Litigation notes No. 17

Number 17
8 October 2008

This issue

Waiver of legal professional privilege

Recent High Court decision on binding nature of implied undertakings in litigation

Police powers of entry without warrant on private premises where criminal activity apprehended

Liability for damages over road design and construction

High Court rules on breach of duty of care


Paul Sykes

Paul Sykes Senior Lawyer
T 02 6253 7050
F 02 6253 7302

The recent High Court decision in Osland v Secretary to the Department of Justice [2008] HCA 37 (7 August 2008) deals with waiver of legal professional privilege (at common law). It also raises important issues in relation to the operation of provisions of the Freedom of Information Act 1982 (Vic) (the FOI Act), principally s 50(4). Section 50(4) confers power upon the Victorian Civil and Administrative Tribunal (the VCAT) to decide that access should be granted to an exempt document (subject to some exceptions) if it is of the opinion that the public interest requires access. (The joint judgment of four of the Justices of the High Court describes s 50(4) as 'a unique provision in Australian freedom of information legislation' (at [21].)

Osland v Secretary to the Department of Justice
High Court of Australia, 7 August 2008
[2008] HCA 37


The decision is an affirmation of the 'inconsistency test' for determining waiver of legal professional privilege, as stated by the High Court in its decision in Mann v Carnell (1999) 201 CLR 1. It goes further, though, to emphasise that inconsistency will not arise as easily as some decisions since Mann v Carnell have indicated, particularly those that have found waiver to have occurred where the gist or substance of the advice is said to have been disclosed by the reference made to the advice. As four of the Justices state (at [42]):

Whether, in a given context, a limited disclosure of the existence, and the effect, of legal advice is inconsistent with maintaining confidentiality in the terms of advice will depend upon the circumstances of the case.

The High Court's decision will bring a greater assurance that the risk of waiver under the 'inconsistency test', particularly where appropriate care is taken in the way reference is made to an advice, may not be as great as has sometimes been feared.


On 2 October 1996, a jury of the Victorian Supreme Court found the appellant, Mrs Marjorie Osland, guilty of the murder of her husband, Mr Frank Osland, on 30 July 1991. The same jury was unable to reach a verdict in relation to her son, Mr David Albion, also charged with the murder. Both Mrs Osland and Mr Albion had relied upon defences of self-defence and provocation. The evidence was that Mr Albion wielded the iron bar that killed Mr Osland and that Mrs Osland had planned the killing and assisted in carrying it out. Subsequently, Mr Albion was retried on the charge of murder but was acquitted.

On 12 November 1996, Hedigan J sentenced Mrs Osland to 14 1/2years imprisonment, with a non-parole period of 9 1/2years. In delivering his sentence, Hedigan J noted that, prior to the murder, Mrs Osland had been subjected to repeated violence from her late husband.

Mrs Osland appealed to the Victorian Court of Appeal against her conviction and sentence but was unsuccessful. On 10 December 1998, a further appeal (by special leave) to the High Court of Australia was dismissed by a 3 to 2 majority.

On 5 July 1999, Mrs Osland submitted a petition for mercy to the then Victorian Attorney-General, Mrs Jan Wade, in which she sought a pardon from the Governor of Victoria. On 6 September 2001, following a change of government, the new Attorney-General, Mr Rob Hulls, announced that the Governor had refused the petition. In a press release of that day, the Attorney-General said:

On July 5, 1999, Osland submitted a petition for mercy to the then Attorney-General Jan Wade. That petition set out six grounds on which the petition should be granted.

Following consultation with the State Opposition, I appointed a panel of three senior counsel, Susan Crennan QC, Jack Rush QC and Paul Holdenson QC, to consider Osland's petition.

This week I received a memorandum of joint advice from the panel in relation to the petition. The joint advice recommends on every ground that the petition should be denied.

After carefully considering the joint advice, I have recommended to the Premier that the Governor be advised to deny the petition.

The Governor has accepted this advice and denied the petition.

In addition to the joint advice referred to in the press release, a number of other advices had been provided to the government in connection with the petition. Mrs Osland made an application under the FOI Act for access to the documents containing, or referring to, these advices (11 in all). The Secretary to the Department of Justice denied access on the ground that the documents were exempt documents under the FOI Act. The ground of exemption relied upon was that, under s 32 of the FOI Act, they were the subject of legal professional privilege.

Mrs Osland applied for review of this decision to the VCAT. (The number of documents in dispute at this point decreased to nine, as it emerged that Mrs Osland already had access to one and, in respect of another, she had withdrawn her claim.) On 16 August 2005, the VCAT ordered that Mrs Osland be given access to the nine documents which were in dispute. The VCAT concluded that:

  1. the documents were the subject of legal professional privilege
  2. there had been no waiver of the privilege, but
  3. notwithstanding that they were exempt documents, access should be granted because the VCAT was of the opinion that the public interest required such access (s 50(4)).

The Secretary appealed to the Victorian Court of Appeal, by leave granted on 14 October 2005, against that order of the VCAT. The issue in the appeal was confined to whether:

  1. legal professional privilege had been waived in the joint advice referred to in the press release, and
  2. in any event, access should be granted to all the documents in question on the ground under s 50(4) of the FOI Act that such access was required by the public interest.

In relation to the joint advice, Mrs Osland submitted that, by issuing the press release, the Attorney-General waived legal professional privilege in the contents of the joint advice. This was said to be a case of implied (or imputed) waiver, resulting from the following statements in the press release:

This week I received a memorandum of joint advice from the panel in relation to the petition. The joint advice recommends on every ground that the petition should be denied.

After carefully considering the joint advice, I have recommended to the Premier that the Governor be advised to deny the petition.

Counsel for Mrs Osland relied on the following statement by Gyles J as a member of the Full Court of the Federal Court in Bennett v Chief Executive Officer, Australian Customs Service (2004) 140 FCR 101 (Bennett), 119, at [65]:

The voluntary disclosure of the gist or conclusion of the legal advice amounts to waiver in respect of the whole of the advice to which reference is made including the reasons for the conclusion.

On 17 May 2007, the Court of Appeal dismissed Mrs Osland's appeal. In giving the main judgment, Maxwell P declined to follow the statement of Gyles J in Bennett. He said of this statement (at [29]):

As will appear, this statement has been applied subsequently as if it were a rule of general application. For reasons which follow, I am respectfully unable to accept that any such general rule is either justified by the authorities or compatible with the inconsistency test as enunciated in Carnell [Mann v Carnell (1999) 201 CLR 1].

Maxwell P said later in his judgment (references omitted) (at [51]):

As Carnell demonstrates, the inconsistency test readily accommodates the notion that, in appropriate circumstances, the privilege-holder may disclose the content of legal advice to a third party for a particular purpose without being held to have waived privilege in the advice. Likewise, in my opinion, the test of inconsistency is well capable of accommodating the notion that, in appropriate circumstances, the privilege-holder should be able to disclose publicly that it is acting on advice and what the substance of that advice is, without being at risk of having to disclose the confidential content of the advice.

Ashley JA and Bongiorno AJA agreed with Maxwell P in separate judgments that there had been no waiver of legal professional privilege in the issuing of the Attorney-General's press release.

In relation to s 50(4) of the FOI Act, the Court disagreed with the VCAT, holding that there could be no basis upon which, on the material before the VCAT, an opinion could be formed under s 50(4) of FOI Act that the public interest requires that access to the exempt documents be granted under that Act. The Court made this decision without looking at the documents concerned.

On 14 December 2007, Ms Osland was granted leave to appeal to the High Court on grounds related to the Court of Appeal's rulings that:

  1. there had been no waiver of legal professional privilege, and
  2. there was no basis for the VCAT forming the view for the purposes of s 50(4) of the FOI Act that the public interest required that access to the exempt documents be granted.

Appeal to the High Court

The High Court allowed Mrs Osland's appeal by a 5 to 1 majority as it related to s 50(4) of the FOI Act but unanimously dismissed the grounds relating to waiver of legal professional privilege.

Reasoning supporting the dismissal of the argument that waiver had occurred

In giving the main judgment, Gleeson CJ, Gummow, Heydon and Kiefel JJ made clear (at [42]) that both parties to the appeal accepted that the principles to be applied were those of the joint judgment of the majority in Mann v Carnell. What was at stake here was the application of these principles to the circumstances, involving the issue of the press release.

Their Honours then stated those general principles as they bore on the circumstances, saying (references omitted) (at [45] and [46]):

Waiver of the kind presently in question is sometimes described as implied waiver, and sometimes as waiver 'imputed by operation of law'. It reflects a judgment that the conduct of the party entitled to the privilege is inconsistent with the maintenance of the confidentiality which the privilege is intended to protect. … In the present case counsel for the appellant acknowledged that, if the press release had not included the sentence earlier identified as critical [viz. 'The joint advice recommends on every ground that the petition should be denied.'], privilege probably would not have been waived. This is undoubtedly correct, even though, upon that hypothesis, the press release would have made some disclosure concerning legal advice taken by the Department.

The conduct of the Attorney-General in issuing the press release and including in it certain information about the joint legal advice is to be considered in context, which includes the nature of the matter in respect of which the advice was received, the evident purpose of the Attorney-General in making the disclosure that was made, and the legal and practical consequences of limited rather than complete disclosure.

Applying those principles here, their Honours said (at [48]):

The evident purpose of what was said in the press release was to satisfy the public that due process had been followed in the consideration of the petition, and that the decision was not based on political considerations. … The Attorney-General was seeking to give the fullest information as to the process that had been followed, no doubt in order to deflect any criticism, while at the same time following the long-standing practice of not giving the reasons for the decision. This did not involve inconsistency; and it involved no unfairness to the appellant. If she had a legal right to reasons for the decision, then she still has it. If she had no such right, the press release did not deprive her of anything to which she was entitled. What the Attorney-General said did not prevent the appellant from making public her petition, or any part of it, as and when she desired.

Their Honours thus acknowledged that a limited disclosure of the existence, and the effect, of legal advice could be consistent with maintaining confidentiality in the actual terms of the advice (at [48]–[50]). They further observed (references omitted) (at [49]):

Whether, in a given context, a limited disclosure of the existence, and the effect, of legal advice is inconsistent with maintaining confidentiality in the terms of advice will depend upon the circumstances of the case. As Tamberlin J said in Nine Films and Television Pty Ltd v Ninox Television Ltd (2005) 65 IPR 442 at 447, at para [26], questions of waiver are matters of fact and degree. It should be added that we are here concerned with the common law principle of waiver, not with the application of s. 122 of the Evidence Act 1995 (Cth) which, as was said in Mann v Carnell, has the effect that privilege may be lost in circumstances which are not identical to the circumstances in which privilege may be lost at common law.

Gleeson CJ, Gummow, Heydon and Kiefel JJ stated in conclusion (at [50]) that the reasoning of Maxwell P in the Court of Appeal was correct.

Kirby J, in his separate judgment, came to the same conclusion that there had been no waiver. The main considerations for him in this regard were (at [90]):

  • The press release revealed very little about the actual content of the joint advice, aside from the names of its authors and their adverse conclusions.
  • The purpose of issuing the press release was not, as such, to secure some advantage for the State in legal proceedings affecting the appellant. Rather, the purpose was to show, as far as was compatible with non-disclosure, that the State had taken a proper course in obtaining and considering advice from appropriate persons.
  • Given the purpose of the FOI Act to encourage greater openness in public administration, it would be undesirable, in effect, to require the Attorney-General to reveal nothing at all about procedures that had been followed.

Hayne J, also in a separate judgment, agreed with the position taken by Gleeson CJ, Gummow, Heydon and Kiefel JJ in relation to waiver (at [131]).

Upholding of the appeal against the ruling that there was no basis for the VCAT's view that the public interest required access

The VCAT's view that, under s 50(4) of the FOI Act, the public interest required access had been influenced by, among other things, what appeared to have been differences between several of the legal advices on whether the exercise of the prerogative of mercy should be supported. If it had been the case that the government had received other and materially different legal advice to the joint advice, then, depending on what those differences were, it is possible that this could have been a relevant consideration in deciding the requirements of the public interest under s 50(4). Among other things, the question could arise as to why the government may have been favouring the joint advice over the other advices.

In light of this, Gleeson CJ, Gummow, Heydon and Kiefel JJ regarded the failure of the members of the Court of Appeal to examine all of these different advices, before deciding to overturn the VCAT's decision in favour of access, as involving 'an error of principle in the exercise of a discretion' (at [57]). Accordingly, they upheld the appeal. They remitted the matter to the Court of Appeal to inspect the documents, and deal further with it in accordance with the High Court's reasons.

Kirby J supported this order as well, accepting that the Court of Appeal was in error in failing to look at the documents (at [116] and [124]). He went further, though, and, among other things, ruled that Bongiorno AJA (with whom Ashley JA agreed on this point) in the Court of Appeal had fallen into error in holding that the secrecy traditionally accorded the consideration of petitions for mercy, of itself, opposed any public interest under s 50(4) in granting access to the documents. As Kirby J saw it, this involved taking into account an irrelevant consideration. His Honour viewed this idea of secrecy as deriving from outmoded notions such as the 'unexaminable prerogative of the Crown'. He said that such secrecy could not be a relevant consideration under s 50(4) as it was inconsistent with 'the language and scheme' of the FOI Act (at [120] and [121]).

By contrast, Hayne J took the view that any differences between the legal advices, suggested by the VCAT, did not bear upon the consideration of the public interest for the purposes of s 50(4) (at [154]–[156]). Consequently, he saw no need for the Court of Appeal to have examined the documents. This put him in dissent with the rest of his colleagues and caused him to dismiss the appeal.

Other issues

One feature of the High Court decision worth noting is that Kirby J, in addressing legal professional privilege as a ground of exemption from release under the FOI Act, observed (at [89]):

The ambit of legal professional privilege needs to be defined in the proper context. The privilege referred to in s. 32 of the FOI Act is necessarily that of a governmental party. At least in the case of a minister, it concerns documents of a kind to which the FOI Act is intended to be applicable, unless such documents are 'exempt'. It would be a mistake to assume that all communications with government lawyers, no matter what their origins, purpose and subject matter, fall within the ambit of the State's legal professional privilege. Advice taken from lawyers on issues of law reform and public policy does not necessarily attract the privilege. Especially in the context of the FOI Act and legal advice to government, courts need to be on their guard against any inclination of lawyers to expand the ambit of legal professional privilege beyond what is necessary and justifiable to fulfil its legal purposes.

Text of the decision is available at:


In Hearne v Street [2008] HCA 36 (6 August 2008) the High Court dealt with aspects of an implied undertaking to the court. The decision shows the important role played by an implied undertaking in the conduct of litigation, and the binding effect of the undertaking beyond the parties themselves to other persons in some way involved with the parties in the conduct of the litigation.

Hearne v Street
High Court of Australia, 6 August 2008
[2008] HCA 36; (2008) 248 ALR 609


Two persons who were residents near Luna Park on Sydney Harbour foreshore sued in the NSW Supreme Court the operator of the park and the company which was the operator's largest shareholder. The plaintiffs' claim was in the tort of nuisance for alleged loud noise from amusement activity at the park. They claimed injunctive relief restraining the continuation of the alleged high noise levels.

While the proceedings were in their preliminary stages, directions were given for the filing of affidavits and the exchange of expert reports. It was common ground that these were the subject of implied undertakings or obligations to the court imposed on the parties by law.

Two persons–one the managing director of the park's operator and the other one who was active in certain of the operator's affairs on behalf of interests which controlled a major shareholding in the operator–disclosed documents covered by the implied undertaking or obligation to the then NSW Minister for Sport, Tourism and Recreation and her staff. The objective of this was to have the NSW parliament enact legislation to protect the operations of the park from claims for noise nuisance. As it turned out, this objective was successful. Such legislation was enacted.

On account of this disclosure, the plaintiffs moved to have these two persons dealt with for contempt of court for breaching the implied undertaking or obligation applying to the documents.

At first instance

At first instance, it was held that the two persons were not personally bound by the undertaking or obligation.

Appeal to the NSW Court of Appeal

On appeal to the NSW Court of Appeal, this ruling was overturned by a majority (Ipp and Basten JJA, Handley AJA dissenting; see Street & Ors v Hearne & Anor [2007] NSWCA 113). The majority found that the persons had breached the undertaking, as the disclosure to the minister was for a purpose unconnected with the proceedings. This caused them to be found in contempt of court.

Handley AJA dissented on the ground that the appeal was not competent under the Supreme Court Act 1970 (NSW) (given that he saw the contempt motion as involving criminal, not civil, proceedings). He did note, though, in passing that, if competence had not been an issue, he would have supported the primary judge's ruling that the two persons were not bound by the undertaking. Also, there would have been, for him, a real question as to whether the disclosure of the documents to the minister for parliamentary purposes was a breach of the undertaking. He said that, if this question were to have been pursued, there would have been need for notice to be given to the NSW Attorney-General before further argument would have been heard on it. The majority did not give any considered attention to this question (though Basten JA acknowledged it at [138]–[141]). The two persons obtained special leave to appeal to the High Court.

Appeal to the High Court

The High Court unanimously dismissed the appeal, affirming the majority decision in the Court of Appeal.

Reasoning supporting the appellants being subject to the implied undertaking

In their joint judgment, Hayne, Heydon and Crennan JJ enunciated two propositions, which they saw 'as damaging to the appellants' arguments [against being bound by the undertaking]'. They stated these as follows (at [102]–[103]):

The first is that to call the obligation of the litigant who has received material generated by litigious processes one which arises from an 'implied undertaking' is misleading unless it is understood that in truth it is an obligation of law arising from circumstances in which the material was generated and received.

The second is that that obligation would be of very limited protection if it were only personal to the litigant, which is why it is often said to be extended also to a litigant's solicitor, industrial advocate or barrister, and also to third parties like a shorthand writer or court officer. For that reason the authorities recognise a broader principle by which persons who, knowing that material was generated in legal proceedings, use it for purposes other than those of the proceedings are in contempt of court.

In connection with this second proposition, they said (at [109]):

The implied undertaking also binds others to whom documents and information are given. For example, expert witnesses, who are not parties, commonly receive such documents and information and are bound by the obligation. It is likely that, in the future, documents and information will be provided to persons funding litigation, who will likewise be bound by the obligation.

Hayne, Heydon and Crennan JJ went on to reject the proposition that there needed to be specific knowledge of the undertaking on the part of those bound by it, saying (at [112]):

There is no support in the authorities for the idea that knowledge of anything more than the origins of the material in legal proceedings need be established. In particular, there is no support for the idea that knowledge of the 'implied undertaking' and its consequences should be proved, for that would be to require proof of knowledge of the law, and generally ignorance of the law does not prevent liability arising.

The three Justices saw this position as supported by the following considerations (at [120]):

The fact is that because in reality the 'implied undertaking' is an obligation imposed as a matter of law, it would be very hard to prove knowledge of that matter of law against lay persons. The narrower the avenue of liability against third parties, the weaker the incentive for litigants to give full discovery and to provide all relevant evidence. 'The interests of the proper administration of justice require that there should be no disincentive to full and frank discovery' (Riddick v Thames Board Mills Ltd [1977] QB 881 at 912 per Waller LJ)–or to full employment of all of the court's procedures directed to accurate fact finding in litigation.

Hayne, Heydon and Crennan JJ also agreed with the Court of Appeal majority that the appeal of the plaintiffs to that Court had been competent (as involving civil, not criminal, proceedings for the purposes of the Supreme Court Act 1970). Gleeson CJ in a short judgment generally agreed with Hayne, Heydon and Crennan JJ. Given the way the appellants had presented their case, Kirby J found himself constrained to dismiss the appeal on similar grounds to the other Justices.

Other issues

Kirby J was uncomfortable with several features of the presentation of the appellants' case in the High Court. Among other things, he saw the purpose of the disclosure to the minister as warranting serious consideration as to whether a breach of the implied undertaking had occurred. He noted certain parliamentary privilege issues raised by the circumstances (at [37]–[43]). He regretfully observed that the appellants, despite some prompting, had 'repeatedly declined to seek leave to enlarge the grounds of (and thus the issues in) the appeal' (at [45]).

Hayne, Heydon and Crennan JJ pointed out that 'it was not argued that the use of the [documents] to advance the cause of the defendants on the political front … was incapable of being contempt of court'. Likewise, they said that it was not argued that 'any principle of parliamentary privilege' prevented the disclosure from constituting a contempt (at [84] and [85]). Finally, they pointed out that no argument had taken place 'on the question of what exceptions to the rule forbidding disclosure exist–for example, in relation to the disclosure of criminal conduct' (at [86]).

Text of the decision is available at:


The extent of police powers of entry without warrant on private premises in circumstances of apprehended criminal activity is often a vexed subject. This is heightened in many situations by the urgency with which police must make judgments on appropriate responses to uncertain or quickly changing circumstances which are often fraught with complication. The litigation in Kuru v State of New South Wales [2008] HCA 26 (12 June 2008) illustrates the differing legal consequences that these judgments can have in exposure to civil liability in damages. The line between, on the one hand, what is lawful and free of consequence and, on the other, what is unlawful with significant damages implications often may be a very fine one.

Kuru v State of New South Wales
High Court of Australia, 12 June 2008
[2008] HCA 26; (2008) 246 ALR 260

Background to the litigation


On 15 June 2001, the plaintiff, a man in his 20s, and his partner (later his wife) attended a barbeque at the home of the plaintiff's sister. This home was a short distance from the flat which the plaintiff and his partner were occupying in a Sydney suburb. The plaintiff had consumed about six beers during the evening. After arriving back at the flat around midnight, a heated argument, involving yelling and screaming, developed between the plaintiff and his partner. There was no evidence that physical violence was involved.

The commotion caused someone in the immediate neighbourhood to contact the local police. The radio alert that issued to police patrol units in the area was designated as a 'violent domestic' rather than a 'normal domestic'.

The argument between the plaintiff and his partner soon abated. A further message seems to have been received by the police, probably from the same source as the first message, to say that the screaming had quietened down.

The plaintiff's sister and two other male persons came to visit the flat. The plaintiff's partner shortly afterwards left the flat with the plaintiff's sister. The front door was left slightly ajar. The plaintiff had a shower while the two male visitors remained in the living area of the flat.

Following the police alert, six police officers, four male and two female, arrived and entered the flat through the open front door. A male police officer spoke to one of the male persons visiting the flat, explaining that the police were attending because of a telephone call received about an argument occurring in the flat. The male person then went into the bathroom to fetch the plaintiff. When the plaintiff appeared from the bathroom, he asked the police what they were doing in the flat. A police officer repeated the explanation for their attendance. The plaintiff responded that there was no female present, but that the police could take a look around. On finding no female present, a police officer then asked the plaintiff the whereabouts of the female who had been there. The plaintiff responded that the female had gone to his sister's home nearby. The plaintiff then asked the police to leave.

The police asked for the address of the premises to which his partner had gone. The plaintiff tried to explain where the premises were, and then, in more exasperated tones, repeated his request for the police to leave. Pressed by the police, he wrote his recollection of the address of the premises on a piece of paper. He then repeated the request to leave. The police continued to question the plaintiff about the address. Shortly after, the plaintiff jumped onto the kitchen bench and demanded that the police leave. After this, the police claimed that the plaintiff jumped off the bench in a threatening manner, and, while gesticulating with his arms for the officers to leave, made contact with one of them. The plaintiff was immediately arrested for assault.

The plaintiff claimed that some five to eight minutes passed from when he first asked the police to leave to when he made contact with the officer.

The plaintiff claimed that he was punched several times by the police in the course of being arrested. The police claimed that they had to use force to restrain the plaintiff, who was attempting to fight them off. Capsicum spray was used and the plaintiff was handcuffed. He was taken to a police station and lodged in a cell wearing nothing but his boxer shorts. He was released from custody some hours later.

The plaintiff was charged with resisting a police officer in the execution of his duty, assault upon a police officer in the execution of his duty and assault occasioning actual bodily harm. These charges were later tried and dismissed by a magistrate.

The plaintiff instituted proceedings in the NSW District Court, claiming damages for trespass to land, trespass to the person, and false imprisonment against the State of New South Wales, the State bearing the liability for any wrongs done by the individual police officers.1

Powers under which the police purported to act

Important to an understanding of the plaintiff's claims for damages are the powers under which the police were purporting to act in entering and remaining at the plaintiff's flat.

Under s 357F(2) of Crimes Act 1900 (NSW), a member of the police force who believes on reasonable grounds that a domestic violence offence, among other things, has recently been committed in any dwelling-house may enter the dwelling-house and remain there for the purpose of investigating whether such an offence has been committed or preventing its commission. This is provided that the police officer is invited to do so by a person who apparently resides in the dwelling-house, whether or not the person is an adult. Under s 357(3), a member of the police force may not enter or remain in a dwelling-house by reason only of such an invitation if authority to enter or remain is expressly refused by an occupier of the dwelling-house and the member of the police force is not otherwise authorised (whether under the Crimes Act or any other Act, or at common law) to so enter or remain.

However, under s 357(4), a refusal of authority to enter or remain by an occupier will not prevail for the purposes of s 357(3) if the person who invited the police officer to enter or remain for the purposes of s 357(2) is one whom the member of the police force believes to be the victim of a domestic violence offence that has recently been committed in the dwelling-house.

Under s 357H(1) of the Crimes Act, among other things, where a member of the police force enters a dwelling-house following an invitation (as referred to in s 357F(2)) for the purpose of investigating whether a suspected domestic violence offence has been committed, or preventing its commission, the member of the police force (to quote paragraphs (a) and (b) of the subsection):

  1. is to take only such action in the dwelling-house as is reasonably necessary:
    1. to investigate whether such an offence has been committed,
    2. to render aid to any person who appears to be injured,
    3. to exercise any lawful power to arrest a person, and
    4. to prevent the commission or further commission of such an offence, and
  2. is to remain in the dwelling-house only as long as is reasonably necessary to take that action.

Subsection 357H(2) also provides that s 357F does not limit any other power which a member of the police force may have under the Crimes Act or any other Act, or at common law to enter or remain on premises.

There has been judicial support for the proposition that, at common law, where there are reasonable grounds of apprehension of a misdemeanour or breach of the peace occurring on private premises, a police officer is entitled to enter and remain on those premises (Thomas v Sawkins [1935] 2 KB 249 at 2542).

At first instance

The plaintiff succeeded in the District Court. The trial judge ruled that the police were not legally permitted to enter or remain in the flat. The trial judge was of the view that s 357F(2) did not apply because the police were not invited by the plaintiff to enter. The other male persons in the flat at the time had no authority to invite entry. The trial judge was further of the view that s 357H could only apply where a police officer enters premises pursuant to an invitation under s 357F(2). He rejected the argument that there was any common law right of entry supporting the police action in the circumstances. In addition, he refused to find that the plaintiff's contact with the police officer had involved any 'shoulder charging' as was alleged by the police. The trial judge ordered that the State was liable to the plaintiff in damages totalling $418,265, comprising:

  1. for trespass to property: $85,000
  2. for trespass to person, general damages: $150,000 plus out of pocket expenses of $8,265
  3. for false imprisonment: $20,000
  4. for aggravated damages:3 $35,000
  5. for exemplary damages:4 $120,000.

Appeal to the NSW Court of Appeal

The State successfully appealed against this decision to the Court of Appeal of New South Wales. The State essentially argued that:

  1. the entry of the police to the flat, and their continued presence, were permitted by s 357F(2) and, to the extent applicable, s 357H(1), or by the common law, or by a combination of these
  2. if liability in trespass was established, the damages awarded were excessive, and should not include aggravated and exemplary damages.

All members of the Court of Appeal were of the view that the plaintiff, by allowing the police to search the flat for a female person, effectively invited them onto the premises for the purpose of investigating whether a domestic violence offence had been committed. Accordingly, s 357(2) authorised their entry into the flat. In addition, the Court was of the view that the entry was permitted at common law, as there was properly an apprehension of a breach of the peace.

The Court of Appeal held that the plaintiff's revocation of consent did not immediately require that they depart the flat. The Court saw s 357H(1) as applying to enable the police to remain in the flat 'for a period reasonably necessary to take action of the kind they were required by s 357H(1)(a) to take'5 (i.e. principally to investigate whether such a domestic violence offence has been committed). Further, even though the police were directed to leave by the plaintiff, they were entitled at common law to stay until they had taken reasonable steps to satisfy themselves that no offence had been committed in the flat (whether by making a closer inspection of the premises or by telephoning the plaintiff's partner or his sister).6

This analysis led the Court of Appeal to a finding that the police were not liable in trespass when they arrested the plaintiff, and that his detention on arrest did not involve any false imprisonment. In these circumstances, the Court saw no need to address the alleged excessiveness of the damages awarded.

The judgment for damages in favour of the plaintiff in the District Court was set aside, and judgment was entered for the State dismissing all of the plaintiff's claims.

The plaintiff was granted special leave to appeal to the High Court against the Court of Appeal's orders.

Decision of the High Court

The High Court allowed the plaintiff's appeal by a 4 to 1 majority. However, it remitted the case to the Court of Appeal for further consideration of the grounds of appeal by the State to the Court of Appeal relating to the excessiveness of the original amount of damages. Four members of the High Court (Gleeson CJ, Gummow, Kirby and Hayne JJ) in a joint judgment criticised the Court of Appeal for not ruling on these grounds of appeal, saying (omitting references) (at [12]):

This Court has said on a number of occasions that, although there can be no universal rule, it is important for intermediate courts of appeal to consider whether to deal with all grounds of appeal, not just with what is identified as the decisive ground. If the intermediate court has dealt with all grounds argued and an appeal to this Court succeeds, this Court will be able to consider all the issues between the parties and will not have to remit the matter to the intermediate court for consideration of grounds of appeal not dealt with below.


The majority pointed out (at [26]) that:

[T]he distinction between entering and remaining is not marked by a bright line. Any entry upon premises necessarily constitutes remaining upon the premises for at least as long as the act of entering takes.

Section 357F recognises that an invitation to enter or remain in a dwelling-house may be revoked by an occupier of the premises. Once revocation occurred, the inquiry became what were the relevant provisions of the Crimes Act that would permit the police to remain on the premises (at [28]).

The majority disagreed with the Court of Appeal that the permission to enter and to remain, which was constituted by the invitation, persisted for so long as the purpose7 of the entry–that is, to investigate whether a domestic violence offence had been committed–remained unfulfilled. The majority saw the Court of Appeal's construction as precluded by s 357F(3). The majority said (at [31]):

[T]he express provisions of s.357F(3) require the conclusion that, unless sub-s.(4) was engaged, and that was not suggested here, an express refusal by an occupier immediately terminated the authority 'to so enter or remain'.

Further, the majority refused to countenance the Court of Appeal's view that s 357H(1) supported the continued presence of the police after the plaintiff revoked consent. They said that neither subs 357H(1)(a) or (b) could be read as granting a power to enter or a power to remain (at [36]). The powers conferred by s 357H(1) limited rather than expanded the power of entry under s 357F(2). The majority explained the underlying reasoning of their approach as follows (omitting references) (at [37]):

To the extent that, in the end, there was any ambiguity about the meaning and ambit of the authority provided to police by ss.357F and 357H to remain in the appellant's flat after he had made it clear that he was requiring them to leave, such ambiguity must be resolved in favour of the foregoing construction. This is because of the strong principle of Australian law defensive of the quiet enjoyment by an occupier of that person's residence. That principle has been recognised and upheld by this Court on numerous occasions. It derives from the principles of the common law of England. Indeed, it appears to be a principle against which the provisions of ss.357F and 357H of the Act were written. It defends an important civil right in our society. If Parliament were to deprive persons of such a right, or to diminish that right, conventional canons of statutory construction require that it must do so clearly.

The majority said that, while they were 'mindful of the difficulties of police in responding to apparent complaints about domestic violence', ss 357F, 357G (dealing with entry to premises by warrant) and 357H reserved the right to the occupier to withdraw an invitation to police to enter and remain on the premises. They said (at [38]):

If, in the present case, the police considered that it was necessary to re-enter the premises, the remedy was in their hands. They could seek a warrant from a magistrate, and this could be sought and provided by telephone.

The majority finally considered any common law protection for the police action. In this they were guided by the High Court's decision in Plenty v Dillon (1991) 171 CLR 635. They said (omitting references) (at [43]):

[I]t is necessary to approach questions of the kind now under consideration by recognising the importance of two related propositions. First, a person who enters the land of another must justify that entry by showing either that the entry was with the consent of the occupier or that the entrant had lawful authority to enter. Secondly, except in cases provided for by the common law and by statute, police officers have no special rights to enter land. And in the circumstances of this case it is also important to recognise a third proposition: that an authority to enter land may be revoked and that, if the authority is revoked, the entrant no longer has authority to remain on the land but must leave as soon as is reasonably practicable.

The majority was sceptical as to whether at common law there was a right of entry of a police officer to prevent a breach of the peace as broad as that claimed in the English Court of Criminal Appeal decision in Thomas v Sawkins [1935] 2 KB 249. Rather, the majority saw the position as follows (at [51]):

Whatever may be the ambit of the power of police (or of a member of the public) to enter premises to prevent a breach of the peace, that power of entry does not extend to entry for the purposes of investigating whether there has been a breach of the peace or determining whether one is threatened.

In the present case, by the time the police went to the flat, there was no continuing breach of the peace. Nothing in the evidence of what happened thereafter suggested that, but for the police officers not leaving the flat when asked to do so, any further breach of the peace was threatened or expected, let alone imminent (at [53]).
In dissent, Heydon J essentially adopted the same reasoning as the Court of Appeal. He said (at [63]):

Is there an absurdity or anomaly in the [State's] position arising from its contemplation that under the legislation consent to enter can be refused in the first place, and any entry therefore rendered unlawful, even if the police officer suspected a serious crime causing grave personal injury had just taken place, but withdrawal of consent to remain cannot render the continued presence of police already on the premises unlawful until the s.357H(1) actions are complete? … Police officers who have lawfully entered pursuant to consent, or who remain lawfully after entry pursuant to consent, are likely in practice to know much more about what has happened or is likely to happen than police officers who were not given any consent to enter or remain and did not enter or remain. The compromise struck by the legislation is that police officers who have been refused consent to enter at all must obtain a warrant. Those who have entered or remained pursuant to consent which is then withdrawn may remain until the s.357H(1) processes, in the course of which they may well have learned information which makes the completion of questioning those persons desirable, are complete. … [The plaintiff's submissions] would mean that the householder could forestall the lawful activities of police officers just as they were beginning to bear fruit.


  1. This approach reflects the application of ss 8 and 9 of the Law Reform (Vicarious Liability) Act 1983 (NSW) and s 5 of the Crown Proceedings Act 1988 (NSW), rendering it unnecessary to join as a defendant any police against whom a wrong is alleged.
  2. To like effect in Australian jurisdictions, see Dowling v Higgins [1944] Tas SR 32 at 34 per Morris CJ; Todd v O'Sullivan (1985) 122 LSJS 403 at 409 per Legoe J; Panos v Hayes (1987) 44 SASR 148 at 154 to 155 per Legoe J; Nicholson v Avon [1991] 1 VR 212 at 222 per Marks J.
  3. Aggravated damages are compensatory in nature, being awarded for injury to the plaintiff's feelings caused by insult, humiliation and the like: Lamb v Cotogno [1987] HCA 47; (1987) 164 CLR 1.
  4. Exemplary damages are not compensatory, being intended to punish the defendant for conduct showing a conscious and contumelious disregard for the plaintiff's rights and to deter the defendant from committing like conduct again: XL Petroleum (NSW) Pty Ltd v Caltex Oil (Australia) Pty Ltd [1985] HCA 12; (1985) 155 CLR 448.
  5. See [2007] NSWCA 141, at [167], per Ipp JA.
  6. See [2007] NSWCA 141, at [178], per Ipp JA.
  7. This must be a 'purpose' identified in s 357F(2).

Text of the decision is available at:


In the High Court decision in Roads and Traffic Authority v Royal [2008] HCA 19 (Royal) the Roads and Traffic Roads Authority of New South Wales (the RTA) avoided liability for the design and construction of an intersection because any failure by it in that task was not a contributing cause to the motor accident injuring the plaintiff.

Roads and Traffic Authority v Royal
High Court of Australia, 14 May 2008
[2008] HCA 19; (2008) 245 ALR 653


The design and construction of a road can be a causative factor in some road accidents. Where an allegation is made to this effect, it will often be a difficult assessment whether the road design or construction is a relevant or major cause, as against the more immediate acts or omissions of the driver or drivers involved or the mechanical state of their motor vehicles.

The decision in Royal points to a plaintiff needing compelling evidence to establish a causal connection between defective road design or construction and the subject motor accident. The decision is consistent with the Court's earlier decision in Commissioner of Main Roads (WA) v Jones [2005] HCA 27; (2005) 215 ALR 418, where the Court denied a causal connection between an alleged failure to erect a sign warning of the risk of wild animals straying onto the road and a collision between the vehicle driven by the plaintiff and a wild horse that unexpectedly bolted out from the roadside into the vehicle's path.

At a deeper level, Royal is a powerful illustration of a tension that comes through in some recent High Court decisions involving negligence claims between, on the one hand, the pursuit of the correct application of legal principle through the appeal process, as against, on the other, frustration with the delay and expense involved, particularly where the final outcomes do not differ much from those in the original trial. A hint of this tension emerges in following comments (omitting references) about this case made by Kirby J early in his judgment (at [39]):

The spectacle of five Justices of this Court labouring over highway plans and photographs and sifting through four appeal books in relation to such a question [viz. what is the cause or causes of a motor accident] would be bound to cause surprise. The record describes the 12 day trial of these proceedings in the District Court of New South Wales, and the two day hearing in the Court of Appeal. What is, and is not, for legal purposes, a material cause of a motor vehicle collision is a question of fact. Ordinarily, it gives rise to no principle of law, binding on lower courts and future parties. On the face of things, it concerns only the immediate parties and the outcome of their dispute.

Whether the other Justices of the Court who participated in the appeal shared these perceptions may be debatable.

Background to the litigation


On 12 March 2001, the plaintiff was driving his motor vehicle along a road near Wauchope in northern coastal New South Wales that approached an intersection with the Pacific Highway from the west. His object was to cross the highway at the intersection and proceed in an easterly direction along the road that intersected immediately opposite, from the east (passing through the break in the highway's median strip). The intersection was located at the crest of a hill. The northbound lanes of the highway turned a curve from the east approaching the intersection. That curve continued through the intersection and down the other side of the hill. In addition to the two northbound lanes of the highway, there were northbound turning lanes which opened up shortly before the intersection, one on the far left and the other on the far right, to carry traffic wishing to turn off the highway at the intersection into the relevant intersecting road.

For a driver waiting by the stop sign to cross the intersection from the road from the west, vision of the approaching traffic from the south, with the curvature of the highway's northbound lanes, was limited to about 250 metres. (The stop sign would be moved slightly forward to the corner of the intersection in consequence of the plaintiff's accident.) This limitation of vision was alleged to be one shortcoming in the way that the northbound lanes, as they approached the intersection, had been designed and constructed. The maximum speed on the Pacific Highway in the general vicinity was 100 kph. There was a sign on the Pacific Highway 300 metres to the south warning of the approaching intersection and recommending a speed of 85 kph.

The RTA constructed the intersection in 1993 (though the northbound lane for traffic turning off to the right was only added in 1997). From its construction in 1993, prior to 12 March 2001, there had been 26 reported accidents at the intersection which had resulted in personal injury or damage requiring one or more of the motor vehicles involved to be towed away. Seventeen accidents involved traffic crossing at the intersection. Overall, these accidents had accounted for four deaths.

According to a witness driving a motor vehicle directly behind the plaintiff, when the plaintiff arrived at the intersection, he came to a standstill beside the stop sign. There were four cars approaching from the south on the highway. Two vehicles, moving into the far left lane, were turning into the road out of which the plaintiff was emerging. Further away was a third vehicle approaching the intersection in the right northbound lane, followed by a fourth. There was evidence that the third vehicle was veering over into the lane that was there for traffic turning right at the intersection into the road intersecting from the east. This suggested that this third vehicle was going to turn in that direction.

The plaintiff moved out in front of the third vehicle, perhaps expecting that this vehicle, from its position over in the right hand turning lane, would shortly be slowing down to turn right. However, the third vehicle, which was travelling on cruise control set at 105 kph, did not slow down to turn right. In continuing northwards, it failed to avoid the plaintiff's vehicle crossing in front of it, colliding with the plaintiff's vehicle at its right-side centre. The impact seriously injured the plaintiff. These injuries denied him any memory of the collision.

At first instance

The plaintiff sued the driver of the third vehicle in the NSW District Court, claiming damages in negligence for his injuries. That driver, who became the first defendant in the proceedings, denied negligence, and, in the event that he was found to be negligent, claimed that the plaintiff had been guilty of contributory negligence in failing to wait until the first defendant had passed the intersection before crossing. The first defendant also cross-claimed against the RTA, seeking contribution from it. The first defendant alleged that the RTA was negligent in the design and construction of the intersection, and that this negligence had been the cause of the plaintiff's injuries. The plaintiff subsequently claimed against the RTA, it becoming the second defendant in the proceedings, alleging against it the same negligence as had the first defendant in his cross-claim.

The trial judge gave judgment in favour of the plaintiff against the first defendant in damages in the sum of $1.3 million, but reduced that by one-third to $871,019.50 on account of the contributory negligence of the plaintiff. The trial judge exonerated the RTA of any liability, dismissing the claims against it of the plaintiff and of the first defendant.

Appeal to the NSW Court of Appeal

The first defendant appealed to the NSW Court of Appeal, primarily challenging the dismissal of his cross-claim against the RTA, arguing that the contribution of the RTA's negligence should be apportioned at 80% and his negligence at 20%.

The appeal succeeded to the extent that a majority of the Court of Appeal upheld the challenge to the dismissal of the cross-claim against the RTA. The majority, Santow and Tobias JJA (the latter agreeing with the former's judgment), referred to the RTA as having a duty of care to take reasonable steps to alleviate what was a known danger at a specific location. There were specific options available for seeking to remedy the danger. The RTA breached its duty by failing to take steps that would have been reasonable in this regard; namely, not just to shift the stop sign forward (as it did after the accident) but, more significantly, to have constructed a staggered T-intersection rather than a cross-intersection. (What was being referred to here was a configuration pursuant to which drivers, approaching on the road from the west wishing to cross the highway and exit on the road from the east, would make a left-hand turn into the left northbound lane of the highway, shift across to the right, and make a right-hand turn some way to the north so as to reach an opening through the median strip accessing the highway's southbound lanes.)

Important to the subsequent High Court appeal was the majority's view that the negligence of the first defendant in driving his motor vehicle through the intersection did not break the chain of causation between the RTA's original negligent design and construction of the intersection and the collision. A break would have precluded any liability on the RTA's part for the plaintiff's injuries. The minority judge, Basten JA, was of the view that the question of breach of duty did not arise until one identified the cause of the collision. Whatever were the failings of the intersection's construction, he did not see these as materially contributing to the cause of the collision.

The majority of the Court of Appeal concluded that the liability should be apportioned as to two-thirds against the first defendant and one-third against the RTA. (The contributory negligence finding against the plaintiff at first instance remained unaffected by the outcome before the Court of Appeal.)

The RTA obtained special leave to appeal to the High Court against the Court of Appeal's decision, but only as to whether there was any chain of causation between the breach of duty of care determined by the Court of Appeal and the plaintiff's injuries.

Decision of the High Court

The RTA's appeal was upheld by a 4 to 1 majority. This restored the position to the outcome attained after the trial, in which the plaintiff obtained a damages verdict against the first defendant, reduced by one-third for contributory negligence.


Of the majority, Gummow, Hayne and Heydon JJ, in their joint judgment (the joint majority), firmly rejected the design and construction of the intersection as a material cause of the collision, saying (at [26]):

[T]here was no evidence that any aspect of the plaintiff's decision, having stopped at the intersection, to move forward was caused by the fact that the defendant's vehicle was masked by some other vehicle. ... Hence to submit, as the defendant did, that the [RTA's] breach of duty 'restricted the [plaintiff's] view of the intersection' and created 'problems of vision' for him may have been correct for some sets of circumstances, but was not correct for the circumstances preceding the collision in question in this appeal.

The joint majority went on to add (at [29]):

If the plaintiff failed to see the defendant, that could have been one causal factor in the collision. … A further causal factor was the defendant's failure to … use the ample time available to take steps to deactivate cruise control, slow down, stop or change lanes or otherwise avoid hitting the plaintiff's vehicle, when that vehicle was apparently doing nothing to avoid a collision … Another causal factor was the potentially misleading effect on the plaintiff of the defendant being in the right-hand turn lane rather than one of the through lanes.

The joint majority criticised, from several standpoints, the approach of the NSW Court of Appeal majority in deciding that the negligence of the drivers did not constitute a break in the chain of causation between the RTA's negligence in designing the intersection and the plaintiff's injuries. Two of these criticisms bear mention.

First, the joint majority criticised the Court of Appeal majority's affixing of liability to the RTA for the design and construction of the intersection, holding that it involved an incorrect application of the 'but for' test of causation. The 'but for' test basically postulates that, if event X would not have occurred but for event Y, then Y is the cause, but, if X would have occurred regardless of Y, Y is not a cause. The 'but for' test has been rejected as the exclusive test of causation under the law of negligence, though it is nonetheless a useful test to exclude, in the first instance, what is not a cause. Instead, the law of negligence follows a 'common sense' test of whether event Y materially contributed to event X. These principles on causation were confirmed by the High Court in a motor accident context in March v Stramare (E & M H) Pty Ltd (1991) 171 CLR 506; [1991] HCA 12 (March). In March, different members of the Court pointed to the shortcomings of the 'but for' test. For instance, Deane J said (at 123):

Thus, it could not, as a matter of ordinary language, be said that the fact that a person had a head was a 'cause' of his being decapitated by a negligently wielded sword notwithstanding that possession of a head is an essential precondition of decapitation.

In a similar vein, Mason CJ said (at 512):

[I]n the nature of things, there will be some cases in which a court concludes that a precondition does not play such a part in the consequence that it deserves to be characterized as a cause.

The joint majority quoted this passage of Mason CJ, and said that the liability which the Court of Appeal majority had sought to sheet home to the RTA for the intersection's design and construction was but an example of what Mason CJ was addressing in this passage (at [32]). While the design and construction was a precondition to the collision, it did not play such a part as to be properly characterised as a cause.

The second criticism made by the joint majority of the Court of Appeal majority emanated from the latter's reliance ([2007] NSWCA 76 at [96]) upon the following passage of Gaudron J's judgment in Bennett v Minister for Community Welfare (1992) 176 CLR 408 at 421:

… the question whether some supervening event broke a chain of causation which began with or which relates back to an omission or a failure to perform a positive duty, is one that can only be answered by having regard to what would or would not have happened if the duty had been performed. It is only by undertaking that exercise that it is possible to say whether the breach was 'still operating' or continued to be causally significant when the harm was suffered.

Applying that test here, the Court of Appeal majority had said that, assuming the duty was performed, one was required to hypothesise that a staggered T-intersection had been designed and constructed rather than the cross-intersection ([2007] NSWCA 76 at [97]). If the accident would have happened anyway, the RTA would be exonerated of any liability.

The joint majority saw this as involving a flawed analysis of the facts, saying (at [33]):

In the first place, Gaudron J's reasoning proceeds on the assumption that a chain of causation has been established: that assumption is not made out here. In the second place, it is no doubt true that if there had been a staggered T-intersection the plaintiff would not have been trying to negotiate a cross-intersection and would not have been injured doing so. But to say that is only to say that there would not have been a cross-intersection collision if there had not been a cross-intersection. It does not say that there would not have been a collision between drivers as careless as the defendant and the plaintiff as the plaintiff came onto the Pacific Highway in the left-hand lane and began to move over to the right-hand lane to execute a right-hand turn in order to get to Boyds Road [i.e. the intersecting road from the east].

Kiefel J, in a separate judgment, agreed with the orders made by the joint majority. Her Honour stated her position in the following succinct terms (at [145]):

The evidence did not show that the design of the intersection contributed to the accident. It is not sufficient to suggest that there was a statistical possibility of an accident at the intersection because it was not the best design. To hold the RTA liable on this account would be to impose something approaching absolute liability. The accident was caused by driver error.

Kirby J in dissent saw the Court of Appeal majority as having correctly identified the design and construction of the intersection as giving rise to a statistical inevitability of a proportion of collisions involving vehicles crossing the intersection. This was demonstrated by the accident statistics which had emerged since the intersection's construction. Kirby J agreed with the Court of Appeal majority that, while the design and construction did not render the collision inevitable, it did materially contribute to its occurrence by creating a heightened risk of such an accident. In that sense, it was properly adjudged by the majority to have been a cause of the collision under the established principles of law on causation in negligence (at [99] and [100]).

Interestingly, Kirby J seemed to call in aid, in support of the RTA's liability in the present case, a broader jurisprudential reason. As well as it being a function of the common law of tort to provide compensation, it was also a function of that law to provide monetary sanctions against errant conduct of officials and, by inference, public authorities like the RTA. While this is a view that may not command wide judicial assent, it is one worth noting. Kirby J said (omitting references) (at [114]):

The law of actionable civil wrongs exists not only to provide monetary compensation (and contribution) where that is justified, but also to encourage appropriate conduct (including on the part of public officials) by the imposition of appropriate monetary sanctions. I realise, of course, the imperfections, inefficiencies and paradoxes involved in treating the law of torts as a guardian of communal fairness and as a stimulus to accident prevention. Doubtless, there are other, usually legislative, means of attaining these ends. However, so long as the law of torts survives, its role in distributive justice and in promoting safety should be maintained rather than denied.

Text of the decision is available at:


The exercise of reasonable care demanded of public authorities under the law of negligence in guarding against harm from dangerous recreational activity in public places is not so onerous as to require the prevention of conduct carrying the risk of harm. Rather, it requires the undertaking of such precautions as are a reasonable response to the foreseeable risks of harm.

Roads and Traffic Authority of NSW v Dederer
High Court of Australia, 30 August 2007
[2007] HCA 42; (2007) 238 ALR 761


The liability exposure of public authorities for injury suffered in recreational activity in public places, including roadways, under their control has again occupied the High Court. In Roads and Traffic Authority of NSW v Dederer [2007] HCA 42 (Dederer) the High Court wrestled with the question of whether there had been a breach of the duty of care by a main roads authority, the Roads and Traffic Authority of New South Wales (RTA).

This required the proper application of the principles going to the exercise of reasonable care formulated in the Court's 1980 decision in Wyong Shire Council v Shirt (1980) 146 CLR 40 (Shirt). These principles have often been referred to generically as the 'Shirt calculus' (though two members of the High Court have recently deprecated use of this term because it suggests that it involves a calculation rather than a judgment1).

The litigation in Dederer again demonstrates the difficulties which can emerge in the area of breach of duty of care in assessing what is a question of law as opposed to one of fact, and how that assessment can affect the outcome of any appeal.

The litigation also raises a more fundamental consideration of how far rights of appeal should extend, particularly where a perception emerges that well resourced public authorities or corporations are able to press these rights more easily than individual persons of limited means, with the result sometimes, as was the case here, that the individual sees his or her initial successes in the courts whittled away as the litigation progresses on appeal.

Background to the litigation


On 31 December 1998, the plaintiff, a 14-year-old male swimmer, suffered partial paraplegia when he struck his head on the bed of the Wallamba River after diving from a road bridge. The bridge crossed the river's estuary, between the towns of Tuncurry and Forster on the mid-north coast of New South Wales. The bridge had been built in 1959 by the predecessor agency of the RTA, and was controlled and maintained by the RTA.

On each approach to the bridge was a pictograph notice prohibiting diving from the bridge and a notice in words prohibiting climbing on it. These had been erected by the local council, the Great Lakes Shire Council (the Council) in 1995. The Council had done this with funding which the RTA had made available to it (along with other municipal councils) to provide maintenance on main roads. Generally, maintenance of the bridge, and the control of traffic on it, were the responsibilities of the RTA, with the Council being responsible for routine work on it such as sweeping the kerbs and gutters and cleaning the drainage holes.

The depth of the river in the vicinity of the bridge was subject to tidal variation. Also, the river in this area experienced currents which could produce sand movements on its bed.

According to the evidence, despite the prohibition notices, swimmers regularly dived from the bridge into the river. There had been no accidents like that suffered by the plaintiff there before. The plaintiff had swum in the area on previous occasions and had seen other swimmers diving from the bridge. Only the day before, he had twice dived from the bridge without mishap.


On 3 April 2002, the plaintiff sued the RTA in the NSW Supreme Court for damages for his injury. On 20 August 2002, the RTA filed a defence in which it admitted that it was 'responsible for' the bridge and the notices 'positioned at each end and along the length of the bridge'.

The trial was to commence on 8 September 2003. In the preceding week, the RTA informed the plaintiff that it was applying for an adjournment of the trial to enable it to join the Council as a cross-defendant. This was the first indication that the RTA believed that the Council had any responsibility for the plaintiff's injury. The trial was adjourned and an amendment made to the RTA's defence that the Council had joint responsibility for the bridge. In response to this, the plaintiff decided that he should claim damages against the Council as well as the RTA. He did so on 5 September 2003.

This claim of the plaintiff against the Council was subject to changes to the law of negligence made by the Civil Liability Amendment (Personal Responsibility) Act 2002 (NSW) (the CLAPR Act). These applied only to claims commenced on or after 6 December 2002, the date on which this Act took effect. Of principal significance, s 5L of the CLAPR Act excluded liability in negligence for harm suffered from obvious risks of dangerous recreational activities. If the claim against the Council had been instituted prior to 6 December 2002, it would not have been subject to the significant hurdle to its success posed by s 5L.

Further changes to the constitution of the proceedings occurred when, on 14 July 2004, the RTA filed a cross-claim against the Council. Changes to the pleadings followed. On 7 October 2004, the RTA effectively changed its previous defence, alleging instead that the Council, rather than it, had 'immediate control', and the 'day-to-day management and supervision' of the pedestrian way and the 'No Diving' pictograph signs on the bridge.


At first instance, the trial judge ruled that the RTA and the Council were the joint occupiers of the bridge, and that each had been negligent in that, knowing the prohibition notices had been ineffectual in stopping swimmers from using the bridge as a diving platform, neither had adopted more stringent measures to curb this use.2 Liability was apportioned at 80% for the RTA and 20% for the Council. The trial judge reduced the damages agreed between the parties by 25% on account of the plaintiff's contributory negligence in disregarding the prohibition notices.

Appeal to NSW Court of Appeal

The RTA and the Council each appealed to the NSW Court of Appeal on grounds including a challenge against the negligence and contributory negligence findings.

The Court of Appeal unanimously upheld the appeal of the Council.3 Applying s 5L of the CLAPR Act, the Court was of the view that a person of the plaintiff's age and with his knowledge of the area would have realised the obvious risk of the danger in diving from the bridge, even without the pictograph notice prohibiting diving.

On the other hand, a majority of the Court dismissed the RTA's challenge to the negligence finding against it. The majority viewed the RTA, as successor of the agency that built the bridge, as bearing responsibility for any danger that the bridge itself posed. The RTA exercised control over the structure of the bridge, its surface and its maintenance, and had undertaken protections for the safety of pedestrians and vehicles on the bridge. These factors were foremost in imposing a duty of care on the RTA to users of the bridge, including the plaintiff.

The majority agreed with the trial judge that the prohibition notices were an inadequate response to the dangers of diving from the bridge. Further, the majority saw the combined effect of the following factors as leading to a situation in which the plaintiff would probably not have dived off the bridge:

  • the presence of a notice more strongly emphasising the danger
  • the presence of a modified flat-top railing
  • the removal of the horizontal railing and replacement by vertical railings.

On this account, the majority were satisfied that the RTA breached its duty of care. This breach caused the unfortunate dive of the plaintiff.

The majority did allow the RTA appeal, to the extent of increasing the measure of the plaintiff's contributory negligence from 25% to 50%.

As a result of the upholding of the Council's appeal, the plaintiff submitted that he should not have to pay the Council's costs of the proceeding but, instead, those costs should be paid by the RTA to the Council. This type of costs order, often called a 'Sanderson order',4 (one by which an unsuccessful defendant is ordered to pay the costs of the successful defendant directly to that defendant) normally requires that there be some unjustifiable conduct on the part of the unsuccessful defendant that makes it fair that the unsuccessful defendant should be ordered itself to pay those costs direct to the successful defendant. This is as opposed to an order, often called a Bullock order,5 under which the successful defendant would obtain a costs order against the plaintiff who was successful against the unsuccessful defendant, with that plaintiff being able to recover those costs from the unsuccessful defendant in addition to its own costs.

The plaintiff argued that he made a claim against the Council only after the RTA had changed its position to assert that the Council had immediate control of the pedestrian area of the bridge and the warning signs. Had the RTA not taken this course, the plaintiff would not have claimed against the Council at all. While the Court of Appeal saw some force in this argument, it refused to make a Sanderson order, ruling that the plaintiff's claim against the Council was still a course he chose to take in the circumstances, free from necessity to do so notwithstanding the inconsistent approach of the RTA in the conduct of its defence.6

Appeal to the High Court

The RTA was granted special leave to appeal to the High Court against the Court of Appeal's decision dismissing its appeal. At the hearing of the appeal to the High Court, the plaintiff applied for leave to cross-appeal against the Court of Appeal finding of contributory negligence and its refusal of a Sanderson order.

Decision of the High Court

The High Court allowed the RTA's appeal by a majority of 3 to 2 and, by the same majority, dismissed the plaintiff's application for special leave to cross-appeal. The final outcome was that the plaintiff's damages claim failed in its entirety, with him bearing costs orders in favour of the RTA and the Council for the trial and the appeals.


Breach of duty of care

The majority (Gummow, Callinan and Heydon JJ) ruled that there was no breach of the RTA's duty of care to the plaintiff in the circumstances of his injury. In his judgment, with which Heydon J generally agreed (at [295]), Gummow J referred to the duty of care borne by the RTA as a road authority. Gummow J saw this formulated in the High Court's decision in 2001 in Brodie v Singleton Shire Council (2001) 206 CLR 512 as follows (at [45]):

… a road authority is obliged to exercise reasonable care so that the road is safe 'for users exercising reasonable care for their own safety'.

Gummow J said that this was not a duty 'to exercise reasonable care in the abstract', nor one 'to ensure that a road be safe in all the circumstances' (at [46]). The RTA did not owe a more stringent obligation towards careless road users as compared with careful ones (at [47]).

The exercise of reasonable care was not to be equated with an obligation to prevent harm occurring to others (at [51]). The latter was more onerous. Gummow J saw the trial judge and the Court of Appeal majority as having fallen into error here in that each had focused on the failure of the 'No Diving' pictograms and 'No Climbing' notices to prevent diving or jumping from the bridge (at [53] and [54]). Gummow J said (at [54]):

If the RTA exercised reasonable care, it would not be liable even if the risk-taking conduct continued. If the contrary were true, then defendants would be liable in any case in which a plaintiff ignored a warning or prohibition sign and engaged in the conduct the subject of the warning.

The material question here for Gummow J was the reasonableness of the warning, not its failure (at [56]).

Even if the trial judge and the Court of Appeal had ruled that the RTA's obligation extended only to the exercise of reasonable care, Gummow J said that they needed to identify accurately the actual risk of injury. The Court of Appeal majority saw the risk as that of spinal injury through diving off the bridge, being created by the RTA predecessor's erection of the bridge. However, according to Gummow J, the true source of potential injury arose, not from the state of the bridge itself, but, rather, from the risk of injury due to diving from the bridge into potentially shallow water. This led to two errors: first, the Court of Appeal majority had failed to make a proper evaluation of the probability of that risk occurring; and, secondly, they had attributed to the RTA a greater control over the risk than it possessed (at [60]).

Gummow J finally addressed the question of whether, applying the correct principles as he had explained them, there had been any breach of duty of care by the RTA (at [65]–[79]). This involved reference to the judgment of Mason J in Shirt, applying the Shirt principles for assessing what response a reasonable person would have made to the risk. This requires a consideration of:

  • the magnitude of the risk
  • the degree of probability of its occurrence
  • the expense, difficulty and inconvenience of taking alleviating action
  • any other conflicting responsibilities which the defendant may have.

Gummow J referred to the judgment of Hayne J in Vairy v Wyong Shire Council [2005] HCA 62; (2005) 223 CLR 422 (Vairy) where Hayne J pointed out that, in applying the Shirt principles, one must look forward from the position of the defendant before the injury occurred in determining what response was reasonably required of the defendant to meet the risk that led to the plaintiff's injury. The assessment must not start at the point that the injury occurred and look back to how the injury might have been avoided (Vairy at [126]).

Applying the Shirt principles, Gummow J expressed his conclusions as follows (at [78] and [79]):

Though grave, the risk faced by [the plaintiff] was of a very low probability, and a reasonable response to that risk did not demand the measures suggested by him. Those measures lacked evidential support; were of doubtful utility; would have caused significant expense in the case of the modifications to the handrail and fencing; and were in some cases contrary to express findings of fact.

This was not a case in which the defendant had done nothing in response to a foreseeable risk. To the contrary, the RTA had erected signs warning of, and prohibiting, the very conduct engaged in by [the plaintiff]. In the circumstances, that was a reasonable response, … and the law demands no more and no less.

Callinan J wrote a separate judgment. He agreed with Gummow J's criticisms of the courts below in relation to their assessments of breach of duty (at [270]). He said (at [276]):

A defendant is not an insurer. Defendants are not under absolute duties to prevent injury, or indeed even to take all such measures as might make it less likely to occur. They are obliged only to make such responses as can be seen to be reasonable in the circumstances. A proper balancing exercise which takes all of the relevant circumstances into account leads inescapably to the conclusion that the appellant, in responding to a risk that had not been realized for 40 years, by erecting the pictograph signs, acted reasonably and adequately.

In dissent, Kirby J wrote a substantial judgment, generally supported by Gleeson CJ in a shorter judgment. Kirby J referred to the Shirt principles (at [136]) and said that there had been no misapplication of it by the Court of Appeal majority (at [139]). He said (at [152]):

The RTA did not give any, or any reasonable, consideration to the fact that because of its position, construction and configuration in relation to the water below, the bridge presented special dangers, particularly to children and young persons.

Kirby J agreed that the steps seen as needed by the Court of Appeal majority should have been taken, noting that the signs themselves were useless. He saw this as not a case of seeking to reconstruct events by looking back in hindsight. He agreed with the Court of Appeal majority that the plaintiff's injury from the dive was an accident 'waiting to happen'.7

Contributory negligence

Kirby J (with Gleeson CJ concurring) would have declined special leave to the plaintiff to challenge the increase of contributory negligence to 50%. Kirby J hinted that, if the appeal to the High Court had been 'a full appeal by way of rehearing', he might have formed the view that the trial judge's ruling of 25% was appropriate. However, he said that '[t]he principles of restraint which limit intermediate courts' interference in apportionments of this kind are even more clearly applicable to any disturbance by this Court of the new apportionment unanimously arrived at by the Court of Appeal' (at [175]).

Sanderson order for costs

Kirby J (again with Gleeson CJ concurring) would have upheld the plaintiff's application for special leave to appeal against the refusal of a Sanderson order against the RTA. Kirby J saw 'the switch of tactics' by the RTA in October 2004 in effectively changing its previous defence, alleging that the Council had control of the pedestrian way and the prohibition notices on the bridge, as forcing the plaintiff into suing the Council needlessly. The ensuing unfairness of this justified a Sanderson order (at [192]).

Model litigant failure

Heydon J, although a member of the majority in upholding the RTA's appeal, nonetheless felt moved to criticise this change of position by the RTA well into the course of the proceedings, denouncing it as a failure of the RTA's model litigant obligation as a public authority. He said (at [298]):

It is a truism that statutory bodies of that kind should be model litigants: counsel for the RTA accepted that this was so 'without question'. A terrible thing had happened to a child. The solicitors for that child were not busybodies. Their request of the RTA was not a trivial one. It was possible that the RTA–a very wealthy and powerful organisation–was liable in tort. It was also possible that the Council–doubtless much less wealthy, but better resourced than the plaintiff and his parents–was liable. There is nothing wrong with wealthy and powerful defendants requiring plaintiffs to prove their cases, but in the circumstances, as a matter of common humanity, not legal duty, the RTA ought not only to have attempted to tell the plaintiff's advisers who controlled the bridge, as it did, but also to have stated the underlying facts correctly.

Challenge in final appeal to concurrent findings of fact in the courts below

An issue in the case that drew some comment from several members of the High Court was the asserted undesirability of an appellant before the High Court attempting to challenge findings of fact before that Court which had been upheld at first instance and again in the appeal court below. Gleeson CJ and Kirby J described this situation as one involving 'concurrent findings' of fact. They each saw the decisions at those levels as turning upon findings of primary fact, some of which were disputed but available on the evidence, with the appellant's argument on that evidence having been considered and rejected. Gleeson CJ said (at [5]):

In an appeal of this nature, the function of this Court, as a second appellate court and a court of final resort, is not simply to give a well-resourced litigant a third opportunity to persuade a tribunal to take a view of the facts favourable to that litigant. 'It is well settled that a second appellate court, such as this Court is in the present case, should not, in the absence of special reasons such as plain injustice or clear error, disturb such concurrent findings (Louth v Diprose (1992) 175 CLR 621, at 634, per Deane J). This is a principle of long standing, and its importance has not been diminished, but rather has been increased, in the circumstances of modern litigation.

Kirby J did acknowledge the merit of the rationale underlying this principle, but conceded that it cannot prevail against 'the statutory and constitutional functions' that the High Court possesses (at [163]). Callinan J and Heydon J were more sceptical about the principle, Callinan J drawing on the role of the High Court under the Constitution as the final court of appeal (at [265]–[267] and [287]–[294]. He said (at [267]):

The task of an appellate court is not to deny any litigant, whether rich or poor the recourse to it that the Constitution, and the relevant legislation say that the litigant should have. … Both the relevant legislation and the Constitution in providing for appeals draw no distinction between questions of fact and law. For my own part I have no doubt that an error of fact is just as capable of causing an injustice, whether it is to be described as a plain, manifest, or gross error or some other form of error, as a mistake of law.

Heydon J even disputed that the findings of fact here were concurrent (at [285]). Gummow J did not see need to address the issue of concurrent findings at all.


  1. Mulligan v Coffs Harbour City Council [2005] HCA 63; (2005) 223 CLR 486.
  2. See Dederer v Roads and Traffic Authority [2005] NSWSC 185.
  3. See Great Lakes Shire Council v Dederer & Anor [2006] NSWCA 101.
  4. The description of this order comes from the decision of Sanderson v Blyth Theatre Company [1903] 2 KB 533 (see particularly at 539).
  5. The description of this order comes from the decision of Bullock v London General Omnibus Company [1907] 1 KB 264 (see particularly at 272).
  6. See Great Lakes Shire Council v Dederer & Anor [No 2] [2006] NSWCA 336.
  7. At [153], referring to Great Lakes Shire Council v Dederer & Anor [2006] NSWCA 101 at [234], [293] and [342].

Text of the decision is available at:

About the author

Paul Sykes, a Senior Lawyer, is Corporate Adviser in the AGS Litigation and Dispute Management Group. He assists AGS lawyers on a broad range of litigation practice and compliance issues, as well as helping to keep them informed about developments in the law.

AGS contacts

AGS has a large national team of lawyers specialising in litigation. For further information on the articles in this issue, or other litigation issues, please contact Leonie Farrant or any of the other lawyers listed below.

Leonie Farrant

Leonie Farrant National Practice Manager Litigation
T 02 6253 7009 F 02 6253 7303

Andrew Berger
Zita Rowling

02 6253 7405
02 6253 7426


Marcus Jafari
Greg Kathner

02 9581 7467
02 9581 7568


Susan Pryde
Graham Thorley

03 9242 1426
03 9242 1244


Christine Runnegar
Barry Cosgrove

07 3360 5623
07 3360 5647


Peter Macliver
Teresa Ling

08 9268 1100
08 9268 1157


Dusan Uglesic
Greg Camilos

08 8205 4298
08 8205 4225


ISSN 1329-458X (Print)
ISSN 2204-6542 (Online)

For enquiries regarding supply of issues, change of address
details etc.
T 02 6253 7052 F 02 6253 7313 E

Electronic versions of AGS newsletters are available for
clients who prefer to receive issues in this form. Please
contact 02 6253 7052 or email to arrange

The material in these notes is provided for general
information only and should not be relied upon for
the purpose of a particular matter. Please contact
AGS before any action or decision is taken on the basis
of any of the material in these notes.