is responsible for is the mode of the exercise of the power he has conferred on the wrongdoer. If that mode be a contumelious.. mode of exercise I can see no reason in principle why damages should not follow from that mode of exercise of power.
The issue was also considered in Monroe v Attorney-General, where exemplary damages were awarded against the Crown in respect of assaults by unidentified police officers. Prichard J concluded that there were sound reasons why the doctrine of vicarious liability should extend to liability for exemplary damages. First, the admonitory value of allowing exemplary damages to be awarded against employers would encourage employers to exercise closer control and discipline over their servants. Secondly, where the plaintiff had no way of proving the identity of the person who was the subject of his complaint (although there was no doubt that the person was acting in the course of his employment by a readily identified employer), then there was a danger that employer and employee would close their ranks, resulting in the conduct going unpunished, and the court being unable to express its disapproval of the outrageous conduct. Both arguments may be flawed. It has been maintained that in the case of a blameless employer, imposing vicarious liability is unlikely to have any significant deterrent effect, as it is unlikely to induce any significant modification of business practice. Again, an employer who expressly or impliedly approves a wrongful act can be held to have ratified or adopted that act and be personally liable for it on this basis. It has been suggested that failure by an employer to take reasonable steps to identify and discipline the errant employee should amount to ratification. It is unnecessary to attempt to hold them vicariously liable. If the approved act was sufficiently heinous to attract an exemplary award, then the employer can properly be held liable for exemplary damages.
The above decisions are from the (then) Supreme Court and the High Court, thus leaving the way open for the Court of Appeal to take a different line.
In America, there is a split between jurisdictions taking a liberal view and those taking a restrictive view about the extent of an employer's liability. It seems that the majority of courts adhere to the view that an employer's vicarious liability extends to exemplary as well as compensatory damages. This is based on the assumption that such liability will encourage employers to exercise greater care in the selection and supervision of employees and will deter misconduct. However, a minority of courts, following a leading federal case, adopt a narrow rule of liability, imposing vicarious liability for exemplary damages only when a superior officer is shown to have ordered, participated in, or ratified the misconduct. The rationale of courts following this rule is that it is unjust to punish an innocent employer who had neither authorised nor ratified their employees' conduct. This is known as the 'complicity' rule and it eventually became embodied in the Restatement (Second) of Torts. The Restatement states that exemplary damages can be awarded against a principal (employer) if, but only if, (1) the principal authorised the manner and doing of the act; (2) the agent was unfit and the principal was reckless in employing or retaining him; (3) the agent was employed in a managerial capacity and was acting in the scope of employment; or (4) the principal ratified or approved the act. However, it has been pointed out that the first, second, and fourth situations are actually instances of the employer's personal responsibility for the choice of employee or in authorising the relevant conduct. This is compared to the employer's true vicarious liability in the third situation. This is also problematic, as there is no definition of a 'managerial' agent, nor any justification for treating such an agent differently from others.
One English Court of Appeal decision has proceeded along the same line as McLaren Transport and assumed an employer could be held vicariously liable for an award of exemplary damages. There was no substantive discussion of the issue until Lord Scott approached it in the 2001 case of Kuddus v Chief Constable of Leicestershire. This was an action against the chief officer of police on the ground of his vicarious liability for a constable's misfeasance in public office. Although the primary issue was whether exemplary damages were recoverable for the tort of misfeasance in public office, a related question was whether exemplary damages can be recovered in a claim where the defendant's alleged liability is simply vicarious. The point was not argued and four of the Lords reserved their opinion, but Lord Scott considered the issue. He said that the objection to exemplary damages awards in vicarious liability cases is fundamental. The defendant should not be liable to pay exemplary damages unless he has committed punishable behaviour. There is no room for an award of exemplary damages against an individual whose alleged liability is vicarious only, and who has not done anything that constitutes punishable behaviour. Although Lord Scott's view is clearly obiter, it is a well thought-out consideration of the issue and accords with a natural sense of fairness.
The cases thus far, with the exception of Kuddus, provide little helpful guidance in determining whether exemplary damages can be awarded on the basis of vicarious liability only. Thus, an examination of the policy behind exemplary damages and whether that is compatible with vicarious liability for exemplary damages needs to be undertaken. The generally accepted purposes for imposing exemplary damages are punishment and deterrence. In New Zealand, at least, these dual rationales have been adopted. But it may be that there are other functions which are also served by exemplary damages awards. We need to examine the various rationales to help determine whether they can also justify exemplary damages being imposed on a vicarious basis.
A defendant who has unjustifiably or inexcusably committed a wrongful act is deserving of punishment. Exemplary damages are intended to punish egregious misconduct. The punishment extends beyond that which compensatory damages provide.
The essential argument against using punishment as a basis for an employer's exemplary damages liability is the employer itself has done no wrong and thus does not deserve to be punished. It is quite simply unfair to punish an employer who is innocent of any wrongdoing and leave the guilty to go unpunished. As we have seen, Lord Scott thought this was the fundamental objection to vicarious exemplary damages awards. The purpose of exemplary damages is to punish the wrongdoer, and this purpose would be ignored if that punishment was transferred to another. The point cannot be evaded by pointing to the fact that employers may well be corporations or public bodies:
Although vicarious defendants are often artificial persons such as municipalities or business corporations, that serves only to disguise the fact that punishment ultimately falls upon natural persons who have not committed wrongful or harmful acts. In the case of municipalities, the loss falls upon taxpayers, employees, and consumers of the goods and services the city dispenses. In the case of business corporations, punishment falls upon stockholders, creditors, employees, suppliers, sellers of complementary products, consumers of goods and services, and taxpayers.
It is obvious that none of these persons has committed a wrongful act and so does not deserve punishment. This is especially important where sexual abuse is perpetrated in a care facility, as the employer sought to be made vicariously liable is likely to be a local or national body. Indirectly, then, it is the taxpayers who will be punished for the tortfeasor's conduct if exemplary damages are awarded. Yet these are the people who are most removed from the misconduct.
In response, it has been argued that this 'unfairness' objection is not unique to vicarious liability for exemplary damages. A similar objection can be made against vicarious liability to compensatory damages, as an employer can be held liable even in circumstances where he or she has not been at fault. But this loses sight of the fact it is unfair to punish an innocent employer by way of exemplary damages, as opposed to merely holding them liable for compensatory damages, which is fair (see the above justifications), as such damages do not include a punitive element. When determining the size of an exemplary damages award, the wealth of the defendant is a major factor that is considered. This is to ensure that the damages do actually punish the defendant in a significant way. How, then, are the damages to be calculated in the case of a vicariously liable defendant? How exactly is a large corporation or a local authority to be punished? And seemingly the fact that an employer has substantial assets could be extremely prejudicial to them when defending an exemplary damages claim. Yet they are innocent of any wrongdoing.
Punishing the defendant for his or her misconduct, via exemplary damages, should deter similar misconduct in the future. Imposing exemplary damages to achieve deterrence, however, is only justified when compensatory damages alone produce less than optimal deterrence. Exemplary damages not only deter the wrongdoer from egregious misconduct in the future but also deter others from following the defendant's example. The major justification offered for vicarious exemplary damage liability is its deterrent effect. The argument is that such liability will encourage the employer to exercise greater care in his or her selection and monitoring of employees, thereby reducing the incidence of harmful acts. By imposing liability on the employer, this is thought to encourage efficient employment arrangements in which the employer creates incentives for employees to avoid breaches of others' rights. So, the aim of holding the employer vicariously liable for exemplary damages is to force the employer either to exercise greater care in the selection of employees or to supervise them more closely and carefully. The English Law Commission is a proponent of this view. Any concerns about the unfairness of not punishing the employees can be mitigated by allowing the employer a right of recovery against the employees.
This argument is unconvincing. As regards selection, the employer can be primarily liable in negligence if he or she failed to take reasonable care in hiring. So, protection from negligent hiring cannot be used as a justification for awarding exemplary damages on a vicarious basis. It is also questionable whether effective means even exist for testing prospective employees.
Further, vicarious liability for exemplary damages will often fail to bring about any closer supervision by the employer. It has been observed that:
The ability to better control the actions of the employees through greater supervision is often illusory...Employees may perform their duties where direct supervision is impossible. Further, increased supervision may well be ineffective to prevent the occurrence of certain torts for which punitive damages may be assessed.
Similarly, it has been suggested that employees who engage in conduct that gives rise to exemplary damages awards are likely to go well beyond the threshold standard for the type of conduct needed to justify such damages when they intentionally behave in an undesirable way. When this is so, the conduct of the employees is unlikely to be altered by any preventive measures put in place by the employer. In addition, an employer's excessive monitoring of employees may be both inefficient and unacceptably intrusive of privacy rights. Moreover, efforts aimed at supervision may not be cost-effective enough to motivate employers to undertake them.
If exemplary damages are to have a real deterrent effect, they must be imposed on the employee. Given the range of conduct that can give rise to an award, an employer cannot tell in advance what he or she must to do avoid punishment or how severe punishment might be. If it is nearly impossible to predict whether an exemplary damages award will eventuate and the size of an award, then its deterrence value decreases. An employer is hard pushed to prevent the occurrence of acts that it cannot anticipate. It is the employees who know their own potential for committing harmful acts and who control their own actions. Any threat of exemplary damages should thus be directed at them if they are to have an incentive to act responsibly.
Exemplary damages may also serve other functions. In Daniels v Thompson Thomas J recognised other broad objectives of tort law as including condemnation, education, the avoidance of abuses of power, the symbolic expression of society's disapproval, vindication and appeasement of the victim with any consequential therapeutic benefits for him or her arising out of the civil trial. These functions, he thought, were exceedingly well served by exemplary damages. Indeed, most of these purposes do support the imposition of exemplary damages personally. Exemplary damages condemn the wrongdoer for his or her outrageous behaviour and, by punishing him or her, symbolise that such behaviour is not tolerated by society. They may also educate the tortfeasor and others that wrongdoing does not pay. But do these justifications also apply to vicarious exemplary damage awards? Arguably not. It is hard to see that there is any symbolic value in punishing someone other than the wrongdoer for heinous behaviour. Similarly, can it be said that awarding exemplary damages against an employer actually educates the wrongdoer or other potential wrongdoers not to repeat the behaviour? Probably not, as they are unlikely to be held liable themselves. Further, it is questionable whether any therapeutic purpose is served when the victim goes through the trials of bringing an action and is then not awarded exemplary damages due to their exceptional nature.
Although the courts have persistently stressed that exemplary damages are not intended to compensate the plaintiff, they may still play a residual compensatory role. But it would make more sense for compensatory damages to be expanded if it is felt that victims were insufficiently compensated, rather than justifying exemplary damages on this basis. In New Zealand, because compensatory damages cannot be awarded for personal injury which is covered by the Act, and because the amount of compensation recoverable under the scheme is relatively low, it is likely that plaintiffs in exemplary damages actions will in fact be concerned with compensation. However, if the whole purpose behind the accident compensation scheme (to remove tort as a mechanism for compensating victims of accidents) is not to become redundant, then the courts must take care not to include a compensatory element in an exemplary damages award. The problem of low levels of compensation may be alleviated by the reintroduction of lump sum awards. Indeed, many have argued that these awards are going to be far too easily obtainable. These various points suggest that compensation cannot be used as a justification for awarding exemplary damages either primarily or vicariously. Another, associated, suggestion is that exemplary damages are mere monetary awards that complement compensatory damages by providing remuneration for intangible injuries such as hurt feelings, embarrassment and mental anguish. However, the current availability of expanded compensatory damages shows the fallacy of this rationale. Again, this cannot justify primary or vicarious exemplary damage awards.
The English Law Commission thought that in two categories of case, the purposes of exemplary damages could only be met by awarding such damages vicariously. These are where employees are unlikely to be able to satisfy an exemplary damages award of any significant size, and where it is problematic for a plaintiff to identify the particular wrongdoer from the employer's workforce. However, as regards the first category, this is really just a question of who has the deepest pocket, and the fact that that is the employer cannot possibly justify vicarious exemplary damages. Furthermore, it is irrelevant to consider problems of satisfying the award, as the damages are not intended to compensate. The identification problem can also be resolved without resorting to vicarious liability. The point has already been made that an employer who expressly or impliedly ratifies or adopts the wrongful act is personally liable and that ratification may include failing to take reasonable steps to identify and discipline the culpable employee. The Commission also thought that there was nothing intolerably unfair about vicarious liability in respect of exemplary damages, as the correlative to the employer deriving a benefit (financial profit) from the employee's employment is that the employer should also bear the risk of loss therefrom. Yet although this is a justification for imposing vicarious liability for compensatory damages in the first place, it does not say anything about the particular problems associated with vicarious exemplary damages. Further, that justification is concerned with ensuring persons injured in the course of business enterprises do not go uncompensated, and that concern has no part to play in a claim for exemplary damages. Vicarious exemplary damage liability cannot be justified. Not only is it unfair to punish an innocent employer, but making the employer liable will have little effective deterrent value. The other ideas advanced in support of such liability also do not hold up.
The doctrine of vicarious liability has sound justifications. These help to dictate whether and when the doctrine should apply. Consequently, vicarious liability has been imposed not just for an employee's negligent acts but also for an employee's intentional (and correspondingly criminal) acts. A new step is to find vicarious liability for sexual wrongdoing. The Supreme Court of Canada and the House of Lords have recently confronted that problem. Despite their differences in approach, the conclusion reached by each as to when vicarious liability for an employee's sexual assault ought to be imposed is the same. There must be a close connection between employment and the tort, which resulted in a material increase in the risk that the tort would occur.
In New Zealand, any action by the victim against either the perpetrator or the employer for compensation for physical or mental injury suffered as a result of the sexual assault is barred by the Injury Prevention, Rehabilitation, and Compensation Act in nearly all cases. However, an action for exemplary damages can still be brought against the wrongdoer, regardless of whether criminal proceedings have been brought or are contemplated, where the victim has suffered injury covered by the Act. But whether the victim can pursue the wrongdoer's employer for exemplary damages is a controversial question. If the arguments against holding an employer vicariously liable for exemplary damages are accepted in New Zealand, then civil claims for sexual abuse will only be able to be brought against the perpetrator. However, in the particular case where the Crown is the employer, there may be an alternative action under Baigent 's Case, giving a direct action against the Crown for breach of the New Zealand Bill of Rights Act 1990.
[*] Alison Todd is clerk to the Right Honourable Justice Tipping in the Court of Appeal, Wellington. This paper was written as part of the undergraduate Honours programme.
 1999 CanLII 692 (SCC); (1999) 174 DLR (4th) 45.
 (1999) 174 DLR (4th) 7.
  2 WLR 1311.
 PS Atiyah, Vicarious Liability in the Law of Torts (1967) 3.
 R Tobin, 'Vicarious Liability: Recent Developments' (2001) 7 New Zealand Business Law Quarterly 196, 196.
 B Feldthusen, 'Vicarious Liability for Sexual Abuse' (2001) 9 Tort Law Review 173, 174.
 Atiyah, above n 4, 12.
 See R Flannigan, 'Enterprise Control: The Servant - Independent Contractor Distinction' (1987) 37 University of Toronoto Law Journal 25; H Laski, 'The Basis of Vicarious Liability' (1916) 105 Yale Law Journal 105, 111.
 The traditional theories in justification of vicarious liability were set out in Baty's work: T Baty, Vicarious Liability (1916). Baty summarised 9 grounds that had been advanced in justification for vicarious liability up to the time of writing. Although Baty and other writers have in fact rejected the listed reasons as proper justification for the doctrine, elements of them appear in modern justifications and so will be briefly considered. 1. Control. In a sense a master controls the conduct of his or her servants and this justifies the imposition of vicarious liability. 2. Profit. The idea that a master obtains the benefit of the servant's work means the master should also bear any burdens caused by the servant's torts. 3. Revenge. This was based on the historical notion that a master's slave had to be surrendered to the victim of the wrong in order that he could wreak his vengeance on him. 4. Carefulness and Choice. The master is free to choose his or her servants and any bad choice is his or her own fault, thus justifying the master's liability to the victim of the servant's torts. 5. Identification. This ground is similar to the concept of corporate personality, whereby the act of the servant is the act of the master, as the two are identified as one. 6. Evidence. One concern of innocent tort victims was it could be difficult to identify the actual individual who committed the tortious act. If all possible perpetrators were servants of one employer, then it could only be by the imposition of vicarious liability that the plaintiff could succeed. 7. Indulgence. This ground is based on the somewhat dubious notion that people should each do their own work. Because the law allows people to employ others to do the work for them, they must pay the corresponding price of accepting liability for the servant's acts. 8. Danger. In this instance an analogy was drawn between vicarious liability and other illustrations of strict liability, such as where liability is imposed on the keeper of wild animals or the keeper of dangerous operations under the Rylands v Fletcher (1868) LR3 HL330 (HL)rule. The imposition of vicarious liability is justified because the employment of servants is similar in kind to setting in motion a dangerous activity in the above circumstances. 9. Satisfaction. This is the concept whereby vicarious liability is justified because an employer is wealthier and more capable of paying damages than the servant. This is in fact the only ground accepted by Baty as a valid reason for the doctrine. He stated that 'in hard fact, the real reason for employers' liability is the ninth: the damages are taken from a deep pocket'.
 J G Fleming, The Law of Torts (9th ed, 1998) 410.
 Adherents of the Chicago School of Law and Economics have criticised this approach. They argue that it is best to apply a cost-benefit analysis to the issue. On this view it is economically more efficient for potential victims of accidents to insure against loss, and determine who can insure at the lowest cost. This school of thought is not prominent in the courts' reasoning and further analysis is beyond the scope of this paper.
 1999 CanLII 692 (SCC); (1999) 174 DLR (4th) 45.
 Ibid 59.
 Ibid 59, 60.
  EngR 1005; (1839) 6 Cl. &F. 894, 910.
 Of course, the deepest pocket principle is based on the assumption the employer is in fact better able to compensate for the victim's loss. It has been said that cases where the employer's 'risk- bearing capacity' is not superior to the victim's are statistically too insignificant to throw doubt on that policy. Fleming, above n 11, 410.
 For a full exploration of the corrective justice rationale for vicarious liability, see R Townsend- Smith, 'Vicarious Liability for Sexual (and other) Assaults' (2000) 8 Tort Law Review 108, 121- 4. Towsend-Smith, however, does warn against construing the corrective justice argument as little more than an argument that employers should be liable on the basis they have the deeper pocket.
 Flannigan, above n 9, 28.
 W Douglas, 'Vicarious Liability and Administration of Risk I' (1929) 38 Yale Law Journal 584, 586. Atiyah has pointed out the consumer may also play a part in spreading the cost, as not all consumers are individuals but may be corporations themselves. Thus the cost of tort liabilities can be spread over a considerable section of the public: Atiyah, above n 4, 23.
 Y Smith, 'Frolic and Detour' (1923) 23 Columbia Law Review 444, 456-7. Smith reaches this conclusion after drawing a comparison to justifications offered in support of workmen's compensation statutes. He believes that 'if it is socially expedient to spread and distribute throughout the community the inevitable losses occasioned by injuries to employees engaged in industry, is it not also socially expedient to spread and distribute the losses due to injuries to third persons which are equally inevitable'. (457).
 Flannigan, above n 9, 29.
 Ibid 34.
 Bazley v Curry 1999 CanLII 692 (SCC); (1999) 174 DLR (4th) 45.
 Ibid 60.
 An analogy can be drawn with the Health and Safety in Employment Act 1992 (NZ), which has deterrence as a broad policy objective of the Act. Under this Act employers may be criminally liable for unsafe acts performed by employees. But the employer's liability is not truly vicarious, since it can arise only from proven fault in aspects of employment, such as a failure to train or to supervise: Campbell, 'State Regulation of Occupational Health and Safety' in Slappendel (ed), Health and Safety in New Zealand Workplaces (1995) Ch 3, as referred to in S Todd (ed), The Law of Torts in New Zealand (3rd edn, 2001) Ch 22 ( J Hughes), 1084-5.
 Feldthusen makes the point that at a certain stage, it becomes less expensive to allow the injury than to prevent it. He suggest when that is so, the law should not hold the employer negligent for failing to incur a greater loss to prevent a lesser one. See B Feldthusen, 'Vicarious Liability for Sexual Torts' in N J Mullany and A M Linden, A M (eds), Torts Tomorrow: A Tribute to John Fleming (1998) 225.
 See S Todd (ed), TheLaw of Torts inNew Zealand (3rd ed, 2001) 18, 19.
 K Davis, 'Vicarious Liability, Judgment Proofing, and Non-Profits' (2000) 50 University of Toronto Law Journal 407, 409.
 Ibid 410.
 S Croley, 'Vicarious Liability in Tort: On the Sources and Limits of Employee Reasonableness' (1996) 69 Southern California Law Review 1705, 1737, as referred to in Davis, above n 29, 410.
 B Chapman, 'Corporate Tort Liability and the Problem of Overcompliance' (1996) 69 Southern California Law Review 1705, 1737, as referred to in Davis, above n 29, 411.
 Feldthusen, above n 27, 224.
 Related to this is the question of sexual harassment. An employer can face liability for sexual harassment committed by one employee against another under the Employment Relations Act 2000. This liability, however, is not vicarious, since it arises only if the employer fails to take practicable steps to prevent such harassment once it is brought to his or her notice: s 108. This can be compared to s 68 Human Rights Act 1993, whereby an employer is liable for sexual harassment committed by employees, unless he or she took all reasonably practicable steps to prevent the employee from doing the act: S Todd (ed), above n 28, Ch 22 (J Hughes) 1084-5.
 It should be noted the most obvious cases where the doctrine would be thought to apply - where the employee's act was expressly or impliedly authorised or ratified - are in fact cases where the employer would be primarily, as opposed to vicariously, liable: S Todd (ed), above n 28, Ch 22 (J Hughes) 1084-5.
 J Salmond, Salmond on Torts (1st ed, 1907) 83-4..
 Hughes has questioned whether this reluctance is due to the fact the underlying policy considerations are less applicable where employees commit intentional wrongs, or whether it is just more difficult to classify intentional wrongdoing as one way of performing employment obligations: Todd (ed), above n 28, Ch 22, 1084-5.
 F D Rose, 'Liability for an Employee's Assaults'  Modern Law Review 420, 422.
 Daniels v Whetsone Entertainments Ltd  2 Lloyd's Rep 1.
 Dyer vMunday  1 QB 742.
 Similarly, in Petterson v Royal Oak Hotel Ltd  NZLR 136, a bar attendant had been instructed to keep order in the bar. In retaliation to a glass being thrown at him by an intoxicated customer, and as an expression of personal resentment, the barman threw part of the glass back at the offending customer, injuring another customer in the process. The employer was vicariously liable for this action, as the throwing of the glass was a wrongful mode of keeping order. This can be compared to Deatons Pty Ltd v Flew  HCA 60; (1949) 79 CLR 370, where an employer was not liable for a bar attendant's actions in throwing a beer glass, as the action was not incidental to keeping order. This decision was followed in Auckland Workingmen 's Club and Mechanics Institute v Rennie  1 NZLR 278.
 There are many further examples. An employer was held to be vicariously liable for a porter's actions in violently ejecting a passenger from a train after a mistaken belief that he was on the wrong train in circumstances where part of the porter's duties was to see that passengers were on the right service: Bayley v Manchester, Sheffield and Lincolnshire Railway Co (1873) LR 8 CP 148. Compare Rutherford' v Hawke 's Bay Hospital Board  NZLR 400 where an employee who was under instructions to ensure the job ran sweetly argued with another employee about the efficiency of the work they were doing, and subsequently assaulted and killed that employee. The employer was not liable for this, as it was not part of the employee's duty to use physical force on other employees. In Keppel Bus Co Ltd v Sa 'ad bin Ahmad  UKPC 2;  2 All ER 700 part of a bus conductor's duties was to keep order but the employer was not vicariously liable when the conductor struck a passenger as there was no disorder at that time.
  UKHL 1;  AC 716.
  1 QB 716.
 The employee was put in charge of cleaning the fur and the manner in which he conducted himself in doing that work was to convert it. What he was doing, albeit dishonestly, he was doing in the scope or course of his employment: Diplock LJ, 736-7.
 This was the analysis taken by the English Court of Appeal in Trotman v North Yorkshire County Council  LGR 584.
 Compare Leesh River Tea Co Ltd v British India Steam Navigation Co Ltd [ 1966] 3 All ER 593 where an employee stole a cover plate off a ship which he had been employed to load. The employer was not vicariously liable for this act of theft, as it was unconnected to the loading or unloading of the ship.
 1999 CanLII 692 (SCC); (1999) 174 DLR (4th) 45.
 (1999) 174 DLR (4th) 71.
  2 WLR 1311.
 For example, EDG v Hammer  BCJ No.992 (QL) (SC); Goodwin v Commission Scolaire Laurenval  RRA 673, 8 CCLT (2d) (Que SC).
 In B (J-P) v Jacob (1998) 166 DLR (4th) 125, a male nurse sexually assaulted a sleeping patient. Again, the employment merely provided the opportunity for the assault to occur and the hospital corporation was not vicariously liable.
 McDonald v Mombourquette (1996) 152 NSR (2d) 109 (CA).
 Ibid 123. See also two similar US cases where vicarious liability was denied for sexual assaults committed by clergymen: Destefano v Grabrian  USCA6 1854; 763 P 2d 275 (Colo, 1988); Tichenor v Roman Catholic Church of the Archdiocese of New Orleans  USCA5 2567; 32 F 3d 953 (5th Cir, 1994).
 K (W) vPornbacher (1997) 32 BCLR (3d) 360 (SC). Of course, in order for vicarious liability even to be considered, it must first be found that an employer-employee relationship existed between the Church and the priest. In both Pornbacher and Mombourquette, the priest was held to be an employee of the Church, as all the common law indicia of an employment relationship were present. This point was not in real dispute. It may be more of an issue in New Zealand, as in Mabon v The Conference of the Methodist Church ofNZ  NZEmpC 283;  ERNZ 690 it was held a Minister was not an employee of the Methodist Church. This was affirmed by the Court of Appeal: (1998) 5 NZELC 95, 834. However, the decision was made on the facts of the case and did not necessarily extend to the relationship between other churches and ministers. See MacKay (ed), Butterworths Employment Law Guide (5th ed, 2001) Ch 6 for further discussion about classifying employment relationships.
 Boykin v District of Columbia 484 A 2d 560 (1984).
 Ibid 562. It was argued that because physical touching between teacher and pupil was an essential activity, the subsequent abuse was an unauthorised mode of performing an authorised activity. This argument was rejected. The Court of Appeal stated 'we do not believe that a sexual assault may be deemed a direct outgrowth of a school official's authorisation to take a student by the hand or arm in guiding her past obstacles in the building'.
 John R v Oakland Unified School District 769 P 2d 948 (1989).
 However, it is not at all clear that preventive measures would have an overly negative impact on school activities. One writer has stated that 'simply requiring that two teachers be present at off- campus events and encouraging school administrators to enter classrooms randomly may serve as effective deterrents with minimal costs. Furthermore...these additional measures do not necessarily intrude into the student-teacher relationship. The trust that teachers are encouraged to foster with their students is more severely undermined by a teacher's sexual assault on a student than by the safeguards designed to prevent such assaults. These safeguards should not cause students to form less substantial relationships with their teachers, especially because most students will be unaware of the reasons for the restrictions.' Anon, 'Case Note: Mary M v City of Los Angeles' (1992) 105 Harvard Law Review 947.
Further, there is definitely potential for a dangerous power relationship to develop between students and teachers, thus creating the opportunity for its exploitation. This potential power relationship can arise because teachers are encouraged to foster trusting, dependent relationships with their students with whom they interact on a daily basis in a way that continually reinforces the authoritarian relationship. Anon, 'Case Note: Mary M v City of Los Angeles' (1992) 105 Harvard Law Review 947.
 Mary M v City of Los Angeles 814 P 2d 1341 (1991).
 Ibid 1343. The court set out the policies underlying why risk was allocated in this way. First, it promoted deterrence of future tortious conduct. Secondly, it gave greater assurance of compensation for the victim. Thirdly, it ensured that the victim's losses will be equitably borne by those who benefit from the enterprise that gave rise to the injury.
  LGR 584.
 B Harvey and J Marston, ‘ST v North Yorkshire County Council  IRLR 98'  Journal of Social Welfare and Family Law 376, 382.
 The British Columbia Court of Appeal's decision in B (PA) v Curry 146 DLR (4th) 72 would have provided a useful starting point.
 The chambers judge, Lowry J, held that the foundation was vicariously liable and the foundation's appeal to the British Columbia Court of Appeal was dismissed. The Court of Appeal's decision was applied in several subsequent decisions to impose vicarious liability in similar factual situations. For example, in B (KL) v British Columbia (1998) 51 BCLR(3d) 1 (SC)the Crown was held vicariously liable for foster parents' abuse and neglect of children in their care.
 In the Court of Appeal, although all five justices agreed that vicarious liability should be imposed, they had rejected the Salmond test as the appropriate tool to use in this context, and applied two different tests: a 'conferral of authority' test (given by Madam Justice Huddart) and a 'sufficient connection' test (given by Madam Justice Newbury).
 Para 18.
 Para 19.
 Para 20-1.
 See above Part II.
 Para 37.
 P Cane, 'Vicarious Liability for Sexual Abuse' (2000) 116 Law Quarterly Review 21.
 Para 58.
 N Des Rosiers, 'From Precedent to Prevention - Vicarious Liability for Sexual Abuse' (2000) Tort Law Review 27,29.
 Cane, above n 73, 25.
 M Hall, 'Responsibility Without Fault: Bazley v Curry' (2000) 79 Canadian Bar Review 474 , 485.
 Cane, above n 73, 26.
 See N Des Rosiers, above n 75.
 This is hardly surprising considering the same Justices sat on each case.
 The decision of the British Columbia Court of Appeal, (T (G) v Griffiths (1997) 31 BCLR (3d) 1) declining to impose vicarious liability, was thus upheld.
 Para 80.
 Para 17.
 For discussion of relationships involving power by one party over the other see C Itzin (ed), Home Truths About Child Sexual Abuse (2000); Feldthusen, above n 27, 230-7.
 See Feldthusen, above n 6, 177; Des Rosiers, above n 75, 31.
 This issue is outside the scope of this discussion but see paras 68-78 of Jacobi for the arguments for and against this proposition.
 Todd (ed), above n 28, Ch 22 ( J Hughes) 1106.
 The Bazley approach has been applied in several recent Canadian decisions: B(M) v British Columbia (2001) 197 DLR385; B(KL) v British Columbia (2001) 197 DLR 431. The High Court of Australia also followed this approach in Hollis v Vabu Pty Ltd  HCA 44, holding that an enterprise is liable for the cost of injury to a third party if the tortious conduct of a person representing that enterprise may fairly be said to be characteristic of that enterprise.
  2 WLR 1311.
 The appellants, at first instance, had alleged that the employers were negligent in their care, selection and control of the warden. This claim was dismissed by the trial judge and was not further pursued.
 See above Part III.
 The trial judge did however hold the employers were vicariously liable for the warden's failure to report to his employers his intentions (before the acts of sexual abuse) and the harmful consequences to the children (after acts of abuse). This finding was appealed. The House of Lords, having found the employers vicariously liable for the warden's torts, did not need to express a concluded view about the employers' vicarious liability for the employee's alleged breach of a duty to report his sexual intentions or the consequences of his misdeeds. It was acknowledged, however, that the issue may require further consideration at some point. This does, however, seem a bizarre and artificial way of overcoming Trotman. With the House of Lords deciding as it did on the issue of vicarious liability, this issue is unlikely to arise again.
 Para 28.
 Lord Steyn (para 25) stated that Trotman:
resulted in the case being treated as one of the employment furnishing a mere opportunity to commit the sexual abuse. The reality was that the county council were responsible for the care of the vulnerable children and employed the deputy headmaster to carry out that duty on its behalf. And the sexual abuse took place while the employee was engaged in duties at the very time and place demanded by his employment. The connection between the employment and the torts was very close.
Lord Steyn also agreed with McLachlin J's criticism of Trotman in Bazley where she stated (para 24) that:
[t]he opinion's reasoning depends on the level of generality with which the sexual act is described. Instead of describing the act in terms of the employee's duties of supervising and caring for vulnerable students during a study trip abroad, the Court of Appeal cast it in terms unrelated to those duties. Important legal distinctions should not turn on such semantics. The other Lords overruled Trotman on similar lines.
 Para 37.
 Para 69-70.
 Para 55.
 G Douglas, 'Vicarious Liability' (2001) 31 Family Law 595, 596.
 I Smith, 'Employment Law Brief’ (2001) 151 New Law Journal 1243, 1243.
 It would be interesting to know precisely why their Lordships refused to take the policy considerations from Curry into account.
 B Feldthusen, above n 6, 178.
 R Coe, 'A New Test for Vicarious Liability' (2001) 151 New Law Journal 1154, 1155.
 C Foster, 'Vicarious Liability - Back to Basics' (2001) 145 Solicitors Journal 554, 554.
 BNZ v NZ Guardian Trust  1 NZLR 664, 683 (CA).
 See Todd (ed), above n 28, 1005 for further discussion of this area of the law.
 This Act comes into force on 1 April 2002. All further section references are to the 2001 Act. The provisions under consideration are substantially the same as in its predecessor - the Accident
 Insurance Act 1998. The discussion hereafter is as if the Act is already in force.
 Todd (ed), above n 28, Ch 2, 74.
 The offences are sexual violation; attempt to commit sexual violation; inducing sexual connection by coercion; incest; sexual intercourse with girl under care or protection; sexual intercourse with girl under 12; indecency with girl under 12; sexual intercourse or indecency with girl between 12 and 16; indecent assault on woman or girl; sexual intercourse with severely subnormal woman or girl; indecent act between woman and girl; indecency with boy under 12; indecency with boy between 12 and 16; indecent assault on man or boy; anal intercourse; compelling indecent act with animal; assault on a child, or by a male on a female; infecting with disease; female genital mutilation; further offences relating to female genital mutilation.
 Note that s 21(5)(a) provides that there is cover under the Act irrespective of whether any person has been charged or convicted of a criminal offence involving the conduct alleged.
 The major change the Injury Prevention, Rehabilitation and Compensation Act 2001 brings about is the reintroduction of lump sum payments for permanent incapacity, including mental injury caused by a listed sexual assault. These lump sum payments aim to provide compensation for non-economic loss. The eligibility for and amount of the lump sum payments are set out in the new Act and regulations.
 See L v Robinson  3 NZLR 499, where professional misconduct of a sexual nature resulting in mental injury was not covered by the Accident Insurance Act 1998.
  1NZLR97.
  1NZLR81.
 Per Richardson J, 90.
 3 NZLR622.
 Indeed, one of the most frequently litigated areas of exemplary damages in New Zealand in the 1990s was sexual abuse. See B Blower, 'Exemplary Damages: Applying Natural Justice to Ensure Fundamental Fairness'  New Zealand Law Review 313, 338. Plaintiffs in sexual abuse cases argue they were intentionally injured by a defendant with contumelious disregard of their legal rights.
 Compare the recent Court of Appeal decision in Residual Health Management Unit v Downie (CA 147/01, 11.12.2001, Gault, Heron and Gendall JJ), where a case of medical negligence failed to meet the Bottrill test for an award of exemplary damages.
  NZCA 3;  3 NZLR 22.
 This reasoning applied similarly to discharge following conviction. The criminal court had decided the appropriate degree of punishment and so there was no basis for an award of exemplary damages.
 The decision was approved by a majority of the High Court of Australia in Gray v Motor Accident Commission  HCA 70; (1998) 73 ALJR 45. However, the High Court limited the bar on exemplary damages to cases where the criminal law had exacted a 'substantial' criminal penalty for 'substantially the same conduct'.
 See Wv W  2 NZLR 1.
 Manning has argued that general deterrence is regarded as a primary goal of exemplary damages, and on this view the concern about double punishment loses much of its force: J Manning, 'Daniels v Thompson: Double Punishment of Double Trouble?'  New Zealand Law Review 721, 728. See also J Manning, 'Exemplary Damages and Criminal Punishment in the Privy Council' (1999) 7 Tort Law Review 129, 133. According to Manning, the majority decision also created an anomaly. Victims of intentional wrongs that did not involve personal injury and did not constitute crimes (notably defamation) were unaffected by the bar on compensatory damages in the accident compensation scheme and the bar on exemplary damages in Daniels. Victims of unlawfully inflicted injury to the person, however, were faced with both bars. But the first alleged anomaly has nothing to do with the Daniels bar. As regards to the second, the double punishment/jeopardy argument does not exist. It is true the victim receives the exemplary damages but not (necessarily) the criminal penalty. Arguably a better option was for the criminal court to make an order for compensation in respect of the emotional harm suffered by the victim: s 22(1) Criminal Justice Act (NZ) 1985. Again, the Court could direct the whole or part of a fine be paid to the victim: s 28(1) Criminal Justice Act (NZ) 1985. See R Mackenzie, 'Lump Sums or Litigation? Compensation for Sexual Abuse. The Case for Reinstatement of a Compensation for Criminal Injuries Scheme' (1993) 15 New Zealand Universities Law Review 367, 391 for further analysis of these sections. Mackenzie argues that an updated Criminal Injuries Compensation Scheme should be reinstated, as the accident compensation scheme fails to provide adequate compensation to personal injury victims.
 Todd has outlined further reasons why the majority reasoning can be favoured. First, it had been argued the majority decision provided an incentive to women to elect to avoid the criminal justice system. However, even if no absolute bar existed, it would still be exceptional to make an award, thus still leaving a disincentive to co-operate with the police. Secondly, it was suggested the majority ruling disenfranchised victims of crime, especially women, of a legal right. But victims are not forced to make a criminal complaint or to help the police, thus leaving the path open for a civil proceeding: S Todd, 'Exemplary Damages' (1998) 18 New Zealand Universities Law Review 145, 175-7.
 Ibid, 178-9.
 The precise wording of s 319 is as follows:
(1) Nothing in this Act, and no rule of law, prevents any person from bringing proceedings in any court in New Zealand for exemplary damages for conduct by the defendant that has resulted in -
(a) Personal injury covered by this Act; or
(b)Personal injury covered by the former Acts.
(2) The court may make an award of exemplary damages for conduct of the kind described in subsection (1)even though -
(a) The defendant has been charged with, and acquitted or convicted of, an offence involving the conduct concerned in the claim for exemplary damages; or
(b) The defendant has been charged with such an offence, and has been discharged without conviction under section 19 of the Criminal Justice Act 1985 or convicted and discharged under section 20 of that Act; or
(c) The defendant has been charged with such an offence and, at the time at which the court is making its decision on the claim for exemplary damages, the charge has not been dealt with; or
(d) The defendant has not, at the time at which the court is making its decision on the claim for exemplary damages, been charged with such an offence; or
(e) The limitation period for bringing a charge for such an offence has expired.
(3) In determining whether to award exemplary damages and, if they are to be awarded, the amount of them, the court may have regard to -
(a) Whether a penalty has been imposed on the defendant for an offence involving the conduct concerned in the claim for exemplary damages; and
(b) If so, the nature of the penalty
 Todd (ed), above n 28, 91.
  3 NZLR 424.
  NZLR 1106.
 Per Henry J, p 1111.
 (27.3.85, Prichard J, HC, Auckland, A 617/82); noted  New Zealand Recent Law Review 342. See also Harris v Attorney-General (23.7.99, Durie J, HC, Masterton, CP 7/96).
 J Smillie, 'Exemplary Damages for Personal Injury'  New Zealand Law Review 140, 166. This is discussed in more detail below.
 Ibid 168.
 See, for example, Miller v Blanton 210 SW 2d 293 (1948); Davis v Merrill Lynch, Pierce, Fenner & Smith Inc  USCA8 720; 906 F 2d 1206 (1990);Al Parker Buick v Touchy 788 SW 2d 129 (1990).
 W P Keeton (ed), Prosser and Keeton on the Law of Torts (1984).
 Lake Shore & Michigan Southern Railway Co vPrentice  USSC 9; 147 US 101, 13 S Ct 261, 37 L Ed 97 (1893).
 See, for example, Lake Shore & Michigan Southern Railway Co v Prentice 147 (US) 101 (1893); Parris v St Johnsbury Trucking Co  USCA2 353; 395 F 2d 543 (1968); Trauth v Dunbar 448 NE 2d 1368 (1983).
 Todd, above n 124, 182.
 Further weaknesses of the Restatement (Second) of Torts rule can be found in T Moran, 'Punitive Damages in Fair Housing Litigation' (2001) 36 Harvard Civil Rights — Civil Liberties Law Review 297, 330-2.
 Lancashire County Council v Municipal Mutual Insurance Ltd  3 WLR493.
  UKHL 29;  2 WLR 1789.
 Formerly claims were confined to those causes of action which were recognised as supporting claims for exemplary damages prior to the decision in Rookes v Barnard  UKHL 1;  AC 1129 (HL). Kuddus removed this restriction.
 Para 131.
 See Taylor v Beere  NZCA 15;  1 NZLR 81; Donselaar v Donselaar  NZCA 13;  1 NZLR 97 and, more recently, TVNZ Ltd v Quinn  3 NZLR 24 and Daniels v Thompson  NZCA 3;  3 NZLR 22.
 Owen has suggested imposing punishment advances several goals. First, it helps restore the plaintiff's equilibrium, as when the defendant is punished by the judicial system, the injured plaintiff can see the defendant suffer. Secondly, imposing punishment is a form of revenge for the public at large, for a person who intentionally or recklessly injures another has breached a norm of societal behaviour. A third and related goal is that punishing the law-breaker indirectly rewards the law-abider and educates the offenderto society's legal values. David Owen, 'Punitive Damages In Products Liability Litigation' (1976) 74 Michigan Law Review 1257, 1279-81.
 See Cooke J inDonselaar, 104, 106-7 and Richardson P in Bottrill v A  3 NZLR 622, para 44.
 R Stern and J Loughhead, 'Vicarious Liability for Punitive Damages: The Worst Side of a Questionable Doctrine'  29 Defense Counsel Journal 32-3.
 D Ellis, 'Fairness and Efficiency in the Law of Punitive Damages' (1982) 56 Southern California Law Review 1, 65.
 English Law Commission, Aggravated, Exemplary and Re stitutionary Damages, Report No 247 (1997), para 5.216.
 Stern and Loughhead, above n 148, 35.
 Ellis, above n 149, 69.
 F Miller, 'Exemplary Damages in New Zealand: A Law and Economic Analysis' (1997) 3 New Zealand Business Law Quarterly 228, 250.
 English Law Commission, above n 150, para 5.220. The Commission believes employers who are potentially liable for exemplary damages will have an incentive to control and educate their workforces. Further, employers' disciplinary powers will enable them to penalise and deter individual guilty employees, or discourage potential wrongdoers. The Commission goes as far as saying the potential loss of employment, coupled with impaired employment prospects, may be a more severe form of sanction for wrongdoing by employees than a punitive damages award could directly provide.
 Stern and Loughhead, above n 148, 34.
 Tolle v Interstate Systems Truck, 41 Ill App 3d 771, 356 NE 2d 625 (1976).
 R Cooter, 'An Economic Analysis of Punitive Damages' (1982) 56 Southern California Law Review 79.
 Moran, above n 40, 327.
 This point has been well summarised by Stern and Loughhead, above n 148, 34:
If it costs less to pay the damages than to prevent the conduct which gives rise to them, the goal of deterrence will not be achieved. In light of the fact that many episodes of agents' wrongful actions will result from causes over which the principal cannot exercise control, it stands to reason that the principal will not deem it financially desirable to expend its valuable resources in attempting to prevent what is not preventable.
 Todd, above n 124, 183. See generally Elliot, 'Why Punitive Damages Don't Deter Corporate Misconduct Effectively' (1989) 40 Alabama Law Review 1053.
 T Olsen and T Boutrous, 'Punitive Damages' (1991) 77 American Bar Association Journal 40, 40.
 Ellis, above n 149, 69.
  NZCA 3;  3 NZLR 22.
 For New Zealand examples, see Daniels v Thompson [  NZCA 3; 1998] 3 NZLR 22 and Bottrill v A  3 NZLR 622.
 See D McLoughlin, 'Cash for Unproved Sex Abuse', The Dominion, Wellington, New Zealand, 9 January 2002; A Moffat, 'Law Firm In Bid for Sex Abuse Cases', The Press, Christchurch, New Zealand, 9 January 2002; Editorial, 'The Lump-Sum Gravy Train', The Dominion, Wellington, New Zealand, 11 January 2002.
 See, for example, Hink v Sherman 164 Mich 352, 357, 129 NW 732, 734 (1911) and Stuart v Western Union Tel Co. 66 Tex 530, 18 SW 351 (1885).
 J Sales and K Cole, 'Punitive Damages: A Relic that has Outlived Its Origins' (1984) 37 Vanderbilt Law Review 1117, 1130. Damages for mental upset are readily available following the commission of an intentional tort.
 Para 5.221.
 Todd, above n 124, 183.
 Smillie, above n 133, 168.
 Para 5.223.
 Todd, above n 124, 184.
 The American complicity rule may be more appropriate, whereby an employer is only vicariously liable for exemplary damages where the employer authorised, ratified, or approved of the wrongful act of its employee. However, these can actually be seen as instances of primary liability.
  3 NZLR 667.
 S Todd, 'Tort Review'  New Zealand Law Review 537.