University of New South Wales Faculty of Law Research Series
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
University of New South Wales Faculty of Law Research Series |
|
Last Updated: 3 November 2010
Straddling the Public/Private Divide: Tortious Liability of Public Authorities[1]
Professor P Vines[2]
Citation
This article was published in the Judicial Review (2010) 9, 445-475.
Abstract
This paper examines the difficult issues associated with imposing tortious liability on public authorities at common law, focussing her discussion on negligence. It considers the public and private nature of these bodies and their functions, first examining the possible approaches to treatment of these at common law, and then following the enactment of the civil liability legislation in the various States and Territories. A preferred approach to the determination of tortious liability of public authorities is then offered, which, provided a principled approach is taken to statutory interpretation, will accommodate the civil liability legislation.
Introduction
The tortious liability of public authorities has been problematic for the last hundred years or more. Until 2002 in Australia we managed this liability through the common law. It was not easy and it was very messy, particularly when negligence was in issue. In this paper I try to set out where the messiness really bites, what has created the problems in the common law approach to the issue, and then I turn to look at the legislative response to this. When looking at the legislative response it is useful in turn to consider how judges might approach the interpretation of this legislation and some of the pitfalls it creates.
In this paper I consider “public authorities”. This is shorthand for public bodies of all kinds. In most cases these are also statutory authorities. The tortious liability I consider is negligence, with a few exceptions.
I argue that the determination of the liability of public authorities for negligence is unnecessarily complicated and that some of the distinctions which are used are untenable. I further argue that a more conceptual approach to statutory interpretation may help to solve the problem and that although the civil liability Acts are unprincipled in their approach to this question, a more serious approach to their statutory interpretation may be useful in relation to this part of the Acts as well. Due to time constraints and the problems created by the differences across Australian jurisdictions, I will focus largely on New South Wales.[3]
Case one
A man connects the hose of the exhaust pipe of his car to the front window and shuts himself in. Two police officers see this and stop and talk to him. After some time he assures them that he has changed his mind about committing suicide and will go home and talk to his wife. The police have power under mental health legislation[4] to detain a person they suspect of being a danger to himself or herself. They do not do this. The man goes home and kills himself in private. His wife sues the police. The High Court of Australia says they owe the man and his wife no duty of care.[5]
Case two
A probation officer who has a statutory duty to supervise parolees fails to closely supervise a parolee despite knowledge that he is a violent offender and that there is a recommendation that he should be closely supervised. The probation officer misses several appointments with the parolee and fails to intervene to prevent his placement for work experience at a liquor provider despite his known problems with alcohol. He carries out a robbery and kills three people and badly injures another. The injured woman sues the parole authority. The Supreme Court of New Zealand says the parole authority could owe a duty of care to the injured woman,[6] overturning the decision of the Court of Appeal which struck out the claim.
Both these cases concern personal injury where the defendant is a public authority — the police and the parole authority. Suppose the defendant had been a private individual. In the first case, the outcome might have been the same because there is no duty to rescue; although that would mean ignoring the existence of the mental health legislation and the fact that the police began a course of conduct which might look like an assumption of responsibility. In the second case, the outcome might have been different because the injured woman was not in any special class of people likely to be hurt by the parolee. In fact, the majority in the second case held that it should be treated as if it were private law and decided the duty did arise; but the minority (who also thought there should be a duty) thought that the public nature of the parole authority was part of the reason why a duty of care should be owed.
Are these outcomes correct? Are they correct or incorrect for the right reasons?
Why is the tortious liability of public authorities so difficult?
The tortious liability of public authorities is difficult for four major reasons:
The public/private divide
Trindade, Cane and Lunney[7] put it nicely:
“Tort law’s individualism comes under great pressure in cases
where the alleged tortfeasor is an entity whose prime responsibility
is to
protect societal (or ‘public’) interests rather than its own
interest or the partisan interests of any particular
individual(s) or group(s)
within
society.”[8]
The
basic paradox at the heart of tortious liability of public authorities is that
it straddles the public/private divide. Where a
public authority makes an
administrative error we are trained to think the obvious legal remedy is
administrative law: a public law
problem, and we do not think of administrative
law as offering compensatory remedies. We regard them as the domain of private
law.
On the other hand, when one party injures another we think of that as a
tort problem for which compensation should (usually) be available
(that is, a
private law problem). In Precision Products (NSW) Ltd v Hawkesbury City
Council[9] Allsop P
addressed whether there should be civil liability for an invalid notice issued
by a council officer which had caused pure
economic loss. What really troubled
Allsop P was that this created a:
“... lack of coherence between administrative law doctrines and the imposition of monetary compensation for the flawed or failed exercise of governmental power ... What tends to strike at the coherence of administrative law here is the positing of a duty to exercise reasonable care not to make a flawed decision by, for instance, failing to give procedural fairness or failing to confine the power within statutory limits. Such a duty, as contended for here, would tend to open public authorities to the spectre of compensation for flawed decision making, in circumstances where the validity of the exercise of power can be tested and resolved by judicial review, and where standards of competence and skill are well able to be dealt with by an appropriate regime of governmental administration.”[10]
It is worth noting that Allsop P is talking about a “lack of coherence” and the “spectre of compensation for flawed decision making”. His point is that the scope of judicial review and its intensity (its grounds, if you like) have expanded exponentially. There are some good reasons for this that have nothing to do with compensation; indeed judicial review by itself does nothing to compensate the harm done to that initial person affected by the decision. The argument about coherence is a taxonomic argument rather than an argument about why compensation should not be given. The High Court’s emphasis on coherence has give us the Sullivan v Moody[11] incompatibility test and changed some aspects of the law of negligence;[12] but in my view any emphasis on taxonomy in the common law system is essentially misconceived. It therefore does not significantly advance the argument about compensation for wrongs done by public authorities or in the exercise of public functions. He also talks here of the “failed exercise of governmental power”, but with respect, the term “governmental power” here seems too broad to be helpful with the basic problem of when public authorities’ wrongs should be compensated.
When a public authority injures a person we recognise that some amalgam of public and private law is at issue. The problem is that exactly how it is to be straddled is a constant headache about how to justify a departure from the Diceyan[13] view that everyone, including public officials, should be subject to the same rules. Why should some parties get special protection? For example, Trindade, Cane and Lunney’s view that public authorities are distinguishable from private bodies because they are required to put the interests of society first (or whatever interest their empowering statute stipulates) might lead to the conclusion that they should therefore have less asked of them and thus be protected from liability. Similarly it is argued that public authorities necessarily have to do things which may harm some people while benefiting society as a whole, and that in doing this they must therefore be given more leeway when they harm someone than a private individual would be. For example, where a road authority demolishes homes to build roads (the statutory authority to do so is never read as exonerating reasonably avoidable harm caused by carelessness). However, it is also arguable that because public authorities are supposed to be other-interested rather than self-interested (for example, the police), they could be asked to do more than a private individual.
The rule of law
When Crown immunity was abolished it was on the Diceyan basis that as an element of the rule of law, public officials should be treated like private parties and the Crown should be vicariously liable for its employees.[14] Two things to note about Dicey in this context are that he was writing before the modern tort of negligence existed, so he was more likely to think of assault or battery as the relevant tort; and that he was writing in a context which regarded administrative law as a separate body of law with some disfavour[15] despite the fact that judicial review was extremely common. Some torts, such as misfeasance in public office, did allow for compensatory actions of public officials short of suing the Crown, but those actions made public officials personally liable and did so on the basis of malice — surely a very private emotion!
Separation of powers
In the context of the tort of negligence, public authorities continue to give judges pause because they have both obligations and resources that private individuals don’t. They are also different from private individuals because they are often created by statute, which raises the question of whether it is permissible for judges to interfere with a statutory body or whether that would be a breach of the separation of powers.
In 1982, Aronson and Whitmore wrote:
“There are many situations in which the fact that the defendant is a public body or official is unimportant. But there are also cases in which the public character of the defendant is crucial ... The Treasurer cannot be held liable in negligence to the businessmen he injures as a result of his economic mismanagement. A municipality should not be held liable in negligence for promulgating a town planning scheme which has the effect of rendering some businesses uneconomic. The reasons for conferring immunity from suit will be different in each case ...”.[16]
The points they make are important. However, this is not just about conferring immunity but also about other ways of making determinations which might affect public authorities’ liability, including deciding there is no duty of care, or deciding that there is no breach for a reason which includes the public nature of the defendant. Causation is unlikely to be determined on the public nature of the defendant so this paper will concentrate on the questions of immunity, duty and breach.
The concept of “public”
In attempting to straddle the public/private division, this area of law strains a whole series of tensions and definitions. What is public? Should the public party be treated differently from private parties, and if so, when, and why? If we are going to accept that public authorities may be sued for some things and not others, how will the line be drawn? How are we to deal with the statutory entanglements of public authorities?
I have said that the liability of public authorities is problematic because it straddles the public/private divide. When we talk about public and private law we distinguish them in these ways:
However, public authorities are not the State, although they may include its emanations. What is public about them? Firstly, they are creations of the State. If a State creates a car, (consider the volkswagen) is it public? There are many ways to define public authorities, for example, public authorities are entities created for public purposes. That is to say, many of them would not exist except for public purposes. Thus, an authority such as the NSW Roads and Traffic Authority (RTA) is created for the public purpose of overseeing road use and development.
So what is public? The first puzzle concerns what we see as a public authority and what we see as a public function. There are real difficulties in determining this in the modern regulatory State. The Statutes of Diet and Apparel of the 1360s have nothing on the reach of the modern regulatory State. As Harlow says:
“To the modern tort lawyers, the emphasis on personal liability is distinctly old-fashioned. Already, writing between the two world wars, Jennings was critical of Dicey for glossing over the growing volume of statutory powers vested in public officials. Today the process has gone much further; it is no longer public servants but public services and public authorities that we want to hold accountable and whose funds we want to access. I cannot too often reiterate the extent to which the functions of the state have changed. The state has welfare functions. It is interventionist. There has been an enormous growth in supervisory and regulatory powers. This steady expansion of state power has come together with the trend I have been describing to replace personal liability by an impersonal set of non-delegable obligations vested in systems, corporations and institutions.”[17]
It may be argued that part of the characteristic of public purposes is its unbalanced nature in relation to individuals. Compare this with Weinrib’s view of private law as essentially dyadic or bilateral.[18] Public authorities are necessarily set up to deal with multiples; multiple individuals with competing interests not all of which can be met. They therefore will often have an allocative function. When a public authority looks much more like a private entity — that is, more dyadic — the law tends to deal with it as if it is a private individual. The Law Commission for England and Wales in its Consultation Paper, Administrative Redress: public bodies and the Citizen,[19] defines “truly public” as things involving a prerogative power or a “special statutory duty or power”. These are defined as “a power that allows the public body to act in a way not open to private individuals” or “a duty placed on the public body that is specific to it and is not placed on private individuals”.[20] This still doesn’t tell us what it is.
Public authorities will very often have to make decisions which necessarily damage one part of the public for the benefit of another part of the public.[21] For example, the RTA may have to demolish homes in order to build a road. Mark Aronson argues that in order to be “truly public” (to use the language of the Consultation Paper, not his) there must be a coercive element.[22] This seems to mean non-consensual alteration of one’s legal rights and obligations. If a coercive element is the marker of the public it would explain the definition of schools as public (because children under 16 must go to school), but it may not necessarily apply to the water board or other utilities.
Public “authorities” may also include persons exercising public functions; and thus not be corporate bodies, but real people.[23]
Harlow also points out that what is public and what is private in terms of both bodies and functions seems to be endlessly fungible[24] — roads have and can be built and maintained by private bodies (compare Sydney’s tollways); sewerage systems were built in Victorian times in cities by private entities[25] and now are maintained by public ones; fire fighters have been private as well as public; and the police have also been private as well as public. The privatisation of utilities, etc, simply reinforces the point. Is banking different when done by a nationalised bank? It seems very clear that there is a difference between the Reserve Bank of Australia and Westpac Bank, but what about when the Commonwealth Bank was nationalised? These examples point in the direction of not accepting the public/private distinction. It is absolutely clear that we cannot defensibly argue that the public/private distinction applies to the classification of defendants; the question remains whether we can meaningfully apply the public/private distinction to functions.
However, we are having it forced on us by the civil liability Acts, even as they emphasise the very point by making entities such as private schools deemed “public authorities”, and treating medical practitioners exercising their powers under the mental health legislation as public authorities as well.
The public/private distinction in relation to public authorities clearly cannot be entirely discarded, despite its problematic nature. I will come back to this.
Approaches
I would suggest that the English pattern has been to give public authorities more protection than private parties, but this has changed under the Human Rights Act 1998 (UK). The Australian pattern has been to be more likely to treat public authorities like private parties. In both cases I think the real approach is to test the authority or its function for “publicness”, and if it is “very public” to give it some protection. The civil liability Acts unambiguously aim to give public authorities a great deal more protection than private parties, and even to define public authorities as broadly as possible.
Two further approaches are to:
Attempts to solve the riddle — why they have only partially worked
Relevance of alignment with private law negligence paradigm
The paradigm private law negligence case involves two private parties. It involves an act of some kind. The act is a breach of a recognised duty, and it causes physical harm which is easy to identify. In such a case, negligence is easy to prove or disprove if the facts are there. The more one of these elements varies, the harder it becomes to establish a duty of care or liability, for example, where a case involves:
In Sutherland Shire Council v Heyman[26] the High Court rejected a public law basis for the tortious liability of public authorities. This continues to be the Australian position at common law. It is easiest to deal with this when a public authority as defendant is clearly acting in a way which a private individual could. Thus where the public authority is the employer[27] or guardian[28] of the plaintiff, or the occupier of land[29] on which the plaintiff was injured, there may be no difficulty and indeed sometimes no mention or discussion of the fact that a public authority is indeed the defendant. However, where the public authority is doing something which private individuals cannot do, the position becomes more difficult. It is in these situations that courts have often balked at using the ordinary tests for negligence without something more.
When establishing the question of whether a duty of care is owed by a public authority, the basic private law test is the starting point. One of the earliest cases concerning public authorities clearly took the approach that public authorities should be treated like private defendants.[30] The current approach to the duty of care in negligence in Australia involves establishing reasonable foreseeability of harm and a scrutiny of the salient factors in the case. The salient factors include matters such as control, vulnerability and reliance, depending on the category of case being considered. There is also a requirement that where there is a novel category of case, the imposition of a duty should not interfere with another duty, either statutory or common law, which the defendant already has; nor should it override some area of law which already covers the field.[31]
This is the basic Australian position. Where the defendant is a public authority one will begin with this approach, taking into account the fact that one of the salient factors will be the public nature of the authority or its statutory function. The duty must not be owed to all the world — such a duty is no duty at all.
The interface of common law and statute
Public authorities are often creatures of statute. The cases on public authorities’ liability can be divided according to the classification of the alleged wrong in relation to statute.
Where a public authority exercises a statutory duty it must exercise it reasonably. The question of liability in negligence may arise, as may the tort of breach of statutory duty. For the purpose of the liability in negligence, the question will be whether the authority acted like a reasonable authority in exercising that duty. For the purpose of the breach of statutory duty, the question of breach is answered by statutory interpretation. There are many cases involving a public authority exercising a statutory duty where a duty of care and the possibility of liability have been found.[32] This is relatively unproblematic. Where a public authority exercises a discretionary power it must also exercise it reasonably.[33] Where a public authority fails to exercise a statutory duty it may also be relatively easy to establish whether there is a duty or liability.[34]
However, where a public authority fails to exercise a statutory power a number of questions arise. The fact that it is a power given by the statute means that the legislature has given the power to decide whether to exercise it or not to the public authority. This in turn means that if the judiciary says the power should have been exercised, they run the risk of being seen to interfere with the legislature’s disposition. That is, the imposition of a duty of care in the form of a duty to exercise the power may be regarded as interfering impermissibly with the legislature. This question is quite separate from the question of whether the failure to exercise the power was a policy/political/quasi-legislative or any other kind of decision. That is a further question which, of course, can be asked about the exercise or non-exercise of both statutory powers and statutory duties.
Is the question of whether this is interfering impermissibly with the legislature necessarily an issue created by the fact that the authority may be a public authority? It seems to me it is not. This same question could apply to the exercise of any discretion whoever it is exercised by. The fact that there are cases which have decided that the failure to exercise a statutory power does give rise to a duty of care[35] and cases which have decided it does not[36] suggests that this is not the critical issue.
Perhaps the way to deal with the question of whether this is interfering impermissibly with the legislation is to do it through a statutory interpretation technique which might better resonate with people’s sense that particular public functionaries actually have a common law duty. This might be done by taking a higher level purposive approach to the enabling legislation.
For example, in Yuen Kun Yeu v Attorney General of Hong Kong[37] depositors took an action against the Commissioner for Banking. The plaintiffs had deposited money with a company registered by the Commissioner as a licensed deposit-taking company. The company went into liquidation causing economic loss to the depositors. It turned out that the company had been in breach of its obligations as licensee. The depositors argued that the Commissioner owed a duty of care to them in relation to his activities in registering and licensing the company. They argued that the Commissioner should have shut down the company. However, if the Commissioner did that, the Commissioner would have protected future depositors at the expense of past depositors who might well lose their deposits. How is a court to determine whether such a decision should have been made? In Yuen Kun Yeu the Privy Council said this was not justiciable so the duty was held not to exist. However, perhaps the issue could have been held over to breach and the calculus of negligence would have solved this problem. The statutory duty owed by the Commissioner was to both present and future depositors and the balance chosen may well have been the correct one. There is no need to regard this as non-justiciable; rather it requires the interpretation of the statutory duty of the body to be taken seriously at the breach stage.
Immunities and duties
One of the ways to deal with the public/private divide has been to give public authorities actual immunity. Crown immunity was an example of this. The immunity of the military for actions during wartime is another. Some immunities have proved problematic for the English courts once they came under the jurisdiction of the European Court of Human Rights. For example, in Hill v Chief Constable of West Yorkshire,[38] the House of Lords held the police were immune from a suit brought by the mother of a victim who had been murdered after the police negligently released her killer who they had been detaining for questioning. In Brooks v Commissioner of Police of the Metropolis,[39] Lord Steyn observed:
“.Since the decision in Hill there have been developments which affect the reasoning of that decision in part. In Hill the House relied on the barrister’s immunity enunciated in Rondel v Worsley [1969] 1 AC 191. That immunity no longer exists: Arthur J S Hall & Co (A Firm) v Simons [2002] 1 AC 615. More fundamentally, since the decision of the European Court of Human Rights in Z and others v United Kingdom 34 EHRR 97, para 100, it would be best for the principle in Hill to be reformulated in terms of the absence of a duty of care rather than a blanket immunity.”[40]
Thus the difference between the absence of a duty of care and a blanket immunity is significant in the United Kingdom situation where the Human Rights Act 1988 (UK) and the European Convention on Human Rights are controlling the situation. When considering the differences between immunities and the absence of the duty of care there are a number of options:
Outside the context of human rights and the European Court of Human Rights, the distinction between the first two options may be less critical. However, the articulation of an immunity creates a profound policy statement, which makes it important to consider what the approach to these cases is.[41]
What causes confusion between the first two options is the fact that decisions about whether the duty of care itself exists rely on policy decisions in themselves. This is quite explicit in the English Caparo[42] test which requires the court to establish reasonable foreseeability of the harm, proximity of relationship and whether it is fair, just and reasonable to find there is a duty in the circumstances. In Australia, the policy elements of the duty of care are not so clearly articulated, but clearly exist. For example, the incompatibility doctrine[43] is clearly a policy doctrine. It does not rest on argument by analogy with previous cases. It rests on a decision by the court that a duty of care in a novel case should not be allowed to arise where a duty might be incompatible with some other area of law. Thus we seem to have a situation where there may be a public policy decision that there is an immunity or a decision that no duty of care arises which is based, partly at least, on public policy.
The policy/operation distinction and its relatives
In Home Office v Dorset Yacht Co Ltd and Anns v Merton London Borough Council, the House of Lords introduced the policy/operation distinction into Anglo-Australian law. This was accepted in Sutherland Shire Council v Heyman, but did not gain much purchase in Australian jurisprudence after that. The Anns test for duty was to decide whether there was reasonable foreseeability of harm and if there was to hold that prima facie a duty existed, unless there was some policy reason why it should not apply. Harlow said of these two cases:
“Lords Diplock [in Dorset Yacht] and Wilberforce [in Anns] proffer as sole ‘explanation’ that ‘The defendant is a creation of public law, a statutory authority exercising statutory powers which may be discretionary in character’.”[44]
She went on to comment:
“Surely this is nothing more than a ‘special rule’ of administrative liability and one, moreover, of very recent creation. What we are seeing ... is the creation of an exception to the general principle of civil liability in favour of public authorities, based on a supposed distinction between ‘public’ and ‘private’ law.”[45]
The question of whether the judiciary should interfere with a public authority’s policies or actions has often been considered in terms of the policy/operation distinction which implicitly relies on the separation of powers doctrine in the United States Constitution.[46]
In Australia as you know, the policy/operation distinction has had its day and now has been supplanted by the quasi-legislative activity idea;[47] or possibly by the ‘political’ versus justiciable distinction.[48] The only way to make sense of these is to look at what underlies them. Every one of them is aimed at trying to work out what is “truly public”. It is a doomed task, but people keep trying.
Specific and general reliance
In Sutherland Shire Council v Heyman[49] the concepts of special and general reliance were introduced into Australian law by Mason J. Specific reliance arose where a public authority (or indeed, I would suggest, anybody) knew specifically that a person was relying on them to carry out their role — for example, certifying that a property is habitable. This is really no different from reliance which is significant in cases of pure economic loss caused by negligent misstatement or any other situation. However, general reliance was a concept which had a more “public” nature. Mason J referred to aircraft controllers as an example — no aircraft controller knows that Mr Bloggs is sitting in seat 35A in an aircraft, but nevertheless, Mr Bloggs is entitled to rely on the air traffic controller to do his or her job because the air traffic controller’s role is necessarily one the public must rely on. Justice Mason treated general reliance as a concept which might be regarded more or less as on judicial notice. Thus he said that in some situations, plaintiffs could argue that general reliance existed and could give rise to a duty of care.
This distinction also reflected a particular concern about public authorities, namely the limited resources they may have from the public purse, and often policy matters are taken to include budgetary allocation decisions. Later Australian cases have followed Heyman’s case (paying particular attention to the judgment of Mason J), and refined and expanded it.
In Pyrenees Shire Council v Day[50] special and general reliance were considered again. The majority emphasised the fictional nature and the breadth of operation of the doctrine of reliance. Justice Gummow rejected general reliance as having too much affinity with public law theories of procedural fairness, and emphasised the need for the duty of care to arise independently as a matter of private law through the consideration of reasonable foreseeability and the salient factors relevant to the situation. These could include vulnerability, control and knowledge. Chief Justice Brennan preferred to use statutory intention rather than the doctrine of general reliance to found a duty of care.
The rejection of the doctrines of general and specific reliance does not mean that the facts of general or specific reliance could not be argued. Thus they have been demoted from legal issues to factual issues.
The Civil Liability Act
The pre-Civil Liability Act position
Justice Ipp drew the threads of the Australian jurisprudence together in relation to the failure to exercise statutory powers in Amaca Pty Ltd v State of New South Wales.[51] Justice Ipp (with whom Mason P and McColl JA agreed) drew on the judgment of Mason J in Sutherland Shire Council v Heyman to put forward the following propositions:
“(a) Generally, a public authority, which is under no statutory obligation to exercise a power, owes no common law duty of care to do so.
(b) An authority may by its conduct, however, attract a duty of care that requires the exercise of the power.
(c) Three categories are identified in which the duty of care may so be attracted.
(i) Where an authority, in the exercise of its functions, has created a danger.
(ii) Where the particular circumstances of an authority’s occupation of premises or its ownership or control of a structure attracts to it a duty of care. In these cases the statute facilitates the existence of a duty of care.
(iii)Where a public authority acts so that others rely on it to take care for their safety.
...
Mason J did not suggest that the categories of circumstances in which an authority may attract a duty of care are closed. Later cases, which are in accord with the general propositions laid down in Sutherland Shire Council v Heyman, have extended the categories to situations where an authority has control over a particular situation that carries with it a risk of harm of which the authority knows or should know. That is to say, where an authority has such control, a duty of care may (not, I stress, must) be recognised.”[52]
Justice Ipp said later:
“I start the inquiry from the standpoint that, generally, a public authority, which is under no statutory obligation to exercise a power, owes no common law duty of care to do so. A common law duty of care will, however, arise when an authority, by its conduct, acts in a manner that requires the exercise of the power. The totality of the relationship between the parties has to be taken into account in determining whether a duty of care is to be recognised. The mere fact that statutory powers are provided and the authority concerned knows that harm may result from a failure to take affirmative action is insufficient to create an affirmative duty of care.”[53]
The later cases he referred to included Pyrenees Shire Council v Day,[54] Crimmins v Stevedoring Industry Finance Committee and Graham Barclay Oysters Pty Limited v Ryan.[55]
The Ipp Report[56]
The Review of the Law of Negligence: Final Report (the Ipp Report) took the view that public authorities do need to be treated differently from private individuals, largely because one can have a choice which benefits the wider community even though it damages an individual or small group.[57] They recommended the test of Wednesbury unreasonableness should apply to claims in negligence based on policy:
“In any claim for damages for personal injury or death arising out of negligent performance or non-performance of a public function, a policy decision (that is, a decision based substantially on financial, economic, political or social factors or constraints) cannot be used to support a finding that the defendant was negligent unless it was so unreasonable that no reasonable public functionary in the defendant’s position could have made it.”[58]
This appears to be aimed at reducing the standard of care which public authorities had to meet. It was not intended to provide immunity. However, it was only aimed at reducing the standard of care public authorities had to meet in a situation where up until then a public authority had been regarded as having no duty of care. It therefore was actually aimed at extending the duty of public authorities.[59] Elizabeth Carroll has argued that the test “is a means of preventing the courts from undertaking merits review when applying the standard of care in negligence”.[60] By this she means that in administrative law the Wednesbury test is aimed at constraining the courts from setting aside administrative decisions on the basis of the merits as opposed to the legality of the decision.[61]
“Public”
Section 41of the Act defines “public authorities” relatively broadly, but not by reference to any principle — this is different from the Consultation Paper.
The reform provisions vary across the jurisdictions. One matter which varies considerably is the class of bodies or people regarded as covered by the legislation. The broadest coverage is that of New South Wales which covers statutory authorities, governmental authorities and bodies having some sort of public function. Section 41(e1) extends the definition of public or other authority to persons having “public official functions”. This was enacted to extend the coverage to doctors exercising powers of detention under the Mental Health Act (see below). Queensland, the Australian Capital Territory, Tasmania, Western Australia and Victoria limit the protection to governmental bodies, although all jurisdictions except Queensland allow coverage to be extended by regulation. In New South Wales, for example, regulations extend the coverage of the protection to private schools.[62]
Justiciability and policy and when they should be considered
The Act has not made this any clearer except for a general perception that it should be harder to sue public authorities.
Section 42 which concerns allocation of resources implicitly invokes the policy/operation distinction insofar as budgetary allocations were frequently characterised as policy matters which were not justiciable. However, it does not even determine for us the question of whether this is to be determined at the duty stage or the breach stage. It thus fudges the question of immunity (despite the fact that the thrust of the Act is clearly to reduce the possibility of liability wherever possible). The section refers to civil liability so it presumably refers to any tortious or contractual matter which is not precluded by s 3B of the Act.[63]
Further, s 42 is so broadly worded that it fails to create clear limits — the provisions refer to financial and other resources “reasonably available” and the “general” allocation of resources. “General procedures” and “applicable standards” are referred to as available evidence of proper exercise of functions but there is no indication of whose general procedures and standards these are or where they are to be found. In State of New South Wales v Ball[64] this was one of the problems — what is the “general” as opposed to the “particular” allocation of resources. For Mr Ball the question of whether the police allocated sufficient resources to have enough staff to cover the child sexual assault unit so that he would not be overstretched and suffer post-traumatic stress disorder seemed fairly particular. However, the court did not agree that this was so.
The power section 43A applies to is a “special statutory power” — which is defined as “of a kind that persons generally are not authorised to exercise without specific statutory authority”. The section goes on to say that there will be no civil liability created by the exercise or failure to exercise a special statutory power unless the act or omission was “in the circumstances so unreasonable that no authority ... could properly consider the act or omission to be a reasonable exercise of, or failure to exercise, its power”: s 43A (3)..
It was passed to counter Presland v Hunter Area Health Service[65] in which a man having a severe psychotic episode was negligently released by a psychiatrist whereupon he went home and murdered his brother’s fiancé. He later sued the psychiatrist for the time he spent as a forensic patient. He won at first instance, and although this was reversed in the Court of Appeal,[66] this was seen as an attempt to profit from his crime and the Parliament passed s 43A in response. It makes it impossible to sue a person exercising “special statutory powers” unless Wednesbury unreasonableness can be established, thus reducing their standard of care. A person exercising “special statutory powers” clearly includes psychiatrists operating under the Mental Health Act, but it also may include persons such as prison guards and others who meet the definition of exercising a “special statutory power”.
Section 43A was considered in Precision Products (NSW) Pty Ltd v Hawkesbury City Council.[67] There the special statutory power purported to be exercised had been exercised invalidly. The invalidity concerned the authority of the council officer to issue the relevant notice. The council officer and the council itself were not aware that there was a deficiency in their notices. The primary judge held that s 43A protected the council.
Justice Allsop (Beazley and McColl JJA agreeing) said that it was not clear whether s 43A provides an immunity or affects the duty of care.[68] He treated it as a matter to be considered at the stage of breach, although he did not want to express a concluded view on the meaning and scope of s 43A.
He held that the approach to the duty of care in this context had to be considered by carrying out three steps:
Here reliance, assumption of responsibility and vulnerability were absent. The purposes of the Protection of the Environment Operations Act 1997 emphasised the problem:
“The imposition of a duty of care to have regard to the economic interests of a person in the position of the appellant in the way proposed would be to subject the Council, whose responsibility is to exercise the power in the public interest, to a duty to have regard to the conflicting interests and claims of the party whose conduct ... may have endangered the environment and the public interest. The setting up of this tension between the statute and common law should not be permitted: Sullivan v Moody at 582.”[69]
He noted that to confine s 43A to validly exercised statutory powers would require the exercise of particular statutory interpretation rules such as the rule that common law rights should not be cut down unnecessarily and should be read jealously.[70] He went on to say that if s 43A was to be seen as affecting purported exercise of powers then some other limiting factor might be needed, such as bona fide attempts, etc, as set out by Dixon J in R v Hickman; Ex p Fox.[71] I do not see why any further limiting factor beyond those already in operation in the law of negligence might be needed. This is a new statutory application of the law of negligence, I would submit, and it is fairly plain in its meaning. This was a case of pure economic loss which has very strong limiting factors in its requirements for the duty of care.
The strongest argument for saying that s 43A could only apply to validly exercised statutory powers would be that this would increase the protection of public authorities from private actions for compensation (which is, of course, quite consistent with the general approach of the Act). The counter argument is that if s 43A does not apply to something then its protection does not apply either. This somewhat paradoxical result suggests that s 43A should not be so confined.
Justice Allsop also considered the Wednesbury unreasonableness test and suggested that such tests are also to be found in the law of company directors and other areas of private law. In the event, he did not think that here there was an act so unreasonable that no public authority could consider it to be a proper exercise of the power and therefore held that it did not fall within s 43A’s proviso.
Section 44 is another section which looks as if it might refer to the policy/operation distinction. However, it focuses on the availability of mandamus. This is more like Lords Reid and Diplock in Dorset Yacht who wished to decide whether an action was ultra vires before they would determine whether there was negligence. However, this approach has been firmly rejected in Australia until now. In Heyman the court was determined to avoid going through a public law process in order to determine a private law matter.
Failure to exercise a statutory power
At common law the failure to exercise a statutory power was the most problematic category of liability of public authorities.
The Civil Liability Act has made it clear that the failure to exercise a regulatory “function”[72] is not to create liability unless the authority “could not have been required to exercise the function in proceedings instituted by the plaintiff”.[73] In other words, unless mandamus would be available to the plaintiff.
Where an act has been begun but not completed, the Act by s 46 emphasises the protection by holding that the fact that a function has been exercised does not “of itself” indicate a duty. What does the “of itself” add to this? In Pyrenees[74] the council was held liable partly because it began the process of notifying the parties without continuing it; but did that “of itself” create the liability the High Court found? I don’t think so. Mark Aronson suggests that s 46 reverses Pyrenees but perhaps it doesn’t.
Omissions or nonfeasance used to be significant in determining whether a road authority could be held liable. This was abolished in Brodie and Ghantous.[75] Section 45, which reverses the new common law position, applies to roads authorities and road work within the meaning of the Roads Act 1993 (NSW).
Applying the Civil Liability Act
Consider how the Civil Liability Act might apply to Kirkland-Veenstra, the first case set out at the beginning of this article. The police are clearly within the definition of public authorities pursuant to s 41. Because there would be few implications for the allocation of resources section 42 would probably not apply – the police would have to do no more than make another decision or check on the person again . Section 43A concerns the exercise or failure to exercise of a special statutory power. Does it apply to this case? The Mental Health Act 2007 s 22 (1) provides:
(1) A police officer who, in any place, finds a person who appears to be mentally ill or mentally disturbed may apprehend the person and take the person to a declared mental health facility if the officer believes on reasonable grounds that:
(a) the person is committing or has recently committed an offence or that the person has recently attempted to kill himself or herself or that it is probable that the person will attempt to kill himself or herself or any other person or attempt to cause serious physical harm to himself or herself or any other person, and
(b) it would be beneficial to the person’s welfare to be dealt with in accordance with this Act, rather than otherwise in accordance with law.
(2) A police officer may apprehend a person under this section without a warrant and may exercise any powers conferred by section 81 on a person who is authorised under that section to take a person to a mental health facility or another health facility.
Clearly this involves a power which is not available to the general public, so it is a special statutory power. If s 43A applies then the question which has to be answered is whether the police in exercising or failing to exercise the power acted in the words of sub-section (3):‘so unreasonably that no authority having the special statutory power in question could properly consider the act or omission to be a reasonable exercise of, or failure to exercise , its power’.
We cannot evaluate the role of the police here without considering the objects of the Act which they are supposed to facilitate.
3. The objects of this Act are:
(a) to provide for the care, treatment and control of persons who are mentally ill or mentally disordered, and
(b) to facilitate the care, treatment and control of those persons through community care facilities, and
(c) to facilitate the provision of hospital care for those persons on a voluntary basis where appropriate and, in a limited number of situations, on an involuntary basis, and
(d) while protecting the civil rights of those persons, to give an opportunity for those persons to have access to appropriate care, and
(e) to facilitate the involvement of those persons, and persons caring for them, in decisions involving appropriate care, treatment and control.
At common law it might be arguable that a reasonable person in the position of the police officers (the circumstances including the objects of the Act giving them their power) would do more in such serious circumstances than these police officers did. That is, since the power is supposed to be exercised in a facilitative way, there would be some prospect of the plaintiff being able to persuade a court that where the police simply checked and then went away, that that was not the act of a reasonable police officer in the circumstances. In this case the High Court was not convinced, but the possibility remains far greater in such a situation than it does using the Wednesbury unreasonableness test set out in s 43A. It would be extremely difficult to say that the actions of the police officers was so unreasonable that no police officer could every think that was a reasonable response. Justice Allsop was right to suggest that s 43A creates something approaching an immunity.[76]. The question of whether the police in stopping and talking to the plaintiff’s husband began to exercise their power is neutralised as a liability issue for the police by s 46. Liability under the Civil Liability Act is extremely unlikely.
The second case, Couch concerns the negligent exercise of power in the form of a failure to closely supervise. Without more information it is unclear whether the failure to supervise in any way was affected by allocation of resources, so it is unclear whether s 42 would have an impact. This may be a section 43A special statutory power - it is conferred by statute; the power is a power conferred to supervise parolees and is in part conferred to protect the public but s 43 is clearly relevant. In either case, the test for breach is the Wednesbury test. Section 43 is confusingly drafted -it is not clear whether it concerns the separate tort of breach of statutory duty , (hereafter ‘BSD’) or whether it is the tort of negligence when a statute is involved. It also does not clarify whether to consider the Wednesbury test at the duty or breach stage. Since the provision uses a test of reasonable behaviour presumably it is to be used at breach stage; furthermore, since it is framed in this way it seems that the provision must concern negligence rather than BSD.[77] The Wednesbury test turns the question into this: was the parole officer’s failure to closely supervise so unreasonable as to be something no reasonable parole officer could do? In Couch itself which was an action for exemplary damages to take it outside the scope of the Accident Compensation Commission of New Zealand, the standard required ‘truly exceptional and outrageous’[78] conduct on the part of the parole officer. Clearly it and the Wednesbury test would both be much easier for the defendant to meet than the ordinary standard of care. Whether s 43 and s 43A of the Civil Liability Act confer immunity on public authorities is not clear; but it is quite clear that they make it very much much harder to succeed in an action against them.
An approach to the liability of public authorities
General
The Civil Liability Act prevents us from discarding the public/private divide. The following section concerns how I would define and deal with this if I were not constrained by the Civil Liability Act. If I could, I would prefer to look at functions rather than the creation of the body in order to determine “publicness”. I would suggest that the preferable view would be to have two stages of determining “publicness” of function — a justiciability stage which needs to be considered at the duty of care stage, and a second look at it which needs to be considered at the breach stage of a negligence action.
“Public” functions which should suggest that there should be no duty of care would include the following:
Further, the fact that a function is something that private citizens cannot do on their own, in my opinion should not come into this non-justiciable category of public function. The example of the Mental Health Act power exercised in Presland is similar to the example the Ipp committee thought should give a policy defence.[80] However, I don’t think the Presland example is a good one. That is why I would prefer the formulation above, that someone necessarily will be harmed, before the public nature of the function should give rise to a “defence” of no duty.
All other areas regarded as “public” should be regarded simply as circumstances to be taken into account in determining breach of duty.
What has been problematic here, in my opinion, is not that the public authority is ultimately held liable or not liable. It is that the courts have often tried to deal with this by using some kind of immunity; that is, by shutting the gates before the circumstances and the reasonableness of the behaviour has been tested; in particular by treating a larger number of categories of “publicness” as significant for the purposes of justiciability than they should.
In Australia the common law appears to have been moving to a position where the issue of justiciability or policy/operation (or whatever it is at the moment) is dealt with at the breach stage. This has the advantage of meaning that there is no immunity.[81] The same position seems to have been reached in Britain through the Human Rights Act and the European Court of Human Rights opposition to the concept of an “immunity” from suit. However, in my view where matters are non-justiciable should be a very small number area, so the corresponding immunity, if any should be small. However, aspects of “publicness” should be considered at both duty and breach stage.
After the non-justiciability stage I would suggest that the public issues should be regarded as one of the “salient factors”, requiring a frank consideration of the policy nature of the issues. This has not been overt in the Australian duty of care context where the High Court has not been able to establish a principled approach to the liability of public authorities. What would be required would be something like the Caparo test with a frank assessment of what the public nature of the authority required at the policy stage of the test. The advantage of having this dealt with at the duty stage is the ability of solicitors in their offices to be able to advise clients about liability with ease because there is a principled basis for the decision. For example, the use of some variant of the policy/operation distinction to distinguish between matters which are properly the province of judges and those which are not. “Policy” on its own is not necessarily public.
At the breach stage the usual test should be used: the standard test for breach of duty is whether the defendant acted like a reasonable defendant in the circumstances. “The circumstances” is a fairly wide net. It allows the consideration of the public or other nature of the defendant, it allows the weighing up of a range of possibilities including the allocation of resources and the extent to which the defendant is required to act as it did. Generally I would prefer not to use the Wednesbury test because the reasonable person standard of care is like the onus of proof in Briginshaw — it shifts with the difficulty and importance of the case. Wednesbury[82] unreasonableness may say only where there is gross negligence will there be liability.
However, the real reason for arguing that this should be done at the breach stage is that it is here that all the facts of the case are considered, and that matters such as the need for a public authority to carry out some particular program which benefits the community as a whole while damaging some individuals can be dealt with as part of the calculus of negligence. As the example of the Yuen Kun Yeu case above shows, I think this can be done. It might also make us more congruent with the civil law world.[83]
The disadvantage, if it is one, is that putting this into the breach test will deliver it into the hands of the lower courts, and may make it less possible to predict the liability (or not) of a public authority. However, I would suggest that the current situation where the High Court give pronouncements, such as “the word ‘reasonable’ must do real work”[84] etc gives no more guidance than this would.
Interpreting the Civil Liability Act
The Civil Liability Act also does not give much guidance on many of these matters. We have already established that the definition of public authorities in the Civil Liability Act seems to be problematic. It has been defined as broadly as possible with a view to reducing litigation in general, rather than on the basis of a principled view of what matters judges should properly determine.
Given the Civil Liability Act as it is, how is Pt V to be interpreted? The Act does not exclude common law. Throughout it there are references which are only understandable on the basis of common law and this applies to Pt V as elsewhere. Consideration of the Second Reading Speech makes it clear that the Act is designed to reduce litigation, but it is not particularly helpful apart from that. It clearly pushes towards immunity for public authorities but does not do so cleanly or on any principled basis.
It deals with the allocation of resources[85] but it does not tell us whether we should consider this at the duty or breach stage, although there are reasons to consider that the breach stage is preferable, not least because the common law has been moving in that direction. .[86] It greatly expands the definition of “public” and public function. It brings in the Wednesbury unreasonableness test. It restores the highway authority nonfeasance protection. It does not clearly explain how the relationship between negligence involving a statute and the tort of breach of statutory duty applies.
So how should the interaction of my preferred approach and the Civil Liability Act be managed? The following steps are proposed:
Has the duty been breached? How would a reasonable public authority in the circumstances act? For example, how would they allocate resources in the circumstances? At this stage the court should take a further detailed examination of the legislation creating the power the authority is to exercise, this time with a view to determining what a reasonable public authority would do in the circumstances (including the statutory circumstances). Section 42 would limit the extent to which the allocation of resources could be considered in comparison with the common law. Sections 43 and 43A of course would change the standard of care, making it harder for plaintiffs to succeed. In my opinion this standard of care is inappropriate since it confers something very close to immunity. Section 44 should be repealed as inappropriate in a negligence action because it incorporates mandamus Section 45 should also be repealed because it re-establishes an immunity, and the test for breach is affected by the actual knowledge provisions. Section 46, as discussed above, may not have very much effect because of the words “of itself” in the provision. With these changes the combination of the common law and the Civil Liability Act might develop some coherence in its treatment of public authorities.
Conclusion
The Civil Liability Act has certainly not solved the problems of dealing with the liability of public authorities. The problems are real, which is why they have not gone away over the past two centuries. No doubt the approach I have put forward has flaws which I have not yet appreciated. The major principles underlying the approach are that the idea of special treatment for public authorities is problematic for the rule of law and that therefore, while there are good reasons to sometimes treat them differently, these should be minimised. While it is clear that the Civil Liability Act is designed to make it harder to sue public authorities, this is not a good enough justification in itself to create immunities or allow them to continue. I hope that judges will be able to remedy the lack of principle demonstrated in the Act by a more principled approach to the interaction of the Act with the common law’s understandings if not of the solutions, at least of the problems in this difficult area of law.
[1] Edited version of a paper presented to the Supreme Court of New South Wales Annual Conference, 21–23 August 2009, Hunter Valley.
[2] Professor, Co-Director, Private Law Policy and Research Group, Faculty of Law, University of New South Wales.
[3] Unless otherwise specified, references to sections below are references to sections of the Civil Liability Act 2002 (NSW).
[4] Mental Health Act 1986 (Vic).
[5] Stuart v Kirkland-Veenstra [2009] HCA 15; (2009) 237 CLR 215.
[6] Couch v Attorney-General [2008] NZSC 45.
[7] F Trindale, P Cane, M Lunney, The law of Torts in Australia, 4th edn, Oxford University Press, Melbourne, 2007.
[8] ibid at p 608.
[10] ibid at [119]–[120].
[12] For example, the drive to coherence seen in the rejection of the critical nature of some of the former requirements for the duty of care in nervous shock cases: Tame v New South Wales [2002] HCA 35; (2002) 211 CLR 317.
[13] AV Dicey, Introduction to the study of the Constitution, 10th edn, Macmillan, London, 1959.
[14] Not contrary to this point, but of interest is the fact that the immunity was abolished early in New South Wales (and Australia) because of the very high level of engagement of the government in acvities involving commerce and construction.
[15] As S De Smith and R Brazier say: ”In the latter part of the last century Dicey’s polemic against the tyrannical nature of droit administratif prejudiced whole generations against administrative law in any form ... [The fact that the French administrative courts are entirely separate from the courts judging other matters] declared Dicey, was contrary to the rule of law, sacred to all Englishmen, which demanded that the conduct of the Administration and its officials be judged in the same common-law courts and governed by the same principles applied to ordinary citizens in private disputes ... [but] Like it or not, the common-law courts were forced to develop a system of principles of administrative law” in Constitutional and Administrative Law, 6th edn, Penguin Books, London, 1989, p 534. The Diceyan model, however, arises out of a polity which does not use the separation of powers as its model for the restraint of power; and Dicey was writing before the rise of negligence, so he saw trespass as the relevant action in these cases. Things have definitely changed.
[16] M Aronson and H Whitmore, Public torts and contracts, Law Book Company, Sydney, 1982, pp 34–35.
[17] C Harlow, State liability and beyond, Oxford University Press, New York, 2004, p 23.
[18] E Weinrib, The idea of private law, Harvard University Press, Cambridge, Massachusetts, 1995.
[19] Consultation Paper No 187, July 2008.
[20] ibid at [4.131].
[21] S Deakin, A Johnston and B Markesinis, Markesinis and Deakin’s Tort Law, 5th edn, Oxford University Press, Oxford, 2003, p 356 suggest that one of the issues for public authorities is the fact that they normally cannot become bankrupt and therefore one of the normal practical barriers to litigation does not exist. Often they get sued because the person more proximate to the problem has gone bankrupt, for example in Anns v Merton London Borough Council, the builder, etc. This is practically true for public authorities in Australia, although strictly speaking not legally true.
[22] M Aronson, “Government liability in negligence” [2008] MelbULawRw 2; (2008) 32(1) MULR 44.
[23] As distinct from the position where an employee of the Crown or other does something wrong and the Crown is vicariously liable.
[24] C Harlow, State liability:tort law and beyond, OUP, 2004, p 31.
[25] These private entities required statutes and many of them were not ordinary incorporations of commercial ventures. As Mike Taggart has shown they had to go through a hard fought committee process to get the statute passed, at least in the early days of the nineteenth century: Private property and abuse of rights in Victorian England, OUP, 2002.
[26] [1985] HCA 41; (1985) 157 CLR 424.
[27] New South Wales v Fahy [2007] HCA 20; (2007) 232 CLR 486; Sydney Water Corporation v Abramovic [2007] NSWCA 248; Hegarty v Queensland Ambulance Service [2007] QCA 366; Sydney County Council v Dell’Oro [1974] HCA 47; (1974) 132 CLR 97.
[28] Bennett v Minister of Community Welfare [1992] HCA 27; (1992) 176 CLR 408.
[29] Voli v Inglewood Shire Council [1963] HCA 15; (1963) 110 CLR 74; Mulligan v Coffs Harbour City Council [2005] HCA 63; (2005) 223 CLR 486; Vairy v Wyong Shire Council [2005] HCA 62; (2005) 223 CLR 422; Romeo vConservation Commission of the Northern Territory (1998) 192 CLR 431. This was also hardly mentioned in Sullivan v Moody (2001) 207 CLR 562. Medical cases where a public authority for example, an Area Health Authority, exists also seem to ignore the “public” factors.
[30] Geddis v Proprietors of Bann Reservoir (1878) 3 App Cas 430. Lord Blackburn disregarded the public nature of the defendant’s activities and took the view that where a statute created a body and gave it the power to supply water to mills on a riverbank, and the body failed to dredge the river so that it silted up and overflowed onto the plaintiff’s land, that this failure was unreasonable and amounted to a breach of the defendant’s duty of care to the plaintiff. Cases following this tended to accept the position that the public nature of the defendant should be generally disregarded until Anns v Merton London Borough Council [1977] UKHL 4; [1978] AC 728.
[31] Sullivan v Moody (2001) 207 CLR 562.
[32] For example, Leichhardt Municipal Council v Montgomery (2007) 230 CLR 22 (but no non-delegable duty owed for third parties actions); Kuru v New South Wales [2008] HCA 26; (2008) 236 CLR 1; Roads and Traffic Authority v Dederer (2007) 234 CLR 330; Home Office v Dorset Yacht Co Ltd [1970] UKHL 2; [1970] AC 1004.
[33] Anns v Merton London Borough Council [1977] UKHL 4; [1978] AC 728.
[34] Brodie v Singleton Shire Council; Ghantous v Hawkesbury City Council [2001] HCA 29; (2001) 206 CLR 512; although in relation to the police this makes it harder to establish a duty.
[35] Pyrenees Shire Council v Day [1998] HCA 3; (1998) 192 CLR 330 (depending on whether you think they had already started and therefore it was not a failure to exercise the power or not); Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; (1999) 200 CLR 1 (not breached); Swan (by his next friend) v State of South Australia [1994] SASC 5079; (1994) 62 SASR 532; Parramatta City Council v Lutz (1988) 12 NSWLR 293.
[36] There are far more of these though: for example, X (Minors) v Bedfordshire County Council [1995] UKHL 9; [1995] 2 AC 633; Sutherland Shire Council v Heyman [1985] HCA 41; (1985) 157 CLR 424; Stovin v Wise [1996] UKHL 15; [1996] AC 923; Stuart v Kirkland-Veenstra [2009] HCA 15; (2009) 237 CLR 215; Hill v Chief Constable of West Yorkshire [1989] AC 53; Graham Barclay Oysters Pty Ltd v Ryan [2002] HCA 54; (2002) 211 CLR 540; Amaca Pty Ltd v The State of New South Wales [2004] NSWCA 124.
[37] [1987] UKPC 16; [1988] AC 175.
[39] [2005] UKHL 24; [2005] 1 WLR 1495.
[40] ibid at [27].
[41] In Commonwealth of Australia v Mewett (1997) 191 CLR 471, Brennan CJ et al said that for constitutional purposes an “immunity” was not an immunity but a procedural bar. Whether this has an impact on this argument is something I need to think about. At present I don’t think so.
[42] Caparo Industries Plc v Dickman [1990] UKHL 2; [1990] 2 AC 605.
[43] Sullivan v Moody (2001) 207 CLR 562.
[44] C Harlow, “‘Public’ and ‘private’ law: definition without distinction’ (1980) 43(3) MLR 241 at 243.
[45] ibid.
[46] VE Schwartz, K Kelly, DF Partlett, Prosser, Wade and Schwartz’s Torts: Cases and Materials,10th edn, Foundation Press, New York, 2000, discusses governmental immunity’s development as based on sovereign immunity and the doctrine that the government cannot be sued without its consent (because of its status as lawmaker: Kawananakoa v Polyblank, [1907] USSC 90; 205 US 349, 353 (1907)). This was originally a procedural notion. Later it became a substantive notion and in Gibbons v United States, 75 US [1868] USSC 39; (8 Wall) 269 (Ct Cl 1868) it was held that the federal government was immune from all liability in tort. The United States waived its sovereign immunity under certain circumstances for torts (the discretionary function exception — in the statute) through the Federal Tort Claims Act of 1946 (US). The interpretation of the discretionary exception — the discretionary function exception applies only to activities involving an element of judgment or choice and based on public policy considerations: Berkovitz v United States, [1988] USSC 108; 486 US 531 (1988).
State governments in the United States had similar sovereign immunity. Municipal corporations had exceptions to immunity on the basis of the proprietary or private versus governmental functions: see. Prosser et al at p 639. D Dobbs, The Law of Torts, West Group, Saint Paul, Minnesota, 2001, p 718 says: “Courts conceived of the municipal government as operating in several different capacities. If it caused harm while acting in a purely governmental capacity, say in police activities, it enjoyed the immunity. But the municipality had no immunity for its torts when it operated in a corporate or proprietary capacity. For example, a municipality would be subject to liability for torts inflicted in the operation of a municipal electric or water utility”. This looks very like the policy/operation distinction as the proprietary function was one which was an “activity that normally was carried out by the private sector of the economy”: see Prosser et al at p 639. This continues to be used in some jurisdictions despite its messiness. It is also interesting to note that where the State has authorised the municipal corporation to have public liability insurance this is regarded as an implied waiver of the immunity to the extent of the insurance: Bollinger v Schneider, 64 Ill App 3d 758 (1978). However, Dobbs, The Law of Torts at 718-719 says that the test for governmental or proprietary function is not settled. Courts “have variously held that an activity is or tends to be proprietary (1) if it is carried on for profit (2) if a fee is paid, (3) if the activity relates to public service, whether or not a fee is paid, (4) if the city is under no duty to carry it out, or (4) if the activity is historically one carried out by private enterprise.’
The governmental/proprietary test, however, is distinct in the USA from the policy/discretionary immunity. Dobbs at pp 720–722 states: “Judicial and legislative decisions are easily included in this category, but so are some executive-branch decisions and action ... the idea is that social and economic policy is to be fixed by legislative and executive branches of the government, not the judiciary, and that the judicial branch must not intrude on those basic decisions. In some instances, this same policy is advanced by the separate rule called the public duty doctrine. The policy dictates that public entities must not be held liable for passing or failing to pass legislation ... Some state cases refuse to grant the immunity unless the governmental entity made a conscious choice of conduct ... Some states also utilize the distinction between planning and operational decisions, limiting the immunity to cases of ‘planning’ and excluding it for actual operations or executions of decisions. Similarly, state cases have often drawn a distinction between discretionary, legislative, or judicial acts on the one hand, and ministerial acts on the other, with immunity for the former only”.
However, even where there is a governmental function sometimes municipal corporations have not been seen as immune: Ayala v Philadelphia Board of Public Education 453 Pa 584 (1973) where they said: “The distinction between governmental and proprietary functions ‘is probably one of the most unsatisfactory known to the law, for it has caused confusion not only among the various jurisdictions but almost always within each jurisdiction’ Davis, Administrative Law Treatise... (1958)”.
In the United States public officers (not legislators and judges, or the President, etc) may be personally liable for tortious conduct committed in the course of official duties, such as under the Constitution or Civil Rights Act: for example, Bivens v Six Unknown Named Agents, 403 US 388 (1971), The United States also has specific immunities for particular public officials.
[47] Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; (1999) 200 CLR 1 per Gummow and Hayne JJ.
[48] Graham Barclay Oysters Pty Ltd v Ryan [2002] HCA 54; (2002) 211 CLR 540.
[49] [1985] HCA 41; (1985) 157 CLR 424.
[50] [1998] HCA 3; (1998) 192 CLR 330.
[52] ibid at [22], [23]. This is not an exhaustive outline of the situations where a public authority might be held liable.
[53] ibid at [140].
[54] [1998] HCA 3; (1998) 192 CLR 330.
[55] [2002] HCA 54; (2002) 211 CLR 540.
[56] D Ipp et al, Review of the Law of Negligence: Final Report, Commonwealth of Australia, 2002.
[57] Surely this could be dealt with using part of the calculus of negligence.
[58] Ipp, above, n 53, p 158, recommendation 39.
[59] Thank you to Emeritus Professor Mark Aronson for this point.
[60] E Carroll, “Wednesbury unreasonableness as a limit on the civil liability of public authorities” (2007) 15(2) Tort L Rev 77 at 80.
[61] ibid at 79–80 (note of the discussion by Brennan J in Attorney-General (NSW) v Quin (1990) 170 CLR 1 at 35–36).
[62] Civil Liability Regulation 2009 (NSW), cl 4.
[63] Section 3B excludes civil liability in respect of an intentional act that is done with intention to cause injury or death or that is sexual assault or other sexual misconduct. It also does not apply to certain matters such as dust disease, tobacco products, motor accidents Acts (except for some of the negligence provisions of the relevant Act), Workers Compensation Act 1987, etc. Thus only some intentional torts are excluded if they concern an intention to cause injury or death — that is not the same as the general concept of intention for torts such as trespass which only require intention to touch, etc. See Carrier v Bonham [2001] QCA 234; [2002] 1 Qd R 474.
[64] [2007] NSWCA 71; (2007) 69 NSWLR 463.
[66] Hunter Area Health Service v Presland [2005] NSWCA 33; (2005) 63 NSWLR 22.
[68] ibid at [171].
[69] ibid at [114].
[70] But would s 43A have any real meaning if it were to be confined to validly exercised statutory powers? Surely the Wednesbury test as it is stated in that section would have little role if it could only be applied to an already invalid statutory power. It goes against the trend of Australian authority in relation to negligence to apply such public law notions in the case of negligence.
[71] [1945] HCA 53; (1945) 70 CLR 598.
[72] Defined in s 41 as a “power, authority or duty”.
[74] [1998] HCA 3; (1998) 192 CLR 330.
[75] [2001] HCA 29; (2001) 206 CLR 512.
[76] See n 70 above.
[77] The test for breach of statutory duty depends entirely on what the statute says — it can range from strict liability to the ordinary test of negligence. Given that the thrust of this legislation is to reduce liability it is highly unlikely that the legislature would want to reduce the standard for breach from strict liability even to Wednesbury unreasonableness. Therefore, if the legislature is setting out the test for breach they must not be considering breach of statutory duty as a tort, but rather the tort of negligence.
[78] Couch v Attorney General [2008[NZSC 45 per Elias CJ at [11], quoting Lord Nicholls in Bottrill v A [2003] 2 NZLR 721 at para [37]
[79] I would therefore argue that the view in Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; (1999) 200 CLR 1 that “quasi-legislative” act being required of the Stevedoring Industry Finance Committee should not be regarded as sufficient to make the matter non-justiciable. In my opinion, this should not apply to delegated legislation and especially to non-elected bodies which are not really legislatures.
[80] That is, where an individual’s interests had to be considered against a wider set of interests or there was an allocation of resources issue that had to be considered.
[81] Although dealing with it as a policy matter at the duty stage would also be sufficient to prevent an immunity arising.
[82] Harlow, above n 15, p 130 says: “Perhaps the much-maligned Wednesbury test is simply an objective bad faith standard, capable of application alike to private and public bodies”. Harlow wants public law compensation to retract, but actions against public officials (for example, misfeasance in public office) to be emphasised and expanded into private law so that it covers “all those ‘in positions of corporate responsibility’”: per Lord Steyn in Kuddus v Chief Constable of Leicestershire Constabulary [2001] UKHL 29; [2001] 2 WLR 1789 at para 66: see C Harlow, p 130.
[83] See M Vranken, Fundamentals of European Civil Law, Federation Press, Sydney, 1997.
[84] See Vairy v Wyong Shire Council [2005] HCA 62; (2005) 223 CLR 422, Mulligan v Coffs Harbour City Council [2005] HCA 63; (2005) 223 CLR 486, Romeo v Conservation Commission of the Northern Territory (1998) 192 CLR 431.
[85] The aim of the policy/operation distinction.
[86] See, for example, Crimmins v Stevedoring Industry Finance Committee (1999) 200 clr 1 per McHugh J at [131].
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/UNSWLRS/2010/47.html