VICARIOUS LIABILITY IN THE AGENCY CONTEXT

This article will consider the law in relation to whether or not a principal may be vicariously liable for the unauthorised or tortious acts of his or her agent. This article will identify and consider the substantial conflicting academic and judicial approaches to this issue. The impetuses for this article are the recent judicial pronouncements of McHugh J in Scott v Davis and Hollis v Vabu. In these judgements, McHugh J adopts a significantly different approach to the other members of the bench in resolving the questions before the court. In both cases, rather than relying on the traditional distinction between employees and independent contractors to found liability, his Honour approached the cases on the basis of agency principles. In each case, McHugh J found that the principal was vicariously liable for the acts of its agent. This article will consider McHugh J's judgements in detail but will also consider the broader issue of whether or not the law does, or should, recognise the vicarious liability of principals for the acts of their agents.


INTRODUCTION
This article will consider the law in relation to whether or not a principal may be vicariously liable for the unauthorised or tortious acts of his or her agent. This article will identify and consider the substantial conflicting academic and judicial approaches to this issue. The impetuses for this article are the recent judicial pronouncements of McHugh J in Scott v Davis 1 and Hollis v Vabu. 2 In these judgements, McHugh J adopts a significantly different approach to the other members of the bench in resolving the questions before the court. In both cases, rather than relying on the traditional distinction between employees and independent contractors to found liability, his Honour approached the cases on the basis of agency principles. In each case, McHugh J found that the principal was vicariously liable for the acts of its agent. This article will consider McHugh J's judgements in detail but will also consider the broader issue of whether or not the law does, or should, recognise the vicarious liability of principals for the acts of their agents.
At the outset, it is worth noting that this area of the law is beset not only by inherent terminological ambiguities, but also by undisciplined use of that terminology. Together, these factors mean that an examination of fundamental principles is necessary in order to analyse the true current state of the law.
Vicarious liability arises when one person is held liable for the tort of another. It is a form of strict liability. Vicarious liability arises by virtue of the relationship between the actual tortfeasor and the person who is made vicariously liable. As Atiyah recognises, only the master and servant relationship has a general rule imposing vicarious liability in all circumstances. 3 The remaining 'recognised' categories have vicarious liability imposed in certain exceptional circumstances only (if truly at all).
Given the topic of this article, in assessing whether or not the principles of vicarious liability apply to the principal/agent relationship, it is first necessary to define what is meant by the term 'agent' in the legal sense. It is commonly recognised that a precise definition of the term 'agent' is elusive, if not impossible. (a) those that can create legal relations on behalf of a principal with a third party; (b) those that can affect legal relations on behalf of a principal with a third party; and (c) a person who has authority to act on behalf of a principal. 4 According to Bowstead: [A]gency is the fiduciary relationship which exists between two persons, one of whom expressly or impliedly consents that the other should act on his behalf so as to affect his relations with third parties, and the other of whom similarly consents so to act or so acts. The one on whose behalf the act or acts are to be done is called the principal. The one who is to act is called the agent. Any person other than the principal and the agent may be referred to as the third party. 5 In Scott v Davis Gummow J endorsed an earlier High Court definition 6 that the term 'agent' is used to 'connote an authority or capacity in one person to create legal relations between a person occupying the position of principal and third parties'. 7 This is similar to Dal Pont's first, most restricted, definition noted above. Bowstead's definition is closer to Dal Pont's second definition. Obviously, the broader the definition of agent, the greater the scope for potential findings of vicarious liability in the agency context.
In fact, the term 'agent' has been used in circumstances broader than what are captured in the above definitions. This is particularly so in relation to the 'car cases' 8 to be considered. As pointed out by Lord Wilberforce in Morgans v Launchbury, 9 use of the term 'agent' in this context 'is merely a concept, the meaning and purpose of which is to say "is vicariously liable" and that either expression reflects a judgement of valuerespondeat superior is the law saying that the owner ought to pay'. 10 As noted by Gleeson CJ in Scott v Davis: Lord Wilberforce made the point that to describe a person as the agent of another, in this context, is to express a conclusion that vicarious liability exists, rather than to state a reason for such a conclusion. Nevertheless, some judges refer to agency as a criterion of liability, similar to employment. If that is to be done, it is necessary to be more particular as to what is meant. 11 In other words, the term agency is used to find that a relationship creating vicarious liability has been established, not that the elements of a principal/agent relationship have been found to exist.
Given the above range of indicative definitions, it can be seen that the concept of who is an agent is difficult to lay down with certainty. Further, contextual considerations will 4 G E Dal Pont, Law of Agency (Butterworths, 2001)  be very important in determining whether a person is or is not considered to be an agent in any given case.
However, an important aspect of agency is that it is underpinned by contractual notions and principles. Indeed, the very existence of an agency relationship is generally, but not always, based in contract. For example, in Yasuda Fire & Marine Insurance v Orion Marine Insurance 12 Colman J held that an agency relationship is 'almost invariably founded upon a contract between principal and agent'. 13 Agency is essentially a creature of contract, however, this fact inevitably causes confusion and doctrinal tension when it is sought to mesh agency concepts with tortious concepts, such as vicarious liability. Now that these basic concepts have been examined, it is necessary to consider the recognised relationships that sustain vicarious liability.

II RECOGNISED RELATIONSHIPS ESTABLISHING VICARIOUS LIABILITY A Master/Servant
The classic relationship that generates vicarious liability is that of master and servant. A master will be vicariously liable for the torts of his or her servant committed during the course of the servant's employment. The categorisation of a person as a servant is critical in this context as a master will not be liable per se for the torts an independent contractor.
Much has been written about the different tests that are to be used to distinguish between a servant and an independent contractor. 14 However, for the purposes of this article it is unnecessary to do so in detail. The fundamental distinction between a servant and an independent contractor is that a servant is employed under a contract of service whilst an independent contractor is employed under a contract for services. Another essential difference is that an independent contractor acts on his or her own behalf as principal, not on behalf of the employer. Nevertheless, it is not always easy to distinguish between the two relationships.
The High Court most recently considered this issue in Hollis v Vabu. 15 The facts were that an unidentified courier cyclist had negligently injured a pedestrian during the course of his couriering activities. The pedestrian sued the courier company (the 'respondent') who engaged the couriers. Under traditional principles, the respondent could only be liable if the couriers were considered to be employees of the respondent, not independent contractors. Therefore, the High Court had to consider the nature of the couriers' engagement and whether this constituted an employment relationship or an independent contractor relationship. A strong majority of the court 16 found that, on the facts, the couriers were employees of the respondent and consequently the respondent was vicariously liable for the tort of its employee. Whilst there were factors which indicated the existence of an independent contractor relationship, these were outweighed by other factors evidencing an employment relationship. 17 How then, if at all, does the law of agency impact on the above matters? For the purposes of agency law, both a servant and an independent contractor may, depending on the circumstances, also be considered an agent in the legal sense. Further, and as noted in the introduction, McHugh J in this case did not adopt the traditional servant/independent contractor distinction. Rather, he found the respondent liable on the basis of agency principles. Both of these matters will be considered below.
Once we move beyond a consideration of vicarious liability in the master/servant context, confusion reigns. According to Fleming, '[v]icarious liability is incident only to a relationship of controlled employment, traditionally described as that of 'master and servant' (emphasis added). 18 Despite this assertion, there are a number of 'generally recognised' categories of vicarious liability that will need to be considered and analysed before Fleming's statement can be assessed.

B Disparate Principal/Agent Categories
There is much debate about whether or not the relationship of principal and agent is one that attracts vicarious liability. Whilst not universally accepted as such, there are a number of specific 'principal/agent' situations where courts have consistently found that vicarious liability will be established. These specific categories will be reviewed and analysed initially and then an analysis will be undertaken of vicarious liability as it applies to the principal/agent relationship more generally.

1
The 'Car Cases' 19 In what has been described as a 'unique' 20 line of cases, owners of motor vehicles have, in certain circumstances, been found to be vicariously liable for the acts of persons driving their motor vehicle. The basis of liability has ostensibly been on the ground that the driver was the agent of the owner.
The main High Court authority in this area is Soblusky v Egan. 21 The facts were that the owner of a vehicle was asleep in the passenger's seat at the time the driver negligently caused an accident. Another passenger sued the owner on the basis of vicarious liability. After reviewing a line of earlier English cases, the court held that: The owner or bailee being in possession of the vehicle and with full legal authority to direct what is done with it appoints another to do the manual work of managing it and to do this on his behalf in circumstances where he can always assert his power of control. Thus it means in point of law that he is driving by his agent. It appears quite immaterial that Soblusky went to sleep. That meant no more than a complete delegation to his agent during his unconsciousness. The principle of the cases cited is simply that the management of the vehicle is done by the hands of another and is in fact and law subject to direction and control.

22
It is interesting to note that the court was at pains to point out that the finding of liability was not based on any new doctrine or principle or the new application of any old doctrine or principle. 23 According to the judges, it was an 'obvious' 24 case. In the words of Atiyah, they 'denied that the car-driver cases were anomalous and affirmed that they were in accordance with ordinary principles under which a principal is liable for his agent's torts '. 25 The other important case in this context is Morgans v Launchbury. 26 In this case, the House of Lords held that in order to affix liability on the owner of a car for the negligence of its driver, it was necessary to show either that the driver was the owner's servant or that, at the material time, the driver was acting on the owner's behalf as his agent. To establish the existence of the agency relationship it was necessary to show that the driver was using the car at the owner's request, express or implied, or on his instructions, and was doing so in performance of the task or duty thereby delegated to him by the owner. The fact that the driver was using the car with the owner's permission and that the purpose for which the car was being used was one in which the owner had an interest or concern, was not sufficient to establish vicarious liability.
In light of even the broadest definitions of agency considered earlier, the use of the concept of agency in this context to found liability has been criticised as artificial and wrong in principle. 27 Bowstead submits that liability does not, in truth, stem from agency principles at all. 28 After a detailed review of the development of the law in this area, Keeler concludes that agency principles do not support the imposition of liability in this context and that the pragmatic explanation is that liability has been imposed upon the basis of the availability of insurance for the owner. 29 More generally, Fridman notes that: In discussing agency in relation to tortious liability, the idea that an agent is one who has power to effect legal relations between his principal and third parties, must be taken to mean 'legal relations' in the sense of liabilities in tort, as well as contractual and proprietary rights and duties. This would therefore include within the scope of agency Of course, the obvious problem with this statement is that it states a conclusion, not a test whereby the notion of agent can be elucidated. The issue in the car case context is whether or not the driver is properly labelled an agent, not the consequences of what that labelling will entail.
The High Court came to consider the above principles in Scott v Davis. 31 The facts were as follows. The respondent owned an aeroplane. During a birthday party at his property, a guest asked if his son could have a joy ride. The respondent arranged for another guest at the party, who was a licensed pilot, to conduct the flight. Due to the negligence of the pilot the plane crashed, seriously injuring the appellant and killing the pilot. The issue was whether the respondent was vicariously liable for the negligence of the pilot. Whilst this was not the paradigm case of vicarious liability, on the basis of the High Court decision in Soblusky v Egan 32 it was argued that the respondent was vicariously liable.
The majority of the High Court dismissed the appeal and restricted the decision in Soblusky v Egan to its own facts. The majority distinguished Soblusky v Egan on the basis that the respondent had not retained the direction and control of the plane. Nevertheless, a number of the judgements contained thinly veiled criticism of the decision in Soblusky v Egan. These judges clearly intimated that they did not find Soblusky v Egan to be a satisfactory decision and that they would not extend it. For example, Gummow J stated that the decision in Soblusky v Egan 'rests upon insecure and unsatisfactory foundations in principle' 33 and Hayne J stated that 'if the decision in Soblusky v Egan is still good law' 34 he would not extend it beyond motor vehicles.
McHugh J dissented in Scott v Davis and would have allowed recovery, albeit not on the basis of the reasoning in Soblusky v Egan. His Honour's decision will be considered below.
The upshot of these cases is that the resort to agency principles to impose liability is tenuous at best. It is submitted that, if liability is to be imposed, for policy or any other reason, it would be best to formulate a sui generis principle for liability of drivers of motor vehicles, rather than attempting to rely on agency principles. Despite remarks to the contrary in Soblusky v Egan, it is difficult, in accordance with principle, to have liability imposed in this context on the basis of agency and not, in consequence, fracture the law governing the interaction of vicarious liability and agency in other contexts. There are two main problems. Firstly, the driver does not satisfy the legal definition of an agent. At the least, it is difficult to suggest that the driver is a fiduciary or that the driver can affect legal relations (except tortiously, as noted by Fridman not the province of the common law to make exceptions for 'car cases'. Such an artificial demarcation should be left to the legislature. 2

Agent Representing a Principal in Transactions with Third Parties
Another generally accepted line of cases involving vicarious liability are those where an agent represents a principal in a transaction with a third party.
According to Lindgren J in NMFM Property Pty Ltd v Citibank Ltd: The facts of Colonial are straightforward. Ridley was employed by the appellant under an agency agreement to obtain assurance business. Despite an express prohibition in his contract which provided that Ridley should not bring the character of any person or institution into disrepute, he made defamatory statements concerning the respondent, another assurance company.
The appellant argued that Ridley was an agent and in the nature of an independent contractor, rather than a servant. In a short judgement, Gavan Duffy CJ and Starke J dismissed the appeal and held that one is liable for another's tortious act 'if he expressly directs him to do it or if he employs that other person as his agent and the act complained of is within the scope of the agent's authority' 38 (emphasis added). Dixon J (with whom Rich J agreed) also dismissed the appeal. His Honour noted that there was no sufficient reason for supposing the appellant assumed such a control over the manner in which Ridley executed his work so as to constitute him a servant.
Dixon J stated the following principles: In most cases in which a tort is committed in the course of the performance of work for the benefit of another person, he cannot be vicariously responsible if the actual tortfeasor is not his servant and he has not directly authorized the doing of the act which amounts to a tort. The work, although done at his request and for his benefit, is considered as the independent function of the person who undertakes it, and not as something which the person obtaining the benefit does by his representative standing in his place and, therefore, identified with him for the purpose of liability arising in the course of its performance. The independent contractor carries out his work, not as a representative but as a principal. But a difficulty arises when the function entrusted is that of representing the person who requests its performance in a transaction with others, so that the very 36 service to be performed consists in standing in his place and assuming to act in his right and not in an independent capacity. 39 Dixon J held that Ridley did not act independently, but as a representative of the appellant. Contrary to the position of the other Justices in the majority, Dixon J considered that there was no case which 'distinctly decides that a principal is liable generally for wrongful acts which he did not directly authorize, committed in the course of carrying out his agency by an agent who is not the principal's servant…' 40 Dixon J held that the undertaking contained in Ridley's contract not to disparage other institutions was not a limitation of his authority but a promise as to the manner of its exercise. In those circumstances, he held that it was not an extension of principle to hold the appellant liable. Therefore, the key to liability in Dixon J's view was that the 'agent' acted in a representative capacity.
Fleming treats a situation such as that which arose in Colonial as an instance of personal liability imposed on the principal, rather than vicarious liability. This suggestion was doubted by two members of the Full Court of the South Australian Supreme Court in Davis v Scott. 41 The issue as to the distinction between vicarious liability and personal liability will be discussed in more detail below.
In the context of the Colonial decision, Atiyah notes three possible bases of liability. Firstly, liability for agents' representations. Secondly, liability for all statements made by agents, as in Colonial. Thirdly, liability for physical conduct constituting a tort committed in the course of representing the principal in transactions with third parties.
He states that at this point the courts would 'draw the line'. He posits that it would be: inconceivable that a vendor would be held vicariously liable for the negligence of an estate agent who negligently injured a prospective purchaser while driving him to view the vendor's house, although it is hard to put one's finger on a good reason for this distinction. 42 Whilst the facts of Colonial involved a tortious statement, it is not clear that the High Court necessarily limited its judgement to this. It may be that, in an appropriate case, liability in this context could be imposed for more than negligent or defamatory statements. Nevertheless, if Atiyah is correct then again the resort to agency principles is anomalous as it is difficult to justify from a strictly legal perspective having liability only for statements, and not physical conduct. Whilst such a result may be of practical utility, it does little for consistency. Indeed, it will be seen below that McHugh J's approach extends liability beyond statements to cover tortious acts and omissions as well. Further, but in the context of a typical employer and employee relationship establishing vicarious liability rather than an agency context, it is interesting to note that in This is obviously broader than liability for merely negligent acts or omissions and suggests a wider approach to the scope of the principles of vicarious liability.

III PRINCIPAL/AGENT GENERALLY
Whilst the above categories show that the courts are willing to find vicarious liability in various manifestations of the principal/agent relationship, we have not yet considered the position of the principal and agent relationship more generally. As noted above, according to many commentators, both of the specific categories considered above are not true instances of vicarious liability in the agency context at all. Rather, the agency tag has been used as a mere convenience, although not strictly juristically correct.
Indeed, the law in this area has been described as a 'tortuous compromise'. 45 This is because the law is unclear as to the interplay between agency concepts and vicarious liability.

A Analysis Through the Prism of Tort Law
A review of the academic authorities in tort law reveals that there is much disagreement as to whether the principal and agent relationship will support a finding of vicarious liability. Whilst all recognise that the master/servant relationship supports vicarious liability, this is about the only common ground.
Distinguished commentators such as Fleming, Balkin and Davis contend that agency has no place in the context of vicarious liability 46 and that vicarious liability should be limited to the master/servant relationship. 47 The result of this approach is that the only analysis that needs to be undertaken is whether or not the person in question is an employee or an independent contractor. Concepts of agency are best left to the contractual context. Whilst the former approach has the benefit of simplicity, it is arguably inconsistent with many cases which have held principals vicariously liable for the acts of their agent. One possible rationalisation for this approach is the proposition that whilst a principal will not be vicariously liable for the tortious acts of an agent, the principal will nevertheless be liable as principal for the tortious acts of his or her agent in the course of the agency. In this analysis, the first question would be, is the person an employee or an independent contractor? If the person is an employee, then it would be necessary to determine the difference in scope, if any, between the master's liability and the principal's. If the person is an independent contractor, it would be necessary to determine if the independent contractor was an agent for the purposes of agency law, and then determine if his or her actions occurred during the course of the agency such as to render the principal liable. The difference between this analysis and Trindade's, is the fundamental nature of a principal's liability.
The difficulty with the approach of Gardiner and McGlone, Trindade and Cane, and Luntz and Hambly is that in practice, it would often be exceedingly difficult to distinguish between an independent contractor and an agent. Also, courts have expressed reluctance to find liability for the acts of an independent contractor even though that contractor could also be considered an agent. For example, in Gaitanis v Nicholas Moss Pty Ltd 52 Phillips JA 53 in the Victorian Court of Appeal doubted that a principal could be vicariously liable for the acts of his property agent where Phillips JA considered that the agent was also an independent contractor. Further, many of the authors recognise that the term 'agent' in the tortious context has a wider meaning than that generally ascribed to the term, such that it obviates the need to enquire into the person's status as an employee or an independent contractor. 54 It must be queried whether such approach is permissible, particularly as the courts have not adverted to this distinction.

B Analysis Through the Prism of Agency Law
A review of the academic authorities in agency law reveals that there is much disagreement as to the circumstances in which a principal will be liable for the acts of his or her agent.
A fundamental element of confusion that arises is whether a principal (who is not a master) is generally liable for the torts of his or her agent (who is not an servant), on the basis of vicarious liability or otherwise. The broadest view is taken by a number of leading commentators who state that a principal may be liable for the tortious acts of his or her agent per se.
In the leading Australian text, Dal Pont suggests that the 'basic rule' is that a principal is jointly and severally liable with his or her agent for any tortious act or omission committed by the agent whilst the agent is acting within the scope of his or her actual or ostensible authority. 55  A middle view is taken by Markesinis and Munday, who suggest that 'the principal is normally considered liable for any torts involving representations which the agent may have committed whilst acting within his actual or ostensible authority'. 58 This analysis reflects the contractual basis of the agency relationship.
The narrowest view of a principal's liability is that taken by the authors of Bowstead and Reynolds on Agency, arguably the leading agency text book in print. 59 The learned authors do not promulgate the tests considered above. Whilst they acknowledge that the principal is liable for tortious conduct of the agent when he or she was acting in the course of his or her employment, they limit this to cases where the agent is also a servant. Where the agent is not also a servant, liability will only ensue: • when the principal authorised or ratified the tortious conduct; • when the principal has breached a non-delegable duty of care; • (perhaps) in the case of a statement made in the course of representing the principal, made within the actual or apparent authority of the agent.
Although the authors acknowledge the existence of the 'car cases', they submit that those cases can not be rationalised on the basis of agency principles, for the reasons noted above.
Whilst the above effectively surveys all of the modern analysis on this issue, it is also instructive to consider the views of two classic legal scholars, Stoljar and Atiyah. Stoljar takes the middle view. He states that 'when it is stated that a principal is liable for his agent's torts, this is true of a tort such as fraud, but not true of trespass or negligence, ie wrongs relating to physical service, not of contractual agency'. 60 Atiyah notes that: There is no more controverted proposition than that a principal is generally liable for the torts of an agent committed within the scope of his authority. While there can be no doubt that a principal is in some circumstances liable for the torts of a person who is not a servant it is still a question of the greatest difficulty whether these cases are illustrations of a general rule or whether they remain isolated cases… 61 What then, is to be made of all of the above? The main point to note is that the authors advocating the broadest basis of liability for acts of agents, do not do so on the basis of vicarious liability. On their analysis, the principal is made liable not vicariously, but as principal. As noted above, this could be consistent with the position taken by Fleming, Balkin and Davis in that, as those authors suggest, there is no need to resort to vicarious liability at all. However, this analysis raises questions as to the true theoretical nature of vicarious liability.
Dal Pont distinguishes between a direct liability of the principal for the acts of the agent and the vicarious liability for a master for the tort of the servant. 63 He laments that: [T]hough in an agency relationship, the principal is liable directly as principal as opposed to vicariously, this distinction has been treated as of little practical significance by the case law, being evident from judges' reference to principals as vicariously liable for their agents' acts.

(emphasis in original)
Dal Pont clearly uses the theoretical situation in the contract context and applies it to the tortious context. In other words, in strict agency/contract theory, although the agent does the physical act of entering into the contract, the principal does not become bound to the contract vicariously, but rather as principal, as if he or she had entered into the contract himself or herself.
However, again, there is confusion as to whether or not Dal Pont's take on the legal theory is correct. In a number of cases it has been held that vicarious liability is a form of direct liability. 65 Furthermore, it would appear that these cases can not be explained away on the basis of sloppy use of terminology. For example, in McConnell Dowell Constructors v EPA 66 Spigelman CJ (with whom Grove and Kirby JJ agreed) stated that 'vicarious liability is a direct liability. The acts of the servant or agent are taken to be the acts of the employer, or, in this case, head contractor'. 67 Dal Pont equates the distinction between direct and vicarious liability with the distinction between liability of a principal for an agent's torts and liability of a master for a servant's torts. Whilst Dal Pont recognises that there are 'clear parallels' between tortious vicarious liability and tortious liability in the agency context, he suggests that the legal theory behind each form of liability is distinct.
Resolution of the true nature of vicarious liability is a difficult task and may depend upon its interplay with agency law. If agency law is considered principally relevant in the contractual context then it is difficult to disagree with suggestions that a principal will not generally be liable for the tortious acts of his or her agent, except perhaps where the tortious conduct relates to entry into the contract. Vicarious liability would fit comfortably within this framework and this is in line with the middle view noted above and the view of Stoljar. If, however, agency law is considered to play a broader role, then there should be no need to resort to vicarious liability principles as liability will be imposed on the basis of inherent agency principles, not imported tortious principles. If there was to be overlap between an employment relationship and an agency relationship, either basis of liability would be available unless of course the scope of liability is different. The issue of the scope of liability will be considered below.
Given the demonstrated difficulties in reconciling liability in the tortious vicarious sense and the general agency sense, it is perhaps of relevance to consider the fundamental factors underlying liability with respect to each.

IV RATIONALE FOR IMPOSITION OF LIABILITY A Vicarious Liability in the Master/Servant Context
It appears generally accepted that there is no sound theoretical/legal basis for the imposition of vicarious liability, even in the master/servant context. In Hollis v Vabu, all members of the High Court (with the exception of Callinan J who did not address the point) acknowledged that there was no sound legal rationale for the imposition of vicarious liability, but rather, it was grounded in various policy considerations. 68 Given then that vicarious liability is not founded upon fundamental legal doctrine, it may be that it is appropriate for courts to fashion it so that it applies in the agency context, but the difficulty will be whether it applies generally, or only in specific circumstances. Before addressing this issue, it is necessary to consider exactly what are the policy considerations behind vicarious liability.
The Queensland Law Reform Commission has, in a recent review of the law relating to vicarious liability, identified a number of policy reasons for the imposition of vicarious liability in the master/servant context. The Commission identified the following main factors: (a) The plaintiff can obtain compensation from someone who is financially capable of satisfying a judgement. It is likely that an employer will have greater financial resources than an employee. (b) A person or corporation who employs others to advance their own economic interest should, in fairness, be placed under a corresponding liability for losses incurred in the course of the enterprise. (c) Vicarious liability promotes a wide distribution of tort losses as an employer can pass the costs on through insurance and higher prices. (d) The imposition of vicarious liability promotes deterrence of tortious conduct. It provides incentives for employers to encourage employees to perform well on the job and to discipline those that do not. 69 It was not doubted that the availability of insurance also played a significant role in the development of the law relating to vicarious liability. 70 68 Hollis v Vabu (2001)  It may be that many of these factors apply to the paradigm agency situation. For example, where an agent is employed to conduct business on behalf of a principal then all of the above factors would probably apply, as in Colonial. However, where the agency concept is extended, for example, in the car cases, it is difficult to justify finding vicarious liability on the basis of the above policy considerations (although of course, in the car cases at least, the availability of insurance would no doubt be an important factor). Further, in the sense that solicitors (and various other professionals) are considered to be agents, often the solicitor would have greater resources than the principal. Therefore, it can be seen that the fundamental problem to applying vicarious liability generally in the agency context is the broad range of persons who can fall within the umbrella of the term 'agent'. It goes without saying that most employers will have more resources than their employees. It is certainly not always the case with agents.

B Liability in the Agency Context
On the assumption that there is indeed a general liability on the part of principals for the tortious acts of their agents, it is necessary to compare the rationale for this type of liability against the rationale for vicarious liability.
In the agency context, Dal Pont has identified three reasons underlying the imposition of liability on a principal for the tortious acts of his or her agent. 71 First, the principal selects the agent and has better means of ascertaining the quality, strengths and weaknesses of the person. 72 Secondly, as the principal has delegated the performance of a certain class of acts to the agent, it is not unjust that the principal, who will derive the benefit of the agent's efforts, should bear the risk of the agent exceeding his or her authority. 73 Thirdly, the principal has given the agent general authority to commit the wrongs. 74 According to Dal Pont's theory, none of the rationales for imposing vicarious liability in the master and servant context necessarily applies to the agency context. In particular, the concept of 'deep pockets' does not underpin the principal's liability and nor does the wide distribution of losses. Yet it could easily be argued that, in some agency relationships at least, the reasons for imposing vicarious liability in the employment context apply equally to imposing liability (vicarious or not) in the agency context, and vice versa. But again, the broad range of agency relationships means that it is difficult to identify universal underlying principles, particularly with reference to 'deep pockets'.

V SCOPE OF LIABILITY
On the assumption that there is a general principle of liability of principals for the tortious acts of their agents, it is necessary to consider what, if any, are the differences in scope between a master's liability and a principal's liability. As noted above, a master will be liable for the torts of its employees committed in the course of their employment. A principal will be liable for the torts of his or her agent committed whilst the agent is acting within the scope of his or her authority.
Given the potential overlap between an employment and an agency relationship, if either has a wider scope of liability then it could be advantageous in appropriate circumstances to seek to classify the relationship as one type of relationship rather than the other. Alternatively, it could be appropriate to seek to fall within the second, broader category even though the relationship fits more comfortably within the category with the narrower scope of liability.
However, most (if not all) commentators who address the issue suggest that there is little or no practical distinction in the scope of liability. Dal Pont submits that there is 'a convergence' in this context between agency and the law of vicarious liability. 75 Swanton suggests that the scope of vicarious liability and agency liability are synonymous. 76 Atiyah notes that the prevalent view is that the phrases mean the same thing. 77 It is suggested that the above commentators are correct. It would be difficult to conceive of a situation where an action occurred within the scope of an agent's authority, but not (if he or she was an employee) within the scope of his or her employment. If there are differences in scope, then they are slight. It is suggested that it would be futile to force a particular relationship into one or other of the categories for the perceived benefit of obtaining a greater scope for liability.

VI THE APPROACH OF MCHUGH J
Now that we have examined the current (and confused) state of the law in this area, it is necessary to consider the novel approach that McHugh J has taken. As noted in the introduction, McHugh J in both Hollis v Vabu and Scott v Davis adopted a very different method to the other members of the High Court in reaching his conclusions.

A Scott v Davis
The first case in which McHugh J propounded his unique theory as to agents' vicarious liability arose in Scott v Davis. The facts of this case have been noted above. McHugh found that, on the facts: A principal is … liable for the wrongful acts of an agent where the agent is performing a task which the principal has agreed to perform or a duty which the principal is obliged to perform and the principal has delegated that task or duty to the agent, provided that the agent is not an independent contractor. The principal is also liable for the wrongful acts of a person who is acting on the principal's behalf as a representative and not as an independent principal and within the scope of the authority conferred by the principal. 79 He further states that the categories of liability for the acts of agents are no more closed than the categories of negligence. The second of McHugh J's bases of liability will be discussed in the context of Hollis v Vabu below.
The obvious difficulty with the first formulation is how to distinguish between an agent and an independent contractor. It is submitted that, in practice, it will be extremely difficult to establish that a person is an agent, but not an independent contractor, given the significant overlap between the two relationships. Numerous eminent commentators have pointed out the difficulties inherent in attempting to consistently distinguish between the two relationships. 80 The second query that could be raised with McHugh J's formulation is whether it is necessary, in order to make the principal liable, that the principle was in fact under an agreement or obligation himself or herself (presumably as an agent) to perform the act? In other words, McHugh J specifies that the agent is performing a task that the principal has agreed to perform or is obliged to perform. Why, it is submitted, should the alleged 'principal' in Scott v Davis be found liable if the boy's father asked the principal if his son could be taken on a flight, but the principal not be liable if he simply offered the boy a flight and suggested that the deceased pilot do the flying? This seems an arbitrary distinction.
McHugh J was at pains to point out that: Holding a principal liable for the action of its agent occurring within the scope of authority while acting as the principal's representative is not inconsistent with the long standing rule that a person is not generally liable for the negligence of an independent contractor. To suggest that it is ignores the elements of representation, the authority of the principal to state in detail the parameters of the agent's authority and to control the conduct of the agent in so far as there is scope for it. 81 With respect, these justifications do not appear convincing. First, the element of representation only applies to the first of McHugh J's tests, as per Colonial. It does not apply to a situation such as arose in Scott v Davis. Secondly, a person who engages an independent contractor can delimit the scope of the independent contractor's work and does have an ability to control, to a certain extent, the manner in which the contractors performs his or her task. Therefore, this arguably cuts through the long-standing distinction between liability for employees and non liability for contractors.
McHugh J also stated that imposing liability on the principal is likely to assist in avoiding tortious acts by giving the principal incentive to exercise care in selecting, training and defining the scope of authority of agents. right of indemnity against the agent, and if the agent can not satisfy that indemnity, it is fairer that the person who engaged the agent is made liable rather than the innocent third party. Thirdly, the agent has power to bind the principal in contract so it is not outrageous for the principal to be bound by the agent's tortious acts. Fourthly, there is no indeterminate liability that might arise. 82 According to McHugh J, 'the "motor car" cases are not the product of unprincipled, social engineering on the part of the common law judges '. 83 Whatever may be the appropriateness of McHugh J's reasoning in the broader agency context, it is difficult to justify imposing liability on the alleged principal in Scott v Davis, even on McHugh J's reasoning. It is artificial to suggest that there was incentive on the alleged principal to appropriately select, train and define the scope of authority of the pilot. Secondly, it is unlikely that a court would have found that an agency relationship existed in order that the principal could seek indemnity from the alleged agent. In other words, whilst it is convenient to find an agency relationship for the purposes of vicarious liability, it is suggested that it is more difficult to establish an agency relationship when attempting to obtain an indemnity from the alleged agent. In these circumstances, it is submitted that courts would be more astute in determining whether, in truth, a relationship of principal and agent existed. Similarly with McHugh J's third reason, it is submitted that the 'agent' had no power whatsoever to bind the 'principal' in contract in the circumstances of the case, therefore it is difficult to hold the 'principal' liable in tort. This is an example of where McHugh J confuses the breadth of the concept of agency.
It is submitted that McHugh J's conclusions on liability in Scott v Davis are made without proper foundation. The social context and the nature of the engagement suggest that the only person who should have been liable for the pilot's negligence was the pilot himself. Introducing agency concepts and vicarious liability appear artificial to say the least.

B Hollis v Vabu
The facts of this case have been noted above. McHugh J found that the courier was neither an independent contractor, in the sense of a person acting as an independent principal, nor was he an employee. McHugh J stated that, rather than attempt to force new types of work arrangements into the employee/independent contractor dichotomy, it was better to develop the principles concerning vicarious liability in a way that gives effect to modern social conditions.
His Honour held the respondent liable because he found that the courier was an agent of the respondent -but not an independent contractor -and was acting as the respondent's representative in carrying out a contractual obligation of the respondent. 84 As McHugh J also considered that policy considerations dictated that the respondent should be made liable. After reviewing the Canadian Supreme Court case of Bazley v Currey, 85 which had dealt with an employer's liability for sexual abuse by its staff, McHugh J held that the respondent was the only effective entity to compensate the victim and that it was 'fair' to make the respondent pay. The notion of fairness was analogous to fairness in the employer vicarious liability context. Namely, the couriers were controlled by the respondent and they were acting for the economic benefit of the respondent.
His Honour noted that not only does the doctrine of vicarious liability have its basis in policy considerations, but common law courts acknowledge that the evolution of the doctrine continues to be guided by policy. 86 He held that policy considerations demanded that Vabu be held liable. First, it provided Hollis with an effective means of compensation given that the courier rider was unidentified and also that, even if he had been identified, it is likely that he would not have had means to satisfy any judgement. Secondly, it was fair to make Vabu liable as the courier was acting for the economic benefit of Vabu. Thirdly, the imposition of liability was likely to deter similar future incidents because Vabu would increase its supervision of the couriers.

McHugh J held that:
It is true that the couriers employed by Vabu are neither employees nor independent contractors in the strict sense. But there is no reason in policy for upholding the strict classification of employees and non-employees in the law of vicarious liability and depriving Mr Hollis of compensation. Rather than expanding the definition of employee or accepting the employee/independent contractor dichotomy, the preferable course is to hold that employers can be vicariously liable for the tortious conduct of agents who are neither employees nor independent contractors. 87 After reviewing Colonial, McHugh J stated that: [A] principal is liable for the wrongful act of an agent causing damage to a third party when that act occurred while the agent was carrying out some activity as the principal's authorised representative in a dealing with a third party. Nearly thirty years ago Professor Atiyah marshalled the arguments which would justify imposing liability on employers for the acts of independent contractors as well as employees. Those arguments seem as convincing to me today as they did when his work was first published in 1967. However, counsel for the plaintiff did not invite the court to re-examine the basis of the liability of an employer for the acts of an independent contractor. The question whether the common law should continue to draw a distinction between liability for the acts of employees and those of independent contractors must wait for another day. 91 The High Court has yet to consider the issue. However, given the long history of the distinction between liability for servants and liability for independent contractors, it may be a while before a litigant has the courage and the funding to follow McHugh J's invitation. Nevertheless, McHugh J's decisions in Hollis v Vabu and Scott v Davis effectively sidestepped these issues altogether. Rather than reformulate the law relating to independent contractors, he extended the law relating to vicarious liability in the context of agency law.
One positive aspect of McHugh J's reasoning is that he is being overt in his extensions of the current law. Therefore, it is difficult to criticise his reasoning on the basis of any extension as such. Rather, the more pertinent issue in an analysis of his Honour's reasoning is the question of whether it is sound from an agency perspective and whether it unduly impinges upon settled agency principles. greater number of circumstances where an agent, who is not an independent contractor, performs a task that the principal had agreed to perform. The problem with this is that, as noted above, these categories of 'established' vicarious liability have been criticised on a number of bases, in particular, their underlying theoretical rationale.
For example, given Dal Pont's view on the general liability of principals for the torts of their agents, it is not surprising that he ponders exactly why McHugh J needed to resort to the concept of vicarious liability at all. On Dal Pont's view, if an agency relationship was established, then liability would flow if the conduct was within the scope of the agent's actual or ostensible authority or within the course of the agency. 93 There would be no need to rely on concepts of vicarious liability. One difference that would arise is that Dal Pont would not necessarily consider that a concurrent finding that the agent was an independent contractor would be fatal to the claim, whereas for McHugh such a finding would mean that the claim must fail.
Given that, as noted above, vicarious liability is based upon policy considerations, and not strict legal foundations, it is again difficult to criticise McHugh J's approach solely on the basis of his policy justifications. Further, he is not entirely alone in his bid to use the law of vicarious liability flexibly to adapt to novel situations.
In the recent case of S v Attorney-General 94 the New Zealand Court of Appeal dealt with a situation where a claim for vicarious liability was made against the relevant government department for the actions of foster parents who had abused the plaintiff. The majority 95 found the Department vicariously liable on the basis of agency, 'albeit that their agency was of an unusual, indeed unique, nature'. 96 Tipping J concurred in the decision and the majority judgement, however, his Honour did not agree on the basis on which vicarious liability was imposed. Rather than stretching the concept of agency, his Honour considered that this was a novel situation and that this novel category attracted the principles of vicarious liability. By analogy with the existing categories of master/servant and principal/agent, his Honour found that the foster care and Department of State relationship in the present case could found a vicarious relationship as well.
Therefore, the criticism is not for McHugh J's use of policy considerations per se, but rather whether those policy considerations were justified. As noted above, this does not always appear to be the case.

VIII CONCLUSION
There are two fundamental problems in this area, as noted by separate High Court judges in Scott v Davis. The first is, as Gleeson CJ notes, 'the protean nature of the concept of agency, which bedevils this area of discourse'. 97 The second 'legitimate concern' is, as Gummow J notes, the 'the place of agency in the law of torts, particularly as a legitimate source of vicarious liability'. The use of vicarious liability is a convenient tool when courts wish to shift the burden of liability from one party to another. However, the challenge is to use that tool in line with tortious principle and without fracturing settled principles in other areas of the law. The breadth of the agency concept means that agency is perhaps an inappropriate vehicle by which to impose vicarious liability. If necessary, resort should be had to inherent agency principles, not imported tortious principles of liability. Nevertheless, and as shown above, there appears to be fundamental confusion as to the true nature of tortious liability in the agency context. Until this is resolved, it may be difficult, if not impossible, to attain consistency in this area.
Whilst McHugh J has adopted a unique approach to the resolution of these issues, he does so without an adequate examination of the fundamental principles involved and arguably causes greater confusion in this area of the law. As such, his Honour's approach to this topic should be approached by subsequent judges with great circumspection.