Roads and Traffic Authority of NSW v Refrigerated Roadways Pty Limited

McCOLL JA : I have had the privilege of reading Campbell JA’s judgment. I agree with his Honour’s reasons and with the orders he proposes.

CAMPBELL JA : Nature of the Case

In the early hours of Sunday, 23 August 1998 Mr Mark Evans was driving a semi-trailer truck south on the F5 freeway. The F5 freeway is part of the Hume Highway. An overhead bridge, known as the Glenlee Bridge passes over the freeway near Menangle, somewhat to the south of Campbelltown. That bridge carries a road that leads to a private residence, a horse riding farm, and an entrance to the National Equestrian Sports Centre on the western side of the freeway. The road on the bridge had no pedestrian walkway at the side. As part of the original design of the bridge, waist-high metal railings on a concrete base were constructed along each side of the bridge.

The truck Mr Evans was driving was struck by a large piece of concrete dropped from the Glenlee Bridge. The concrete went through the windscreen of the vehicle, and struck Mr Evans in the chest. His vehicle veered across the highway and up an embankment. Mr Evans died at the scene of the accident. A widow and two children survived him.

Four people were involved in the dropping of objects from the bridge on 23 August 1998. They collected the objects that were dropped from the bridge from the home of one of the men, placed them in the boot of a motor vehicle, and drove to the Glenlee Bridge for the specific purpose of dropping them from the bridge onto trucks passing below. Clearly, dropping the concrete from the bridge onto traffic below was criminal conduct. It was common ground that the four people responsible have been convicted and sentenced for offences concerning this incident.

Mr Evans was an employee of the Respondent, and was acting in the course of his employment when he died. The Respondent made payments to his widow pursuant to the Workers Compensation Act 1987 consisting of a lump sum death benefit and continuing weekly payments.

At the time of the accident, and at all times since the commencement of the Transport Administration Act 1988 the Appellant (the “ RTA ”) had the care and control of the Glenlee Bridge and the Hume Highway at Menangle below the bridge.

The Glenlee Bridge was constructed by a statutory predecessor of the RTA, the Commissioner for Main Roads (“ DMR ”). Design work for the Glenlee Bridge appears, from the date on a plan in evidence, to have been carried out in and about June 1975. While the evidence is not precise about when construction occurred, its construction had been completed by June 1978.

Mr Evans’ widow and children did not take any action for damages against the RTA. In the court, below the Respondent sued the RTA contending that Mr Evans’ death arose in circumstances that created a liability on the part of the RTA to pay damages to the widow and children. In consequence, the Respondent contended, it was entitled, pursuant to section 151Z(1)(d) Workers Compensation Act , to indemnity from the RTA for the payments of workers’ compensation made, plus interest on the death benefit component of compensation.

The trial proceeded on the basis that if the RTA had a liability to the widow and children the damages recoverable by them would be approximately $710,000. The trial proceeded on the basis that, as the total amount of compensation plus interest for which the Respondent sought indemnity was less than $710,000, there was no occasion for the trial judge to quantify the amount of damages to which the widow and children would have been entitled with any greater precision than this.

Adopting that basis was contrary to the principle that a trial judge hearing the first application for an indemnity under section 151Z(1)(d) should determine the quantum of damages to which the worker (or those claiming under him), would have been entitled if the worker, or those claiming under him, had sued: Grant v Royal Rehabilitation Centre Sydney [1999] NSWCA 250; (1999) 47 NSWLR 263. The obvious purpose of this is that, where there are continuing payments of workers’ compensation being made, the amount of damages to which the worker or those claiming under him or her would have been entitled, will provide the basis for calculating the maximum amount of indemnity that will become payable in the future concerning those ongoing payments. The error in proceeding in this way at the trial does not affect the outcome of this appeal.

His Honour Judge Hungerford ADCJ found that the RTA was liable to indemnify the Respondent for the amount it claimed: Refrigerated Roadways Pty Ltd v Roads and Traffic Authority of NSW [2008] NSWDC 91; (2008) 7 DCLR(NSW) 45.

The RTA appeals against the finding of liability in the court below. The issues argued on the appeal were: (1) Even though the RTA admits it owes a duty of care to motorists of the type described in Brodie v Singleton Shire Council [2001] HCA 29; (2001) 206 CLR 512, whether its duty extends to requiring it, either at the time of construction of the Glenlee Bridge or later, to take reasonable care that motorists not be injured by criminal behaviour like that involved in rocks being thrown or dropped from an overhead bridge onto a freeway. (2) If such a duty existed, whether the RTA breached it in the present case (bearing in mind section 42 Civil Liability Act 2002 ) by either (a) not having installed screening on the bridge when it was first constructed, or (b) not having retrofitted screening to the bridge prior to 23 August 1998. There are several sub-issues to these issues concerning breach of duty: (1) Whether the judge was right in making a number of findings that he described as findings of fact. (2) What, if any, significance there is in any limitations on the RTA’s ability to spend money or the usual way in which the RTA obtained money to spend on the F5. (3) Whether it is open to the RTA to rely on section 43A Civil Liability Act in this appeal, and if so what, if any, difference section 43A makes to the result that would otherwise arise. (4) Whether it is open to the Respondent to rely in the appeal on a particular of negligence that had not been pleaded concerning the manner in which RTA decided the order of priority in which bridges should be screened. (3) Whether any breach of duty by the RTA caused Mr Evans’ death. Summary of Decision Reached

I have concluded that the RTA owed a duty of care to motorists such as Mr Evans concerning the dropping of rocks from overpasses over freeways, but that the RTA did not breach that duty either by failing to install screens at the time the Glenlee Bridge was constructed, or by failing to retrofit screens to the bridge at a later time. In so doing I have permitted the Respondent to rely on the particular of negligence that had not been pleaded, but was in substance litigated at the trial.

The conclusion that there is no breach of duty is initially arrived at on the basis of the common law, after taking into account a mass of evidence concerning the funding available to the RTA, and the steps that it took to respond to the risk of objects being dropped from overpasses.

I reject a conclusion that the trial judge arrived at that section 42 Civil Liability Act does not apply to this case, but section 42 does not lead me to a conclusion that is different to the conclusion arrived at from the common law.

I reject the RTA’s application to rely on section 43A Civil Liability Act , but also conclude that, even if it had been permitted to be relied on, it would not have led to a different result.

When there is no breach of duty, the question that was argued about causation of loss does not arise. The Types of Evidence Called

The case of the Respondent in the court below consisted of calling an expert witness on the topic of crash investigation, Mr John Jamieson, and tendering documents, many of which were derived from the records of the RTA.

The case of the RTA in the court below consisted of calling evidence from various RTA officers who had been involved in the RTA’s activities relating to bridges over freeways, expert evidence from Mr Bruce Fishburn (an engineer) and Professor White (a criminologist), and tender of further documents, again mainly from RTA internal records. Some Relevant Roads and Bridges

Details of several freeways around the Sydney region other than the F5 appear in the evidence. These include the F3, the M4, the M5, and the M2.

There are eight bridges over the F5 in the Campbelltown region. As one proceeds south along the F5, there is a bridge that carries Campbelltown Road over the F5 near Prestons and Glenfield, another at Brooks Road, then a bridge that carries Campbelltown Road over the F5 between Ingleburn and Minto, then the St Andrews bridge, then the Raby Road bridge, then a bridge that carries Narellan Road over the F5 at Mt Annan, then one comes to the Glenlee Bridge and finally the Menangle Road bridge. Knowledge of the Problem of Items Dropped from Overpasses

Mr Bruce Fishburn is an engineer who had previously worked for the DMR and then the RTA, but had become a consultant in private practice by the time of the hearing. He gave both expert and lay evidence. He had responsibility for the erection of the Glenlee Bridge. He was unaware of any incidents involving items thrown from an overpass to the freeway below until about 1994.

Analysis of the RTA’s internal documentation assembled for the purpose of this case shows the following concerning the period before Mr Evans’ death: (a) On 17 February 1990 a bag full of sand was dropped from the Mount Kiera Road overpass onto a vehicle on the F5 freeway, resulting in lacerations to the arm of the driver. (b) When the first stages of the M5 motorway were opened in 1992 numerous incidents occurred of objects being dropped from bridges over it. (c) On 2 April 1992 Mr Ron Clayton wrote to the RTA saying that on 28 March 1992 a plastic bag filled with water and eggs was thrown at the windscreen of his vehicle, and three other vehicles, while they were travelling on the Hume Highway between Campbelltown and Liverpool. His letter concluded: “It is extremely disappointing at a time when the NSW RTA is Blitzing nearly every vehicle that moves on the States[’] Roads that they cannot protect motorists from harm and damage through the use of their structures for people to launch missiles. Fortunately in Victoria the RTA saw the error of these structures and erected high wire fences on all Freeway Overhead bridges. I only hope that something can be done about this sort of thing before someone is killed.” An RTA officer replied to Mr Clayton, saying: “No other cases have been reported to the Authority of any person or persons throwing objects from the bridge to which you refer or other bridges on the Freeway. The provision of preventative fencing on bridges along the road could not be justified at this time. However, the Authority will continue to monitor the situation and liaise with the Police Department.” (d) On 16 October 1992 a rock was thrown by “juveniles” from the River Road overpass onto a vehicle travelling on the M5 freeway, injuring the driver. (e) On 8 February 1993 Mr Glen Morgan of the RTA wrote to another RTA officer, saying: “Have discussed these incidents with Picton Police. They indicate that there is a real problem along the freeway in this area & Campbelltown. Suggest discussions with Police or at Traffic Committees to develop strategy.” (f) On 22 July 1993 a concrete block was dropped by “unknown persons” from the Avon Dam Road overpass bridge into the path of a vehicle travelling on the F5. (g) On 24 September 1993 a Mr Saville wrote to the Premier, who was also his Local Member of Parliament, Mr John Fahey. Mr Saville complained that he was the driver of one of three cars damaged when rocks were dropped on the southern expressway from an overpass “near Glen Alpine, several kilometres south of Campbelltown” . That overpass is conceded to be the Glenlee Bridge. His letter continued: “The Campbelltown police officers who were called to the scene were familiar with this form of madness. The only thing that surprised them was that it usually happened closer to Campbelltown. The tow truck driver had had a log dropped on his truck further down the highway and the claims clerk with whom I spoke this morning had seen a similar occurrence on the Central Coast.” Mr Fahey passed this complaint on to the RTA. The Director of the RTA replied to Mr Fahey on 26 October 1993, saying: “Unfortunately there is little the Authority can do. The enclosure of bridges with wire mesh is only practicable in the case of pedestrian bridges. On the M5 only the pedestrian bridge at Mackenzie Street has been treated in this way. On vehicle bridges the height of fencing required may impose loadings on the bridges which they have not been designed to withstand. Given the large number of overhead bridges on the road network any such treatment program would be hugely expensive, and no doubt in the view of many aesthetically undesirable, with no guarantee that it would ultimately deter those determined to carry out these senseless activities.” The reference to “enclosure” of bridges with wire mesh refers to their total enclosure, not to the erection of protective screens at the side. (h) In 1994 the dropping of an object from an overbridge in Victoria caused a fatality. This incident became known to the RTA in NSW. So far as the evidence discloses, it was the first death resulting from this sort of activity in Australia. (i) In March, September and December 1994 there were separate incidents of rocks being dropped from overpasses at three separate Sydney locations (none of which was on the F5) resulting in vehicle damage. (j) In January 1995 a rock was dropped onto the M5, causing a vehicle to overturn. The finding was “RTA and Minister then advising that pedestrian bridges were not being considered” . I take that cryptic statement to mean that screening of pedestrian bridges was not being considered. (k) In April 1995 there were three separate instances, on three different Sydney roads (not including the F5) in which objects were dropped and either the driver was injured or the windscreen was broken. (l) In May 1995 the RTA received a report of an incident that had occurred in March 1994 when an object was dropped from a bridge on the M4, damaging the vehicle. (m) In June 1995 the RTA received a letter from a motorist complaining that on 20 January 1995, while driving along the F5 in a southerly direction, his vehicle was hit by a rock thrown from the Campbelltown Road overpass. The letter stated: “Why the RTA does not install safety net? Before someone gets killed. I am very lucky person to be alive.” (n) In August 1995 there was an incident involving a rock on the M4 at Pomeroy Street, in which a window was broken. (o) In October 1995 there was an incident at a bridge on the M4 that resulted in a window being broken. (p) In October 1995 a pipe was suspended from a bridge on the M4. (q) In January 1996 a brick was dropped from the Raby Road, Campbelltown overbridge and through the windscreen of a vehicle travelling on the F5. (r) There were two incidents in February 1996, one in March 1996, one in April 1996, one in June 1996, one in September 1996, one in February 1997, one in May 1997, one in December 1997, and one April 1998 causing either injury to the driver, windscreen damage or other vehicular damage. (s) An RTA ministerial briefing note dated 24 January 1996 reports that in the year prior to 24 January 1996 there were seven incidents known to the RTA when objects were intentionally thrown from an overbridge onto vehicles travelling beneath. Motorists were injured in three of those incidents.

There was uncontested evidence that in the 1990s there were incidents not only of objects being dropped onto freeways from overhead bridges, but also of objects being thrown onto freeways from embankments at the side of the freeway. These included some incidents recorded in police records that do not seem to have a corresponding entry in RTA records, namely: (a) In November 1994 a semitrailer driving south on the F5 had its windscreen broken by a rock that youths threw from an embankment on the site of the expressway. (b) On 31 July 1997 youths aged 10 to 12 threw rocks at vehicles on the F5 near Campbelltown “from the Fullwood Reserve” . I would infer that these rocks were thrown from the side of the road rather than from a bridge. Action Concerning Bridge Screening

Though the trial judge did not make detailed findings concerning the way in which screening of bridges over expressways progressed over time in the Sydney region, the course of events emerges clearly from contemporaneous documents. The RTA talks of the “Sydney Region” as including not only the Sydney metropolitan area, but also the areas covered by Blue Mountains City, Gosford City, Hawkesbury City, Wollondilly, and Wyong.

In 1993 and 1994 two pedestrian bridges over the M5 were screened. The evidence does not establish that it was the RTA, rather than the operator of the M5 tollroad, that carried out this screening.

In April 1995 the Minister for Police convened a committee comprising the Police, RTA, State Rail Authority (SRA) and local government to consider action following several incidents where motorists had been injured in the preceding months. The RTA committed to screening some bridges, in an order of priority determined by a risk methodology that it had developed in January 1995.

Mr Stephen Dunlop is an engineer employed by the RTA. In mid 1995 Mr Dunlop was appointed as the RTA’s Sydney Asset Manager. Part of his responsibility in that position was for hazard reduction concerning objects dropped from overbridges. He was assisted by Mr S’Muttu Sivarasa, an engineer who worked for the RTA as a Budget Manager and Bridge Maintenance Engineer.

In June 1995 Mr Peter Sansom, an M5 Project Manager with the RTA, sent Mr Dunlop a memo concerning objects dropped from bridges on the M5. He reported that there had been a number of incidents associated with objects being dropped from bridges over the M5, that when the first stages of the M5 were opened in 1992 numerous incidents occurred, and in consequence Interlink Roads (the private company that developed and operates the M5) undertook the removal of loose items that could be used as projectiles in the vicinity of overbridges, and employed security guards to monitor the overbridges. The memo continued: “The Mackenzie Street pedestrian bridge was then caged. It is apparent that publicity associated with incidents, including a Victorian fatality, generated further spates of incidents. In June 1994 the M2 Casula Link was opened to traffic and while the sides of the Box Road pedestrian bridge were screened, incidents where rocks and tree planting stakes were dropped on vehicles. The roof of the pedestrian bridge was subsequently caged. Further incidents have been reported from Council’s North South Link Road. Observations It is evidence that incidents are likely to occur when: - an overbridge is in proximity to a school or shopping centre; - the overbridge is relatively isolated with little passing traffic; and - material that could be used as a projectile is in close proximity to the overbridge.”

On 10 August 1995 Mr Dunlop sent to Mr Ray Wedgwood, the Manager of the Bridge Branch of the RTA, a brief for developing concepts for the screening of four particular bridges. Of these, three were pedestrian bridges, and the other was a combined road and pedestrian bridge. The bridges were over the M4 and the F3. Mr Wedgwood provided cost estimates promptly, and an estimate of completion time of the end of September 1995. He was instructed to proceed with the work.

On 14 November 1995 Mr Dunlop produced an 11-page report on hazard reduction from overpass structures. It reported that four bridges over major roads in the Sydney region had been identified as having the greatest likelihood of an incident. These were the same four bridges as had been identified in the 10 August 1995 memo, but Mr Dunlop’s report explained on what basis they had been chosen – see para [35] below. Other statements in it were: “It is proposed that protective screens be erected on these bridges to deter future criminal acts. It is also proposed that all future overbridges and bridges where incidents are reported be assessed under a similar system and screens erected as required.”

It reported that a working party of five people (including Mr Wedgwood, Mr Sansom, Mr Sivarasa, and Mr Dunlop) had been formed to identify the risks and develop strategies. That working party had identified the scope of the hazard potential as extending to a total of 121 bridges. These were either bridges that were solely pedestrian, or road bridges with pedestrian traffic. It said: “From the limited data available most incidents occur from pedestrians crossing an overbridge throwing or dropping an object which was found around the overbridge. … Police records show a low level of known incidents in the Sydney Region of approximately five per year. … [I]ndications are that most involve opportunistic rather than premeditated behaviour.”

It also reported that a committee had been formed comprising representatives of the police, local government, the State Rail Authority and the RTA. It reported: “… the problem for the SRA is severe. During a two week survey for the committee the SRA recorded 21 incidents of trains attacked by persons throwing missiles from bridges or track side. The Committee considered that the RTA’s approach of analysing sites for the potential hazard and dealing with each site as required the most effective method of treatment.”

By that time the working party had developed an assessment survey for bridges which identified nine different risk factors and two variables that might apply in relation to each risk factor. Using that survey, certain pedestrian bridges over state roads and all bridges over restricted access roads were surveyed. The criteria for a bridge being surveyed was that it was in the Sydney region, was the responsibility of the RTA or over another RTA road, and there was a pedestrian walkway on the bridge. As well, overbridges carrying reasonable traffic volumes were not considered as it was considered that potential miscreants would be deterred by the possible observation. Ten particular bridges were identified using that methodology as those with the highest priority to screen. The most urgent four were put in that category because each was near a school, a hotel or a youth attraction (or more than one of those), and had either no lighting, or remote lighting. The Glenlee Bridge was not one of those identified. As it is a bridge that did not have a pedestrian walkway, it would not have been considered in the survey.

On 12 September 1996 Mr Dunlop prepared a document concerning the risk of objects thrown from bridges onto roads below. It included: “During the past year there has been seven incidents known to the RTA where objects have been intentionally thrown from an overbridge onto vehicles travelling beneath. On three occasions motorists in the vehicles were injured. In Victoria in 1994 a motorist was killed in a similar incident. Police records show a low level of known incidents in the Sydney Region of approximately five per year. More incidents occur outside the urban area. While the Police consider acts of throwing objects onto vehicles a criminal act indications are that most involve opportunistic rather than premeditated behaviour. In comparison a report from the Chicago area of the US has in one year 600 incidents of objects being thrown from bridges over roads. 2. Analysis of Sites The potential sources of problems in Sydney Region are:- a) Pedestrian bridges over State Roads – 61, (including 10 over Freeways and Motorways – four of the latter are already encased). b) Road bridges with pedestrian traffic over Freeways and Motorways – 60. c) Road bridges with pedestrian traffic over State Roads – 110. … The incidents of missiles being thrown from bridges are sporadic in both location and time. Accordingly, there were no bridges identified as having a definite history of incidents.”

Thus, a total of 231 bridges in the Sydney Region were identified as “potential sources of problems” .

Mr Eka Onggo was an engineer who worked for the RTA as a senior project manager. Mr Onggo had, by the end of 1996, developed a methodology for assessing the risk of objects being dropped from overbridges. It was an elaboration and refinement of the methodology that had been used in Mr Dunlop’s report of 14 November 1995. Mr Onggo’s methodology was applied to various bridges under the control of the RTA, and by its use an order of priority for screening those bridges was arrived at. The detail of the manner of operation of the risk assessment process will be considered in greater detail later (paras [409] ff below). Operation of the Budgetary Approval Process

The F5 and the F3 are both classed as National Highways. Funding for both maintenance and improvements on National Highways is ordinarily provided by the Commonwealth Government, by means of tied grants under section 96 of the Constitution.

Our attention was not specifically drawn to any evidence that identified any legal framework that led to the classification of some roads as National Highways, or any clear statement of the basis on which the roads came to be classified as National Highways, or the precise consequences of being so classified. However, it is apparent from documents in evidence, including detailed budgetary documents, that National Highway roads in New South Wales include a link between Sydney and Melbourne constituted by the Hume Highway (of which the F5 is part), a link between Sydney and Brisbane constituted by the Newcastle freeway (the F3) and the New England Highway, certain urban links in the Sydney region connecting the Hume Highway and the Newcastle freeway, a link between Sydney and Adelaide constituted by the Sturt Highway, a link between Melbourne and Brisbane constituted by the Newell Highway, and some Canberra connections constituted by the Federal and Barton highways.

Accounting for road works is done on the basis of financial years running from 1 July to the next 30 June. A process is in existence whereby State Ministers seeking funding for National Highway roads submit to the Commonwealth Government a forward strategy that itemises proposed work in the next five financial years. Those forward strategy documents are forwarded in February or March of a year. They are taken into account in the framing of the Commonwealth Budget for the next financial year, which usually is delivered in May of a year.

Items such as screening of bridges over freeways are classified as safety and minor works. Proposals for safety and minor works need to be taken account of in the overall forward strategy documents and an item covering such works in general terms is included in the Commonwealth Budget. However, during the years relevant to the present case, the State Government made submissions, and the Federal Government made decisions, concerning the detail of the safety and urgent minor works that the Commonwealth would fund in a process that operated separately to the overall Commonwealth Budget process.

The first appearance of screening of overpasses in the Commonwealth budgetary process was on 24 January 1996, when the Chief Executive of the RTA sent the NSW Minister for Roads, Mr Michael Knight, a briefing note relating to bridges over motorways in urban areas. It stated the background as follows: “* Following a number of serious incidents involving injury to motorists and damage to vehicles the RTA completed a full investigation into the throwing of projectiles from overbridges and methods to prevent such action. * The investigation team looked at the circumstances surrounding such incidents and recommended the treatment of high risk overbridges that had not already been treated under previous programs. … * The outcome of the study was the identification of bridges on the M4 Motorway, the F3 Newcastle Freeway and the F5 South Western Freeway as having the highest priority for treatment. There are 43 bridges on these roads requiring treatment. At $50,000 each the total cost is estimated at $2.2M. Both the F3 and the F5 are National Roads and it is proposed that the Federal Government be approached to provide $1.2M in 1996/97 to cover the works.”

By the time of that briefing note, the RTA had already announced a programme for screening of overbridges on the M4 motorway. The M4 is not a National Highway. Presumably, any estimates and appropriation to cover the cost of that work, and any other screening of overbridges on roads that were not National Highways, was carried out through the budgetary process of the State Government. The evidence in the present case did not go into the budgetary process of the State Government concerning roads. Screening of bridges over roads for which the State Government was responsible was touched on only incidentally in the evidence, as such bridges were considered by the RTA as part of the process of risk evaluation concerning the dropping of objects from overpasses in the Sydney region.

The briefing paper continued: “* During January 1996 a brick was dropped from the Raby Road, Campbelltown overbridge and through the windscreen of a vehicle travelling on the F5 Freeway. * Previous investigations had not indicated a problem at this overbridge (it is relatively heavily trafficked, there is clear visibility with no special developments such as schools either side) however, the site will now be included in investigations. Way Forward * Given the wide spread problem of projectiles thrown onto motorways having the potential to cause serious injury and death, the concept of treating the whole length of a motorway such as proposed for the M4 is seen as the only comprehensive approach. The bridge you don’t treat is the one used by the criminals. * It is therefore proposed to make a submission to the Federal Government and outline the action taken by the State on the M4 and request funding to treat the bridges on both the F3 and F5 Freeways in the urban areas of Sydney.”

An annexure to the briefing paper informed the Minister: “During the past year there has been seven incidents known to the RTA where objects have been intentionally thrown from an overbridge onto vehicles travelling beneath. On three occasions motorists in the vehicles were injured. In Victoria in 1994 a motorist was killed in a similar incident. Police records show a low level of known incidents in the Sydney Region of approximately five per year. More incidents occur outside the urban area. While the Police consider acts of throwing objects onto vehicles a criminal act indications are that most involve opportunistic rather than premeditated behaviour. In comparison a report from the Chicago area of the US has in one year 600 incidents of objects being thrown from bridges over roads.”

The briefing paper contained an annexure that included the information contained in the document prepared by Mr Dunlop on 12 September 1996 (and the material quoted at para [36] above). It stated that there were a total of 231 bridges over roads in the Sydney region – some of these were pedestrian bridges, and some were road bridges with pedestrian traffic. The Minister was informed: “The general consensus is that the highest risk occurs with bridges over high speed, restricted access roads. On the F3 freeway, the M4 Motorway and the F5 Freeway (excluding the Tollway) there are 43 bridges to be treated.”

The Chief Executive expressed the view that, of possible solutions: “The use of advertising, press releases or community involvement has not been considered suitable due to the risk that such actions could increase awareness of the activity and actually promote missile throwing. Similarly, erection of signs warning of the danger and criminal penalties is considered counter productive. Increased surveillance is considered impractical at the sites nominated. The Police are unable to provide increased resources, which being intermittent are usually ineffective. Cameras and security patrols are expensive for a very infrequent occurrence. The preferred treatment to reduce the hazard is the erection of physical barriers on the bridges. These barriers are to either prevent or restrict missiles from being dropped or thrown. To enable an assessment of the cost and effectiveness of physical barriers four bridges were selected from the candidate ten for concept design and cost estimates.”

The current programme was explained as follows: “The need exists to reduce the risk on the outer urban freeways and Motorways by the erection of fencing on all overbridges. Currently four bridges have been chosen on the M4 Motorway and the F3 Freeway as having the greatest likelihood of being an incident site. Attached are concept designs for the four bridges under consideration. The treatment for these four bridges involves erection of an enclosure along each side of the bridge at the existing handrails. The concept designs for the addition of an enclosure on each bridge has attempted to complement the existing design, to reduce any aesthetic effects. The program of works on each bridge allows for architectural concept design, a review of environmental factors and possibly community consultation, prior to construction. It is expected that the four bridges currently in design will be completed during 1995/96, with the remaining 39 bridges on the outer urban freeways and motorways to be completed in 1996/97. The estimated costs of these barriers and enclosures are $50,000 per bridge, with a total cost for the 43 bridges of $2.2M.”

In February 1996, the RTA sent to the Commonwealth Department of Transport its forward strategy report for 1996/97 to 2000/01. It proposed total expenditure in NSW over those five years on National Highway roads of $1,795,666,000. The RTA sought $336m in 1996/97, by comparison with the $299m that the Commonwealth had made available in 1995/96. The detailed listing of work proposed in that five-year period included nothing about bridge protection fencing.

The New South Wales Minister for Roads wrote to the Commonwealth Minister for Transport and Regional Development, the Hon J Sharp MP, on 30 April 1996 stating that there had been “a number of serious incidents in Sydney over the past two years where motorists have been injured by projectiles thrown or dropped from overbridges” , referring to the death from such an incident in Victoria in 1994, and stating that to reduce the hazard to vehicles from objects dropped from overbridges “it is proposed to erect fencing along the outer side of all bridges over high speed urban areas” . The letter concluded: “There are 13 bridges on the F3 Freeway and 8 bridges on the F5 Freeway requiring treatment. The expected total cost of the fencing project is $1.1M at $50,000 per bridge. It is proposed that the work be carried out in 1996/97. For your information the State Government has allocated $1M for similar work on the M4 Motorway in Sydney. My purpose in writing is to seek your concurrence to the proposed bridge fencing and to seek funding under the National Highways Program during 1996/97.” Though the eight bridges on the F5 were not specifically identified, I would infer that they were the eight bridges I have listed at para [22] above, and thus included the Glenlee Bridge.

On 19 June 1996 Mr Sharp wrote to Mr Knight, referring to his letter of 30 April 1996, and saying: “I agree there is a serious safety risk of projectiles being thrown from bridges onto heavily trafficked roads and that providing fencing is the best preventative measure. I shall be happy to consider such safety works in the context of developing the 1996/97 National Highway program.”

There had been a change of federal government following the election in March 1996. This resulted in the Federal Budget relating to 1996/97 being delayed past the usual May date. On 20 August 1996 Mr Sharp wrote to Mr Knight advising of the Commonwealth 1996/97 budget decisions concerning road funding. Mr Sharp advised that in 1996/97 New South Wales would receive $306.7m. That figure of $306.7m included a line item of $27.36m for safety and urgent minor works. (The body of Mr Sharp’s letter said that $64m would be allocated for safety and urgent minor works, but that was inconsistent with a detailed table that accompanied the letter that showed $27.36m being allocated for safety and urgent minor works. All subsequent correspondence is consistent with the figure of $64m being a typographical error, and the real figure being $27.36m.) Mr Sharp said: “The safety and urgent minor works component, however, will be administered through the normal program approval process and, accordingly, I request submission of your proposals for funding as soon as possible.”

In other words, even though the budget made provision of $27.36m for safety and urgent minor works in New South Wales, the breakdown of that sum into specific projects was not part of the budget, and was to be decided separately.

On 21 November 1996 the Assistant Secretary of the Commonwealth Department of Transport and Regional Development wrote to the Chief Executive of the RTA informing him of the allocation for construction and maintenance funding for 1996/97, and providing a list of project allocations for several large construction works. The letter stated that the next federal budget would be in May 1997. The list of project allocations in the letter showed that $27.36m was allotted for safety and urgent minor works in 1996/97, but the letter did not break that sum up to identify any specific projects, or types of project. It foreshadowed that the federal funding available for safety and urgent minor works would decrease in future years. The foreshadowed decreases were to $25.99 million in 1997/98, $24.62 million in 1998/99, and $23.26 million in 1999/2000.

On 5 February 1997 the New South Wales Minister (by now the Hon Carl Scully MP) wrote to the federal minister, referring to the budget advice that New South Wales would receive $27.36m in 1996/97 for safety and urgent minor works, and the budget advice about indicative Commonwealth funding for the forward three years 1997/98 to 1999/00. The letter attached a schedule reviewing the safety and urgent minor works component of the National Highways forward strategy to reflect that budget advice and indicative forward funding. It sought confirmation that the proposed use of the $27.36m in 1996/97 was acceptable.

The attached schedule showed (accurately) that nothing had been proposed for overhead bridges protection screening in the forward strategy for 1996/97. It showed that it was still the case that nothing was proposed to be spent for overhead bridges protection screening in 1996/97. It showed that a total of $1.05m was proposed to be spent on bridge screening in 1997/98, made up of $650,000 for expenditure on 13 bridges on the F3, and $400,000 for expenditure on eight bridges on the Hume Highway. Again the document does not identify the eight bridges on the Hume Highway, but I would infer that they included the Glenlee Bridge.

On 27 March 1997, Mr Sharp wrote to Mr Scully referring to his letter of 5 February 1997, and approving the proposed works for the 1996/97 year, to a limit of $26.725m.

In March 1997 the RTA submitted its forward strategy report to the Federal Department, relating to the years 1997/98 to 2001/02. The report included: “The New South Wales National Highways Forward Strategy for 1997/98 to 2001/02 has been developed within the context of funding advice provided in November 1996 by the Federal Department of Transport and Regional Development. This advice provided future funding profiles for major works, maintenance and safety and urgent minor works. The focus of this advice was a five percent per year reduction in maintenance funding over four years, to achieve a 20 percent ‘efficiency dividend’ , and reduced funding for safety and urgent minor works. The priorities of this Forward Strategy are to: complete works in progress, commence and/or complete a limited number of new high priority works, and maintain National Highways system assets. … Previous NSW National Highways Forward Strategies identified a number of major and minor works which, under the funding levels proposed, will not be able to be undertaken within the five year period of this Strategy. NSW seeks funding in 1997/8 of $293M and an average allocation of $302M (including administration costs) per year over the next five years (dollar values expressed in 1996/97 terms). These levels of funding are considerably less than those sought in previous Forward Strategies and reflect advice received from the Federal Department of Transport and Regional Development referred to earlier.”

The detailed listing of proposed works included $1.05m proposed in 1997/98 for bridge protection fencing. The inference is that this figure related to 21 bridges, at $50,000 each. Thus, it sought to cover all eight bridges on the F5, and all 13 bridges on the F3, but did not reflect a revised costing of $70,000 per bridge that the RTA came to adopt around this time or soon after.

Mr Neil Forrest was the RTA’s Road Asset Manager, Sydney Region. Mr Rod Carter was the Network and Road Safety Manager for the RTA. On 13 April 1997 Mr Sivarasa sent a memo to Mr Forrest and Mr Carter, reporting that the RTA’s study had identified 22 bridges over the M4, 13 bridges over the F3 and 8 over the F5 as having the highest priority for treatment. He stated that the Federal Government had agreed to provide $1.1m in 1997/98 to cover the 21 bridges on the F3 and F5. His statement of the current position included: “Cost of screening of bridges has escalated to $70,000. Only 15 of the 21 bridges in F3 and F5 can be screened during 1997/98. Federal Government will be approached to fund the remaining six bridges in 1998/99. However these six bridges have low priority for treatment …”

His recommendation was: “The problem of projectiles being thrown from bridges onto motorists was clearly demonstrated by incidents in 1996. Given the media attention to these issue[s] at the time it is recommended that the program of retro fitting screens to the 55 bridges already identified on high speed roads should continue. The program of works on each bridge allows for architectural concept design, a review of environmental factors and community consultation prior to construction. The estimated cost of these barriers and enclosures is $70,000 per bridge, with the total cost for 55 bridges of $3.85M. It is proposed the work be staged over a 5 year period with 10-12 bridges being treated annually.”

On 30 April 1997, Mr Forward, the Director of Road Network Infrastructure of the RTA, wrote to the Assistant Secretary of the federal department, providing requested information concerning monthly cash requirements for 1996/97 for each federally funded project. It stated: “The end June forecast for safety/minor works is for an under-expenditure of $5.5M (20%). This is largely due to the delay in obtaining approval to the safety/minor works program, particularly the four overtaking lanes and the Chelmsford Drive/Metford Road improvements on the New England Highway in the Hunter which comprise 30% of the program. Attached is a schedule of works (one page) to address this shortfall, involving bringing forward works from 1997/98 and introducing additional works. Approval to proceed with these works is requested as soon as possible in order to expend the funds.”

The attached schedule of work listed various safety and minor works projects, of a total estimated value of $5.49m, beyond the expenditure that had initially been proposed for safety and minor works for 1996/97. In other words, it proposed alternative uses for nearly the whole of the expected under-expenditure. Included in that list was $50,000 for overhead bridges protection screening for the Hume Highway, and another $50,000 for work of that type on the F3. The “comments” concerning each of those items was: “Bring forward commencement from 97/98. Federal Minister’s letter 19/6/96 refers (copy attached).” (That letter is at para [52] above.)

On 9 May 1997 the Commonwealth department approved the additional safety and urgent minor works that the RTA had requested on 30 April 1997.

Mr Vivian Manwaring was at the relevant time a Senior Capital Works Programmer with the RTA. In that position he was involved with the delivering of major projects that were funded by the Federal Government. On 12 May 1997, Mr Manwaring emailed various people in the RTA, saying: “Further to my fax last week about noise abatement works on the National Highway, DoTRD has today approved the following works for 96/97 to try and address under-expenditure: * Noise wall Hume Hwy Lachalsh to Raby St Andrews $1,000,000 * Noise wall Cumberland Hwy OWR near Christine St Northmead $220,000 * Hume Hwy screening of overhead bridges (proj 65540/1) $50,000 * F3 screening of overhead bridges (proj 65540/2) $50,000 Please proceed with these projects to expend as much funds as possible this financial year.” (DoTRD is the Commonwealth Department of Transport and Regional Development.)

On 14 May 1997, Mr Piper emailed Mr Allan Procter, saying: “An approved allocation for 1997/98 of $1.05M exists under National Highways funding for screening of bridges on Hume Freeway [sic] and Sydney-Newcastle Freeway within Sydney Region. The opportunity exists, if funds are available, to spend up to $150,000 this financial year, on procurement of stores for this project. Do you approve?”

On 26 May 1997 Mr Procter emailed Mr Manwaring about that request, saying: “I have no objection – depends on your other moves to balance money?”

Later on 26 May 1997, Mr Manwaring emailed Mr Piper, saying: “Allan, as discussed today, please proceed with procurement. 96/97 National Hwys safety & minor works program is underspent, and the Feds have approved an early start to this bridge screening in 96/97. 96/97 program is underspent because of late approval to proceed with works. This means, of course, a squeeze on 97/98 program due to carryovers and no additional Fed funds. Consequently, after doing some number crunching this morning, I anticipate that 97/98 $1.05M for this bridge screening will have to be reduced to around $0.3M to $0.4M with the rest in 1998/99. This means treating around 6 to 8 bridges in 97/98. I anticipate a detailed review, in liaison with Regions, next month. We won’t get Fed advice on program allocations for 97/98 and forward years until mid-June.”

Still later on 26 May 1997 Mr Piper emailed Mr Sivarasa: “Siva, please advise all bridges in priority, with cumulative $s. Procure as much materials as possible this year, even if we have to then store them for a year. Don’t treat $150k as a limit on this year’s expenditure.”

On 4 July 1997, Mr Manwaring emailed a large number of RTA staff: “I thank, and congratulate, everyone involved in National Highways Safety & Minor Works in [1996/97*]. We achieved expenditure of $24.1M against the Federal allocation of $27.36M despite the fact that Federal approval to the Works did not occur until 27 March 1997 (and some supplementary Works in May 1997). In the circumstances, this was an excellent effort. Unfortunately however, there will be significant carryover of Works to 1997/98. Accordingly I request that Regional ‘bids’ for National Highways Safety & Minor Works for 1997/98 and the following five years (ie 1998/99 – 2002/03 inclusive) be forwarded to me in three separate schedules as follows: 1. Works physically commenced which require continued funding in 1997/98. 2. Works committed in 1997/98 but not physically commenced (nature of commitment to be stated, but ‘included in Electorate Statement’ will be insufficient justification). 3. Works desired (ie not physically commenced or committed) preferably ranked in Regional priority.” (*Date as corrected by Mr Manwaring tp 633.)

Thus, the RTA had not spent $3.26m of the amount that had been allocated by the federal government for the 1996/97 year. However, that did not mean that the RTA could spend that $3.26m in the following year. There is uncontested evidence from Mr Manwaring that if the federal funding attributed to a particular year was not spent it would lapse. The manner in which federal funding took place was that it was not paid to the State in advance, but rather the State was “paid on actual expenditures each month” .

Mr Procter’s reply to Mr Manwaring on 10 July 1997, included: “Not sure that ‘2’ should exclude electorate statement jobs; there is a view here that these must now be seen as govt ‘promises’ to members & should be delivered unless clearly impossible or member agrees with a substitution.”

Mr Manwaring replied to Mr Procter later on 10 July, saying: “I acknowledge your comment re electorate statement jobs and agree … it will be impossible to deliver all of those in 97/98 [National Highway] Safety & Minor Works because there is something like $6.5M in carryover works from 96/97 into $26M available in 97/98. To further compound this funding problem, there is [a railway level crossing] on the Newell Hwy north of Moree where there has been two fatals in the last month. Lew Lang has been dealing with this problem and flashing lights/bells must be installed in 97/98 to appease the community. This was unplanned, and will cost around $300,000 to $400,000.

Mr Manwaring explained the significance for the budgeting process of “carry-overs” as being: “The first call on the next year’s funds is to complete projects that have started – physically started – and therefore need to be finished, or were supposed to have started and for some reason haven’t, and therefore they have a commitment from the previous year. And with a late start to the approved works, unfortunately some of that cost translated over into the next financial year which had not been factored in when the program originally went to Canberra. Therefore, with additional funds being required for the next year to complete those started works, or committed works, but the total funds available not expected to increase, that means the funds allocated to projects will have to move to accommodate that anticipated additional funds required.”

In the meantime, the RTA was taking some statutorily required preliminary steps to fencing any overpasses.

The RTA has obligations under section 111 of the Environmental Planning and Assessment Act 1979 to examine and take into account to the fullest extent possible all matters affecting or likely to affect the environment by reason of its activity. In particular, it is obliged, by section 112 of that Act, to prepare an Environmental Impact Statement (“ EIS ”) before commencing an activity that is likely to significantly affect the environment (including critical habitat) or threatened species, populations or ecological communities, or their habitats. The way in which the authority decides whether a proposed activity is one for which an EIS must be prepared is by carrying out a study called a Review of Environmental Factors (“ REF ”). Factors to be addressed by a REF were listed in clause 82 of the Environmental Planning and Assessment Regulation 1994 .

A document in evidence explained that a REF: “… is an environmental impact assessment document which identifies examines and assesses the likely environmental impacts of a proposal. The REF is assessed to determine if those impacts are likely to be significant and hence whether an EIS is required.”

In May 1997, Mr Terry McCoy, who was the Project Manager of Environmental Services with the RTA, provided a proposal to prepare 20 REFs for proposed screens on pedestrian bridges and overbridges on the M5, SH13 and F3. By the time Mr McCoy commenced preparation of those reports, on 2 June 1997, the Glenlee Road Bridge had become one of the bridges he was to review.

On 10 June 1997, Mr McCoy wrote to the Campbelltown City Council informing it that the RTA was proposing to erect screens on the eight bridges over the Hume Highway. The Glenlee Bridge was specifically identified as one of those bridges.

The REF relating to the Glenlee Bridge was completed by 1 August 1997. It identified the period of construction of screening on the bridge as being 10 to 14 days, and identified that the work would commence “Within 1997-1998 financial year when funds become available.” As a result of the REF it was decided that a full EIS was not needed.

Mr Alex Jugueta was an RTA Project Manager. On 16 October 1997, Mr Jugueta prepared a summary for Mr Sivarasa of the contracts already awarded concerning the overbridge screening project on federal roads. There were six contracts awarded, relating to the environmental study, design, purchase of materials, and installation at both St Andrews Road and Raby Road on the F5. The total obligation pursuant to those contracts was $260,862, of which $24,875 had been expended during 1996/97. That left $235,987 that had already been committed for those two bridges alone. The allocation for the entire project of screening overhead bridges for 1997/98 was $370,000. I infer that by this time the “around $0.3M to $0.4M” that Mr Manwaring had estimated on 26 May 1997 as being available for bridge screening in 1997/98 had firmed to a figure of $370,000. Thus, a balance of $134,013 was available in 1997/98 to carry out bridge screening work that had not already been committed.

Mr Manwaring explained why it was that the amount available for bridge screening was reduced in the 1997/98 year from $1.05m to $370,000: “… some projects, once they’re started you can’t break it up into sections – such as an overtaking lane. Once you start, you must build it in its entirety. An intersection that requires traffic signals, you must do it in its entirety, you can’t do part. Whereas some projects, for example bridge screening, it is composed of a number of discrete bridges, therefore you could do some and defer others.”

Mr Jugueta wrote to Campbelltown council on 28 October 1997, reporting that installation of overhead screens for St Andrews Road and Raby Road was substantially complete, and that installation of overhead screens for the Campbelltown Road bridge at Glenfield was scheduled for completion by the end of November 1997. That latter remark does not appear to take into account the budget pruning process that was underway within the RTA, and had already resulted in a reduction of the funds available for bridge screening.

On 23 October 1997 the Chief Executive of the RTA sent a submission to the New South Wales Minister concerning the National Highways safety and urgent minor works review. It referred to the fact that $25.99m had been allocated as the New South Wales amount for safety and urgent minor works in 1997/98, and that the works proposed for 1997/98 had been reviewed, taking into account carry over works from 1996/97. He explained that the strategy taken for 1997/98 was: “1. complete works in progress – $12.62 million or 49% of funds; 2. continue works in progress – $3.16 million or 12% of funds; 3. include projects committed by the State or Federal Governments (eg Newell Hwy Coolbaggie Creek and Billabong Creek bridges replacement, Newell Hwy overtaking lane 120km north Moree, Newell Hwy Mirrool rail level crossing advance signals, Newell Hwy Camurra rail level crossing signals where recent fatalities have occurred, F3 Cooranbong noise walls, New England Hwy Arthur Street Rutherford traffic signals, Hume and Federal Hwy ‘Remembrance Driveway’ plantings/signs/rest areas upgrading) – $6.79 million or 26% of funds; and 4. include other high priority works (eg Newell Hwy Alleena rail level crossing signals, Newell Hwy Forbes and Parkes heavy vehicle bypass intersections improvement, Newell Hwy various overtaking lanes and rest areas, New England Hwy Blandford School safety improvements) – $3.42 million or 13% of funds.”

He stated: “$8.71 million of the $12.62 million to complete works in progress (item 1 above) is for unscheduled carryover works from 1996/97. this is due to late starts in 1996/97 on seven projects [which he listed]. Late starts on these projects resulted from the 1996/97 program not being approved until 27 March 1997.”:

The Chief Executive included a draft letter for the Minister to send to the Federal Minister.

Mr Scully wrote on 24 November 1997 to the then Federal Minister, the Hon Mark Vaile MP. He enclosed a schedule of safety and urgent minor works to reflect the budget advice and indicative forward funding. That schedule totalled $25.99m for 1997/98. He sought the Federal Minister’s confirmation that the proposed use of that $25.99m was acceptable.

The enclosed schedule, insofar as it related to overhead bridge protection screening, noted that $400,000 had been listed in the forward strategy for 1997/98 relating to the Hume Highway, and $650,000 relating to the F3, that the amounts now proposed for 1997/98 were $140,000 relating to the Hume Highway, and $230,000 relating to the F3 (which I interpolate totals the $370,000 that Mr Jugeuta had referred to on 16 October 1997), that expenditure in 1998/99 was proposed to be $450,000 for the Hume Highway, and $220,000 for the F3, and that the 1997/98 would allow two bridges to be screened on the Hume Highway and three bridges screened on the F3.

On 19 December 1997, Mr Vaile wrote to Mr Scully approving the requested manner of expenditure of the $25.99m.

In March 1998, the RTA forwarded to the Commonwealth Department a revised forward strategy report concerning the National Highways program, for the years 1998/99 to 2002/03. The detailed listing of works continued to show two items totalling $370,000 as estimated expenditure in 1997/98 for overhead bridge protection screening, and two items totalling $670,000 as the estimated expenditure for that item in 1998/99.

The Federal Budget was delivered on 12 May 1998, and included an amount of $24.62m allotted for safety and urgent minor works in New South Wales on the National Highway.

On 25 May 1998 Mr Onggo emailed Mr Sivarasa seeking confirmation “that the proposed funding for the F3/F5 bridge screening is still $670k” . Mr Sivarasa passed that question to Mr Manwaring, who replied, later on 25 May 1998: “Siva, at present there is $670k of Fed funds earmarked to continue bridge screening in 98/99. However, with Regions/Sydney Ops, I will be undertaking a review of National Hwys Safety/Urgent Minor Works Program (the funding source) in July to take account of any carryovers and possible changed priorities. Until that exercise is completed, I can’t guarantee all the $670k.”

A monthly status report summary, as at 31 May 1998, for the screens on overhead bridges on National Highways projects stated that the allocation for the current year was $370,000, and expenditure to date in the current year was $398,000. It stated: “Construction for this financial year is complete. Detailed design for next year’s work to commence in July 98. Further work will depend on availability of funds.”

As at 30 June 1998, the situation remained that the approved allocation for the financial year was $370,000, and actual expenditure as at 30 June 1998 was $398,000. By 30 June 1998 three bridges over the F5 had been screened, namely those at Campbelltown Road, Prestons, St Andrews Road and Raby Road.

On 10 July 1998 Mr Dunlop prepared a memorandum for Mr Carter concerning the status of the screening project. It recorded that the Federal Government had agreed to provide $370,000 in 1997/98 to cover five bridges on the F3 and F5. The cost per bridge had risen to $75,000 per bridge.

Mr Dunlop summarised the current position concerning screening of bridges over all roads for which the RTA had responsibility as follows: “National Freeways * $670,000 has been allocated in 1998/99 to cover nine bridges on National roads. No funding has been provided for the remaining eight bridges. State Freeways * 12 bridges have been completed in M4. A further two bridges will be completed before the end of the 1997/98 financial year. 4 bridges have been programmed to complete in 1998/99. Funding required for 3 bridges. 12 bridges have been identified over M5 which could be fenced. No funding has been requested to date. National Highways * 8 bridges have been identified over the Cumberland Hwy which could be fenced. No funding has been requested to date, yet all of these bridges have high risk assessment. State Highways * The risk study was extended to overbridges on State Roads having speed limits of 80km/h or over in the Region. The outcome of the study was the identification of 41 bridges which have a significant potential risk. * Please see the attached evaluation matrix for priority ranking. * During 1996-1998 there were two incidents reported on the State Road at Parramatta Road Forest Lodge and the Western Freeway near Homebush, where a rock was dropped on a vehicle, without injury to occupants. Homebush incident occurred on 24 April 1998. * There have been no incidents reported from bridges where fencing has been erected.”

Thus, Mr Carter identified 87 bridges over roads that were unfenced. The options that Mr Carter listed included: “a) Do nothing further. This option assumes either the public will not be tempted to drop further items or that the worst cases have been treated. As there has been no media campaign or other efforts to change public opinion the former is unlikely. For the latter the bridges remaining on the F3, M4 and F5 have the same level of risk as the bridges previously treated and hence the worst cases have not yet been treated. b) Treat Outer Urban Freeway This would complete the fencing of the bridges in the worst risk categories. We have required all bridges in the F3, M4, M5 and SH2 to be fenced to remove this case of risk.”

His recommendations included: “The problem of projectiles being thrown from bridges onto motorists was clearly demonstrated by incidents in 1995 and 1996. It is expected that the problem will not go away and the RTA should protect motorists to a reasonable level. Given the continuing risk and the media attention to these issues at the time it is recommended that the program of retro fitting screens to the higher risk bridges already identified on high speed roads should continue. The recommended program will be to treat approximately 13 bridges per annum, ie $1.0M, on the State Network and to request similar funding from the Federal Government. This program would, commencing 1998/99, complete the State outer freeways in 1998/99, National Highway in 2000/01 and remainder of the State Network for bridges (80km/h and over) in 2002/03. (Funding proposal attached).”

The forward estimates for 1998/99 showed total amounts that would have resulted in six of the nine bridges proposed for 1998/99 being on the F5, and the remaining three on the F3. There is an unexplained puzzle with these figures, as there were only eight relevant bridges on the F5, and three of them had already been screened by that stage.

A project status report as at 31 July 1998 identified the allocations for the current year as “not known” . It reported “Design work commencing on Monday 17 Aug 98 for Glenlee and Menangle Bridges.” Acceleration of the Screening Program

On the night of Sunday 9 August 1998, rocks were dropped from three separate bridges over the F5 in the Campbelltown area, including the Glenlee Bridge. One of the bridges from which rocks were dropped, the St Andrews Bridge, was one that had already been screened.

On the night of 10 August 1998 there was another incident in which a rock was dropped from the Glenlee Bridge.

On 11 August 1998 Mr Dunlop emailed various RTA officials, reporting on these facts, and saying: “1. Fence the bridges down there immediately. They were to have been done last year but Feds reduced funding. I have advised RNI today that we will not wait until funding approved, currently under review, but we will do what is needed now and then find money from where ever. 2. Police are not reporting incidents. We have only found out about the extent of Sunday’s incidents this morning. We will contact senior Police in Campbelltown area for better feedback …”

Design work commenced on 17 August 1998 for the fencing of the Glenlee and Menangle Bridges.

By 31 August 1998 investigation and design work was complete for the Glenlee Bridge, and by mid October 1998 the screening had actually been completed. It happened too late to protect Mr Evans, who had been killed on 23 August 1998. The Alleged Acts of Negligence

The acts of negligence alleged against the RTA in the Statement of Claim were: “(a) Exposing the worker to a risk of injury of which it knew or ought to have known and which could have been avoided by the use of reasonable care on its part. (b) Failing to instal[l] a barrier, fence guard or other appropriate structure to prevent or inhibit things falling or being thrown to the highway below. (c) Failing to properly inspect the area and detect the dangers presented by an unfenced overhead bridge. (d) Constructing or otherwise allowing the bridge to be constructed without barriers, fencing or guard rails. (e) Failing to take any or any adequate steps to deter people from throwing things off the bridge. (f) Failing to take any or any adequate steps to ensure the safety of motorists travelling on the highway below the bridge. (g) Failing to have in place any system for detecting persons throwing things from the bridge.” ISSUE 1 – DUTY OF CARE OWED BY RTA? Statutory Duties and Powers of RTA

The RTA is a corporation created by section 46 Transport Administration Act 1988 .

Under section 7 Roads Act 1993 the RTA is “the roads authority for all freeways” . It is common ground that the F5 is a public road, and a freeway within the meaning of section 7. The case has been conducted on the basis that the RTA’s powers and responsibilities concerning the F5 extend to the Glenlee Bridge.

Section 145(1) Roads Act provides that all freeways are vested in fee simple in the RTA.

Section 146(1) Roads Act provides: “ Nature of ownership of public roads (1) Except as otherwise provided by this Act, the dedication of land as a public road: (a) does not impose any liability on the owner of the road that the owner would not have if the owner were merely a person having the care, control and management of the road, and … (d) does not constitute the owner of the road as an occupier of the land …”

The Dictionary to the Roads Act provides: “ road work includes any kind of work, building or structure (such as a roadway, footway, bridge, tunnel, road-ferry, rest area, transitway station or service centre) that is constructed or installed on or in the vicinity of a road for the purpose of facilitating the use of the road as a road, the regulation of traffic on the road or the carriage of utility services across the road, but does not include a traffic control facility, and carry out road work includes carry out any activity in connection with the construction, erection, installation, maintenance, repair, removal or replacement of a road work.”

Section 71 Roads Act provides: “A roads authority may carry out road work on any public road for which it is the roads authority and on any other land under its control.”

Pursuant to section 5(1) Roads Act a member of the public is entitled, as of right, to pass along a public road “whether on foot, in a vehicle or otherwise” . That entitlement applied, in the present case, both to members of the public who wished to drive along the F5, and also to members of the public who wished to cross the Glenlee Bridge. RTA’s Duty of Care

While the terms of section 42 Civil Liability Act 2002 (set out at para [379] below) make clear that the matters to which it refers can be relevant to whether a duty of care exists, section 42 was not deployed in the present case as part of any argument about whether a duty of care existed. It is for that reason that the following discussion of duty of care does not include any reference to section 42.

In Brodie v Singleton Shire Council [2001] HCA 29; (2001) 206 CLR 512 at 577 [150], Gaudron, McHugh and Gummow JJ held that, under the common law of Australia, public authorities who had statutory powers of the nature of those conferred by the Local Government Act 1919 upon the local councils involved in Brodie : “… to design or construct roads, or carry out works or repairs upon them, are obliged to take reasonable care that their exercise of or failure to exercise those powers does not create a foreseeable risk of harm to a class of persons (road users) which includes the plaintiff. Where the state of a roadway, whether from design, construction, works or non-repair, poses a risk to that class of persons, then, to discharge its duty of care, an authority with power to remedy the risk is obliged to take reasonable steps by the exercise of its powers within a reasonable time to address the risk. If the risk be unknown to the authority or latent and only discoverable by inspection, then to discharge its duty of care an authority having power to inspect is obliged to take reasonable steps to ascertain the existence of latent dangers which might reasonably be suspected to exist.” (emphasis added)

The novelty of the majority decision in Brodie lay in its recognition that a highway authority could be liable in the tort of negligence for failure to exercise its statutory powers. It had been recognised since at least Geddis v Proprietors of Bann Reservoir (1878) 3 App Cas 430 at 455-6 that a statutory authority could sometimes be liable for negligence in the way it exercised its statutory powers. Gibbs CJ in Sutherland Shire Council v Heyman (1985) 157 CLR 424 at 443 confirmed that in principle failure to exercise statutory powers could also be the subject of an action in negligence. See also Mason J at 457-8. In particular, it had previously been recognised that a statutory authority which had the care and control of land to which the public had a right of resort could be liable in negligence for failure to exercise a statutory power concerning that land: Aiken v Kingborough Corporation (1939) 62 CLR 179; Schiller v Mulgrave Shire Council (1972) 129 CLR 116; Nagle v Rottnest Island Authority (1993) 177 CLR 423; Romeo v Conservation Commission (NT) [1998] HCA 5; (1998) 192 CLR 431. That there could be a duty of care in such a case was recognised long before the possibility of a statutory authority that did not have the care and control of any relevant land being liable for failure to exercise statutory powers was recognised in Anns v Merton London Borough Council [1978] AC 728. Cases concerning allegations of negligence against statutory authorities that did not have the care and control of any relevant land and who failed to exercise statutory powers (eg, Pyrenees Shire Council v Day [1998] HCA 3; (1998) 192 CLR 330; Crimmins v Stevedoring Industry Financing Committee [1999] HCA 59; (1999) 200 CLR 1; Graham Barclay Oysters Pty Ltd v Ryan [2002] HCA 54; (2002) 211 CLR 540; Stuart v Kirkland-Veenstra [2009] HCA 15; (2009) 237 CLR 215) are clearly distinguishable from the cases where the statutory authority has the care and control of land so far as the basis for the existence of a duty of care is concerned.

The decision of this Court in Amaca Pty Ltd v State of New South Wales [2004] NSWCA 124; [2004] Aust Torts Reports ¶81-749 (65,788); (2004) 132 LGERA 309; 1 DDCR 486 likewise concerned whether a duty of care was owed by a statutory authority with powers to inspect industrial premises that it did not occupy. However I note that Ipp JA (with whom Mason P and McColl JA agreed) at [22] (incorrectly numbered [21] in both the LGERA and DDCR reported versions) recognised that one of the categories where a duty of care might be attracted concerning a public authority’s exercise of a power was “where the particular circumstances of an authority’s … ownership or control of a structure attracts to it a duty of care . ”

The common law had crafted an exception, the “non-feasance” rule, for roads authorities from the usual situation in which a statutory authority having the care and control of land could owe a duty of care concerning failure to exercise its statutory powers. That exception had been accepted in Australian law: Buckle v Bayswater Road Board (1936) 57 CLR 259; Gorringe v Transport Commission (Tas) (1950) 80 CLR 357. It was in abolishing the non-feasance rule for road authorities that the novelty of Brodie lay.

Both in the court below and on appeal, the RTA accepted that it was subject to a duty of care of the type accepted by the majority judgments in Brodie . It contended, however, that the scope of that duty of care did not extend to exercising reasonable care to prevent road users suffering physical injury as a consequence of deliberate criminal action, such as was the immediate cause of Mr Evans’ death. In its argument both in the court below and on appeal, the decision of the High Court in Modbury Triangle Shopping Centre Pty Ltd v Anzil [2000] HCA 61; (2000) 205 CLR 254 played a large part.

In Modbury , the plaintiff (Mr Anzil) was a shop employee, who worked at night in a suburban shopping centre. He was assaulted after dark one night in the car park of the shopping centre by three unknown men, and suffered personal injuries. He sued the owner of the shopping centre on the basis, described by Gleeson CJ at 259 [1], that the owner: “… was the occupier of the car park; that, at the time of the attack, the car park lights were off; that, in the circumstances of the case, which will be set out in more detail below, the failure to leave the lights on was negligent; that the risk of harm of the kind suffered was foreseeable; and that the negligence was a cause of the harm.”

Gleeson CJ (with whom Gaudron and Hayne JJ agreed) proceeded, at 262-3 [13]-[18], on the basis that the duty alleged in Modbury fell outside the usual duty that an occupier of land owes to a person lawfully upon the land, to take reasonable care to avoid physical injury arising from the physical state or condition of the car park. At [14]-[15] he regarded it as important in deciding whether a duty of care existed to identify “the nature of the harm suffered by a plaintiff, for which a defendant is said to be liable” . He identified that harm as being “personal injury, the direct and immediate cause of which was the deliberate wrongdoing of the three men who attacked him.” He reasoned, at [17]: “The nature of the harm suffered was physical injury inflicted by a third party over whose actions the appellant had no control. Thus, any relevant duty must have been a duty related to the security of the first respondent. It must have been a duty, as occupier of land, to take reasonable care to protect people in the position of the first respondent from conduct, including criminal conduct, of third parties.”

At [19]-[21], Gleeson CJ rejected control as a possible basis for such a duty, because (at [19]): “The appellant had no control over the behaviour of the men who attacked the first respondent, and no knowledge or forewarning of what they planned to do. … The inference that they would have been deterred by lighting in the car park is at least debatable. The men were not enticed to the car park by the appellant. They were strangers to the parties.”

He rejected, at [22], reliance on the occupier to light the car park as a basis for the duty. Likewise, at [24]-[25] he rejected assumption of responsibility as a basis for the duty.

At [20], he had quoted the statement of Dixon J in Smith v Leurs (1945) 70 CLR 256 at 262 that: “It is, however, exceptional to find in the law a duty to control another’s actions to prevent harm to strangers. The general rule is that one man is under no duty of controlling another man to prevent his doing damage to a third. There are, however, special relations which are the source of a duty of this nature. It appears now to be recognized that it is incumbent upon a parent who maintains control over a young child to take reasonable care so to exercise that control as to avoid conduct on his part exposing the person or property of others to unreasonable danger.”

However, at [26], Gleeson CJ recognised that there could in some circumstances be a duty to take reasonable care to prevent the person’s criminal conduct injuring another: “Leaving aside contractual obligations, there are circumstances where the relationship between two parties may mean that one has a duty to take reasonable care to protect the other from the criminal behaviour of third parties, random and unpredictable as such behaviour may be. Such relationships may include those between employer and employee, school and pupil, or bailor and bailee. But the general rule that there is no duty to prevent a third party from harming another is based in part upon a more fundamental principle, which is that the common law does not ordinarily impose liability for omissions.” (citations omitted)

Another example of a relationship in which the law has recognised a duty to take reasonable care to prevent harm arising from criminal conduct was given by Hayne J in Modbury at [111], concerning a gaoler’s duty to a prisoner to take reasonable care to prevent assault by fellow prisoners.

A further example arises from the series of cases in this and other intermediate appeal courts that have recognised that the operator of a venue like a hotel or nightclub owes a duty to patrons to take reasonable care that they are not injured by the violent or unruly behaviour of other patrons, and that that duty extends to injury that is caused by criminal behaviour of other patrons: Chordas v Bryant (Wellington) Pty Ltd (1988) 20 FCR 91 (FC); Club Italia (Geelong) Inc v Ritchie [2001] VSCA 180; (2001) 3 VR 447; South Tweed Heads Rugby League Football Club Ltd v Cole [2002] NSWCA 205; (2002) 55 NSWLR 113; Wagstaff v Haslem [2007] NSWCA 28; (2007) 69 NSWLR 1; Spedding v Nobles [2007] NSWCA 29; (2007) 69 NSWLR 100; Collingwood Hotel Pty Ltd v O'Reilly [2007] NSWCA 155; Adeels Palace Pty Ltd v Moubarak [2009] NSWCA 29; [2009] Aust Torts Reports ¶81-997 (62,722).

In Modbury , Gleeson CJ returned, at [29], to control and knowledge, saying: “The control and knowledge which form the basis of an occupier's liability in relation to the physical state or condition of land are absent when one considers the possibility of criminal behaviour on the land by a stranger. … The unpredictability of criminal behaviour is one of the reasons why, as a general rule, and in the absence of some special relationship, the law does not impose a duty to prevent harm to another from the criminal conduct of a third party, even if the risk of such harm is foreseeable.”

As Ipp JA pointed out in Graham v Hall [2006] NSWCA 208; (2006) 67 NSWLR 135 at 144 [64], Modbury recognises that there are various different circumstances where there is a duty of care that extends to requiring the taking of reasonable care to prevent harm arising from criminal conduct. I adopt his Honour’s summary: “Gleeson CJ, with whom Gaudron J agreed, said (at 265 [26]) that while the general rule was that there was no duty to prevent a third party from harming another, ‘… there are circumstances where the relationship between two parties may mean that one has a duty to take reasonable care to protect the other from the criminal behaviour of third parties …’. Hayne J accepted that there could be a duty to control criminal conduct of others ‘in very restricted circumstances’ (at 293 [117]). Gaudron J said (at 270 [43]) that there were situations ‘… in which there is a duty of care to warn or take other positive steps to protect another against harm from third parties’. Her Honour remarked: ‘… Usually, a duty of care of that kind arises because of special vulnerability, on the one hand, and on the other, special knowledge, the assumption of a responsibility or a combination of both’. Callinan J said (at 302 [147]) that a duty could be imposed to take precautions to prevent or reduce the chances of criminally inflicted injury or loss by third party. But for such a duty to arise, ‘there must be something special in the circumstances, or the nature of the relationship between the plaintiff and the defendant’.”

In my view, it is significant that the type of harm suffered by the plaintiff in Modbury is a type of harm that could arise only through criminal conduct. In that respect it differs from the type of harm suffered by Mr Evans in the present case. A motorist travelling on a freeway could suffer physical injury as a result of an object falling from an overhead bridge and striking his or her vehicle even if no criminal conduct was involved. Such an event could happen as a consequence of part of a poorly secured load on a truck going over the overbridge becoming loose and falling over the edge, or as a result of some object that a pedestrian on the bridge was carrying or playing with accidentally going over the edge.

In Ashrafi Persian Trading Co Pty Ltd v Ashrafinia [2001] NSWCA 243; [2002] Aust Torts Reports ¶81-636 (68,314) the plaintiff had been assaulted while she was sleeping in the defendant’s premises when an unknown assailant struck her, from outside the premises, with an iron bar through a small gap in a sliding window. Heydon JA (with whom Mason P and Handley JA agreed) upheld an argument, summarised at [39](19), as being that there was no “duty owed by the appellant to the respondent to protect her from attacks by a third party or parties” . Ashrafi , like Modbury , was a case where the type of harm sustained by the plaintiff did not fall within that concerning which any conventionally recognised duty of care gave protection, and was a type of harm that could arise only through the criminal conduct of others.

The RTA has a unique role to play concerning any dangers of objects falling from an overpass onto cars on a freeway below. It was the RTA (which for this purpose need not be distinguished from its statutory predecessor) that constructed the freeway, a type of road calculated to encourage vehicles to travel fast and thereby be at potentially greater danger if hit by a falling object. It was the RTA that constructed the overpass, that provided the physical source of risk of objects dropping onto the freeway below. It was the RTA that had ownership of the freeway, pursuant to section 145(1) Roads Act . That ownership would extend to ownership of objects constructed in the superincumbent airspace, such as the Glenlee Bridge. Both that ownership, and the RTA’s powers of care, control and management, make it the only entity that has power to take steps to lessen the risk of objects dropping.

People driving on the freeway are vulnerable, in the sense that they have little or no capacity to protect themselves from the risk of an object falling onto their vehicle: cf Crimmins at 40-1 [100] (per McHugh J). All these matters show the sense of the general duty of care of a road authority, recognised in Brodie , applying to the RTA in relation to the risk of personal injury caused by an object falling from an overpass onto a freeway, regardless of the particular cause of the object falling.

Mr Maconachie QC, counsel for the RTA, submitted that before “control” could be relevant to the existence of a duty of care to protect someone from the actions of another, it had to be control over the person who was the potential source of the harm. He submitted that the undisputed fact that the RTA had control over the structure of the overhead bridge was not the relevant type of “control” that could give rise to a duty of care to protect the motorist from the actions of someone dropping objects from the bridge. He provides as instances Smith v Leurs (control of a parent over a child) and Home Office v Dorsett Yacht Co Ltd [1970] AC 1004 (control of a gaoler over an inmate).

I do not accept that submission. In those cases the relevant duty of the defendant was to take reasonable care that the child or the prisoner did not cause harm to someone in the plaintiff’s position. When, as in the present case, the essence of the duty owed is not to control the conduct of others, but simply to take reasonable care to prevent harm to road users, it is legitimate to treat a road authority’s control of a road or structure as a factor relevant to the existence of a duty of care to motorists who could be injured other than through the physical state of that road or structure.

It could not be the case that the duty of care of a roads authority concerning positive actions that it took did not extend to taking reasonable care to protect motorists from harm that resulted from criminal conduct of other motorists. A high proportion of serious motor accidents involving more than one vehicle would be ones in which one of the participants had engaged in criminal conduct, arising from his or her manner of driving. A significant proportion of those would be ones where the criminal conduct had been engaged in deliberately, such as by deliberate speeding, or knowingly driving while intoxicated or under the influence of a drug. If a motorist was injured partly as a consequence of the criminal manner of driving of another motorist, and partly as a consequence of the defective design construction or maintenance of the road itself, the usual result is that there is apportionment of responsibility between the other motorist and the roads authority. It is not that the roads authority does not owe any duty of care to the injured motorist. When that is the situation concerning the duty of care of a roads authority for what in pre- Brodie times would have been classified as misfeasance, it is hard to see why, now, such a duty should not be owed when there is a non-feasance. Particularly is that so when Brodie has demonstrated that the distinction between misfeasance and non-feasance is an unsatisfactory one for marking the distinction between situations where a duty of care is owed, and a duty of care is not owed.

In Modbury at 292 [110], Hayne J explained how the duty of care of an employer to take reasonable care to prevent an employee being robbed could exist. He said: “The duty which the employer breaks in such a case is not a duty to control the conduct of others. It is a duty to provide a safe system of work and ensure that reasonable care is taken” (citation omitted).

In a similar way, if it were the case that Mr Evans’ death resulted from a breach of a duty of care by the RTA, the duty breached would be the RTA’s general Brodie duty, not some exceptional duty to control the conduct of others.

There is another matter that supports the existence of a duty of care. It is that common experience of driving on Australian roads shows that when a road passes at the foot of a cutting, bank or hill, the roads authority sometimes takes steps to protect motorists on the road from the risk of rocks falling through natural forces onto the road surface. Sometimes the action taken is a warning about the risk of falling rocks (presumably given so that motorists keep a look out for rocks that might already have fallen onto the road surface, and so far as possible avoid the danger zone rather than increase the risk to themselves by stopping there), sometimes it takes the form of fencing the dangerous slope. For a roads authority to take such precautions concerning a known and sufficiently serious risk of rocks falling onto a road’s surface accords with the taking of reasonable care for the safety of motorists on the road. The fact that a risk of rocks falling onto a road’s surface arises through deliberate human action, rather than the operation of natural forces, does not seem to be a fundamental difference.

In my view, the duty of care that the RTA owes, of the type recognised by the High Court in Brodie , has a content that does not exclude taking reasonable care to protect a motorist from the criminal actions of another.

While the fact that an object’s falling has been caused by criminal conduct is clearly important in deciding whether the RTA has a liability in negligence concerning it, the criminality of the dropping of the object enters the analysis at the levels of breach of duty, and causation of damage, not at the level of existence of a duty of care. ISSUE 2 – BREACH OF DUTY? Issue 2(a) – Breach of Duty at Time of Construction of Overpass?

The judge identified, at para [15] of his judgment, one of the ways in which the Respondent put its case as being: “… at the time in 1977 when the Glenlee Bridge was constructed the then Department of Main Roads, the defendant’s predecessor, was aware from published experiences in the United States of America that there was a risk of objects being launched from bridges passing over high-speed roads so that it ought to have screened or fenced the bridge during the construction phase – it did not and the readily foreseeable risk eventuated.”

The other way in which the Respondent put its case was that later experience, in the 1990s, of objects actually being dropped from bridges over freeways, should have led the RTA to retro-fit protective screens to the Glenlee Bridge well before the time of Mr Evans’ death.

Mr Sheldon, counsel for the Respondent, submits that the trial judge accepted both ways in which the Respondent put its case below. Deciding whether that submission is correct involves a close examination of his Honour’s judgment. The 1969 American Publication

The sole evidence of “published experience in the United States of America” related to a pamphlet entitled “A Guide for Protective Screening of Overpass Structures” that was published in 1969 by the American Association of State Highway Officials. A copy of that document found its way into the library of the RTA’s predecessor, the DMR, in September 1970. The document said it was “intended to assist the Engineer in determining the need for a protective screen on overpass structures with suggested screening methods” . It stated: “With the advent of extensive freeway development in the United States the problem of objects being thrown or dropped from overpass walks onto vehicles travelling below has become a serious one in some areas. For instance, in the Chicago area in the two year period 1964-1965 there were over 1,200 reported incidents of objects being dropped or thrown from overpasses onto vehicles below. These findings were supported in a more recent study covering the period July 1968 to March 1969 during which 357 incidents were reported. It is logical to assume in addition that there were many more incidents not reported. The seriousness of the consequences that can result from objects being thrown from bridges is brought out by scattered and continuous country-wide reports of fatalities and major injuries caused by this type [of] vandalism. Reports of large objects falling through windshields and fatally injuring occupants of vehicles have been received from various States.”

It identified the nature of the problem posed by such incidents as follow: “The act of dropping objects from bridges may be wilfully malicious by adults or just mischievous by children. There is no practical device or method yet devised that can be universally applied to prevent a determined individual from dropping an object from an overpass. Small objects can be dropped through screens. Solid plastic enclosures have been suggested but these are expensive, may be insufferably hot in summer, and may have the undesirable feature that they obscure and darken the pedestrian travelway, which is conducive to other forms of criminal activity. It is also a known fact that any type of physical preventive measure acts as a motivating challenge to some individuals.”

The pamphlet considered better police surveillance and better education of the public to be of some, though only limited, use. It said: “It is not now possible to establish useful location warrants for when and where barriers should be installed to discourage the throwing of objects from structures. The general need for economy in design and desire to preserve the clear lines of a structure unencumbered by screens must be carefully balanced against the absolute requirement that the highway traveller and overpass pedestrian be provided maximum safety.”

It identified various types and configurations of screens that could be built on overpasses. Its conclusions were: “Screens of various types can definitely reduce the number of incidents of objects being dropped from overpasses and striking vehicles below. However, there can be no absolute warrants for installation of screens or barrier protection. Each location must be analysed individually. In some cases erection of a protective screen can be postponed until such time as there are indications of need since an effective screen can be readily installed at any time. However, where there is any doubt as to their need, screens should be erected initially. The addition of a screen or barrier should not be delayed until a serious incident definitely establishes a need for such a device. Locations where screens definitely should be considered at the time the overpass is constructed are as follows: 1. On an overpass near a school, a playground or elsewhere where it would be expected that the overpass would be frequently used by children unaccompanied by adults. 2. On all overpasses in large urban areas used exclusively by pedestrians and not easily kept under surveillance by police. 3. On overpasses with walks where experience on nearby structures indicates a need for screens. Screens should also be installed on existing structures where there have been incidents of objects being dropped from the overpass and where it seems evident that increased surveillance, warning signs or apprehension of a few individuals involved will not effectively alleviate the problem.”

There is no evidence of this publication being known to, or endorsed by, Australian engineers involved in freeway design or maintenance prior to 1995. In 1995 it became known to Mr Dunlop. It appears to be one of the sources on which he drew for his report of 14 November 1995 (para [32] above).

The Respondent’s expert witness, Mr Jamieson, is a consulting engineer. His initial qualification was a Bachelor of Surveying, obtained in the period 1971-1975. He then obtained a Master of Engineering Science (Traffic and Transportation) in 1976. After completion of that degree he worked for the NSW Traffic Accident Research Unit, involved with all aspects of crash investigation and traffic safety.

Mr Jamieson gave two reports. The first, dated 18 August 2005, he described as “an analysis of a crash from an engineering/scientific viewpoint” . He stated that in his experience “since 1976” the design review process of a bridge involves issues concerning structural integrity, architectural issues, and operational/environmental safety issues. He said that in his experience: “… there is a long period of negotiation commencing at the initial submission of a preliminary bridge design between the civil/structural and architectural teams, and the operational teams.”

He gives “typical examples of such negotiation (from the present author’s experience)” . Those examples do not include anything to do with fencing the sides of a bridge to prevent projectiles being dropped from it. He continues: “In the examples above, the civil/structural-architectural team often presents the operational/safety team with conflicting objectives. That is, there may be traffic safety issues with the bridge, despite the fact that the bridge is perfectly structurally sound. It is considered that railing barrier fencing (preventing projectiles being thrown off) is one of these conflicting design elements between the structural-architectural team versus the operational/safety teams.”

The last-quoted paragraph recognises that the nature or extent of bridge fencing might be a topic of debate in the course of bridge design, but expresses no view about what the outcome of that debate ought be.

He also says: “The actual act of a vandal throwing an object from a bridge deck is a matter of ‘common knowledge’ and probably has been occurring before Roman times. Therefore, such an event is totally foreseeable from every bridge.”

That report thus provides no basis of expert opinion for a conclusion that the taking of reasonable care in the design and construction of the Glenlee Bridge in the 1975-1978 period would require more extensive screening of the sides of the bridge than was in fact constructed.

Mr Jamieson gave a second report, dated 24 March 2006, based upon documents that had been produced from the RTA’s files. One of those documents was the 1969 American publication. His second report does not indicate that he had any personal familiarity with that document before it was produced to him. His second report does not express any view about whether the taking of reasonable care in the design and construction of the Glenlee Bridge, at the time it was designed and constructed, would have required the installation of protective screening.

Mr Jamieson had put to him in cross-examination that in 1975 and 1977: “Q. … The constructing authority was entitled in taking its decision to go ahead with a bridge that did not have a screen, on the fact that its experience revealed that there was no complication, no adverse effect to motorists. From the experience, dating back to the first freeway, of these bridges, there had been no problem. A. Yes. Well, in the previous 7 or 8 years, I suppose, yes.”

Professor Robert White, a criminologist called by the RTA, was unaware of the American publication, and did not know of anything else along those lines. The Judge’s Findings Concerning Screening at Time of Construction

His Honour stated, at [19]: “Mr Jamieson said the F5 was built according to standards laid down by the National Association of Australian State Road Authorities with parameters originally derived from American parameters. Those standards from America, specifically the Association of Highway and Traffic Engineers, were applied since the first freeway in New South Wales from Berowra to the Hawkesbury River (the F3) was constructed in 1968-69. Thus, the connection with American standards and experience as to freeways was established.”

He also said, at [25]: “Given that the American report meant the risk of objects being thrown from overbridges could not be ignored, Mr Jamieson’s evidence was that the addition of an appropriate screen on the Glenlee Bridge when constructed in 1977 would have been less complicated and involved than ‘retro-fitting’ a screen in 1998; he said a marginal amount of extra time would have been involved in installing a fence or screen during the initial construction of the bridge and the extra cost would have been ‘a very marginal amount, an insignificant amount.’”

After finding (at [60]) that “the defendant” had a copy of the American publication from September 1970 at the latest, and that it was not “until November 1995 at the earliest that the defendant became relevantly aware of the American guide” the judge made at [60](8) what he called a finding of fact: “The addition of an appropriate safety screen on the Glenlee Bridge would have been less complicated and involved if fitted during the initial construction in 1977 than being “retro-fitted” in 1998 and for extra cost of a marginal or insignificant amount; it would have been reasonable and consistent with the American experience to have screened the bridge when constructed in 1977.”

He also found, at [60](14): “The erection of screening fences or barriers on overbridges, consistent with the American guide from its experience, was an appropriate means to address the safety risks involved with overbridges for users of the freeways; it would have been a reasonable measure, particularly in remote rural areas, so as to avoid or, at the least, make less likely, injury to those users.”

His Honour’s ultimate finding of breach, at [109] was: “I am satisfied that the risk posed to users of freeways by an unscreened overbridge, such as the Glenlee Bridge, from objects being thrown or dropped from it was real, known and foreseeable. The appropriate and reasonable response by a freeway authority, such as the defendant, was to fence or screen the overbridge, as in fact the Glenlee Bridge was by 17 October 1998. The failure by the defendant over very many years beforehand to do so, in my view, represents a failure to take reasonable care. The count of negligence is made out.” Decision on Breach of Duty at Time of Construction

In my view it is far from clear that the judge has made his finding of negligence on a basis that includes a failure to install protective screening at the time of construction of the bridge. The findings in para [60] (8) and (14) about what was “reasonable” or “appropriate” do not clearly state that they are findings that the DMR failed to take reasonable care by not taking those “reasonable” and “appropriate” steps. There are many things that it might be reasonable (according to some criterion or other) for a person, corporation or statutory authority to do, but concerning which the failure to do such a thing does not amount to a failure to take reasonable care, according to the test appropriate in the law of negligence. The judge is quite unspecific about what he means in para [109] by failure “over very many years beforehand” to screen the overbridge, and in particular it is not clear that “very many years” extends to more than 20 years. There is no Notice of Contention submitting that the judge should have found negligence on the basis of failure to install protective screening at the time of construction of the bridge.

However, if the judge had made his finding of negligence on a basis that included failure to install protective screening at the time of construction, in my view he would have been mistaken to do so.

There is an evidentiary basis for the judge’s finding that the F5 was built according to standards derived from America. Mr Jamieson’s evidence was: “Q. Now, can I ask you when was the first freeway in New South Wales completed? A. I understand in 1968–69. That’s the Berowra to Hawkesbury River F3. Q. To what type of standards was that road built? A. To an American freeway standard. Q. Where did the American freeway standards come from? A. The Association of Highway Engineers. Q. Does that association have an acronym? A. ASHE, now called ASHTO. Q. What does the T in ASHTO stand for? A. Either Traffic or Transport; I can’t recall just at the moment. Q. All right. Now, was the F5 built to any particular standards, leaving aside the Glenlee Bridge? A. By the time the F5 was built, they would have derived their standards from the National Association of Australian State Road Authorities, NAASRA, now called Austroads. The parameters of NAASRA documents, as I understand them, were originally derived from American parameters.”

However, the connection with American standards that emerges from that evidence does not suffice to show, or even contribute to a conclusion, that the DMR had been negligent in not incorporating screens in the original bridge design. Mr Jamieson does not say that any NSW freeway was built in accordance with standards laid down by the American Association of State Highway Officials (the body that published the 1969 publication). Nor does he say anything about those American standards that were adopted or drawn on in construction of the F3 and the F5 having any element relating to the screening of overpasses.

In so far as the 1969 American publication contains a “standard” , it is: “In some cases erection of a protective screen can be postponed until such time as there are indications of the need since an effective screen can be readily installed at any time. However, where there is any doubt as to their needs, screens should be erected initially.”

That statement is inconsistent with good practice always requiring incorporation of screening at the time of construction. In the present case it has not been shown that there were any “indications of the need” at the time of construction of the Glenlee Bridge.

More fundamentally, the judge’s finding (set out at [162] above) that it would have been a reasonable measure to install screening on overpasses does not address the real question that is involved in a question of breach of duty, namely whether by acting (or, here, not acting) in some particular way the defendant has failed to take reasonable care to prevent harm to the plaintiff or members of a group to which the plaintiff belongs. The Legal Test for Breach of Duty

Section 5B Civil Liability Act provides: “(1) A person is not negligent in failing to take precautions against a risk of harm unless: (a) the risk was foreseeable (that is, it is a risk of which the person knew or ought to have known), and (b) the risk was not insignificant, and (c) in the circumstances, a reasonable person in the person’s position would have taken those precautions. (2) In determining whether a reasonable person would have taken precautions against a risk of harm, the court is to consider the following (amongst other relevant things): (a) the probability that the harm would occur if care were not taken, (b) the likely seriousness of the harm, (c) the burden of taking precautions to avoid the risk of harm, (d) the social utility of the activity that creates the risk of harm.

Section 5B is not a self-contained statement of the circumstances in which a liability for negligence will arise. Rather, subsection 1 sets out three preconditions that must co-exist before a liability in negligence arises, when the type of negligence alleged is failure to take precautions against a risk of harm arising. As to whether such cases cover the entire field of negligence cases, see Drinkwater v Howarth [2006] NSWCA 222 at [11]-[13], [24]. Subsection 2 provides a non-exhaustive list of factors the court is required to take into account in deciding whether the third of those preconditions exists. Section 5B presupposes the existence of the law of negligence, and operates against its background.

In Waverley Council v Ferreira [2005] NSWCA 418; [2005] Aust Torts Reports ¶81-818 (68,074) at [45], Ipp JA (with whom Spigelman CJ and Tobias JA agreed) said that the “matters set out in s 5B(2), in substance, are a reiteration of Mason J’s remarks in Wyong Shire Council v Shirt (1980) 146 CLR 40 at 47-48 .” Ipp JA explained at [47] that the purpose of enactment of section 5B(2) was so that the factors there identified “… should be given statutory force so that courts would focus more directly on the issue ‘whether it would be reasonable to require precautions to be taken against a particular risk’ and to avoid conflation of the concept of foreseeability of risk with the conclusion that a reasonable person would have taken precautions against it.”

At [50] his Honour rejected the notion that section 5B(2)(d) added anything novel to the law, saying: “It simply gives expression to the idea that some activities are more worth taking risks for than others.”

At [51] Ipp JA said: “Section 5B(2) provides a framework for deciding what precautions the reasonable person would have taken to avoid the harm and involves weighing the factors set out in ss 5B(2)(a) and (b) against those in ss 5B(2)(c) and (d) (subject, of course, to each being applicable in the particular circumstances of the case).”

While section 5B(2) articulates clearly the process of reasoning that would have been involved under the common law in application of the test in Shirt for identifying what steps the taking of reasonable care requires, it does not show any intention to alter the common law on that topic. That view of section 5B(2) has been applied previously in this Court: Council of the City of Liverpool v Turano [2008] NSWCA 270; (2008) 164 LGERA 16; 51 MVR 262 at [171] per Beazley JA (with whom Hodgson JA substantially agreed), [362] per McColl JA. The Shirt Test for Breach of Duty

Under the Australian common law unmodified by statute, the test laid down in Wyong Shire Council v Shirt (1980) 146 CLR 40 at 47-8 continues to govern whether there has been a breach of a duty of care. There, Mason J (with whom Stephen and Aickin JJ agreed) said that, if a reasonable person in the defendant’s position would have foreseen that the conduct complained of involved a risk of injury to the plaintiff or a person in a similar position: “… it is then for the tribunal of fact to determine what a reasonable man would do by way of response to the risk. The perception of the reasonable man’s response calls for a consideration of the magnitude of the risk and the degree of the probability of its occurrence, along with the expense, difficulty and inconvenience of taking alleviating action and any other conflicting responsibilities which the defendant may have. It is only when these matters are balanced out that the tribunal of fact can confidently assert what is the standard of response to be ascribed to the reasonable man placed in the defendant’s position.”

The ongoing relevance of that passage has recently been confirmed by Gummow J (with whom Heydon J agreed) in Roads and Traffic Authority (NSW) v Dederer [2007] HCA 42; (2007) 234 CLR 330 at 338 [18] and 353-4 [68]-[69].

In New South Wales v Fahy [2007] HCA 20; (2007) 232 CLR 486 an invitation for the High Court to reconsider Shirt was not acceded to, though in unusual circumstances of precedent. An appeal from this Court was allowed, by a 4:3 majority (per Gummow, Hayne, Callinan and Heydon JJ, Gleeson CJ, Kirby, and Crennan JJ dissenting). Gleeson CJ implicitly approved Shirt at 490-1 [6], Gummow and Hayne JJ declined to reconsider it at 504-5 [56]-[57], 511 [78]-[79], as did Kirby J at 523-8 [119]-[133]. Heydon and Callinan JJ made obiter remarks disapproving it at 550-4 [213]-[227], while Crennan J held that the case was not a suitable vehicle for reconsidering it, at 556 [241]. The fact remains that the High Court has recently been asked to reconsider the Shirt test, and declined to do so.

In any event, the attack that counsel in Fahy had made on Shirt was on its “undemanding” test of foreseeability by reference to what is not “far fetched or fanciful” . (That test now in any event needs to be read in the light of section 5B(1) Civil Liability Act – see para [186] below). It was not an attack on how a court should go about establishing whether there has been a breach of duty, once foreseeability of the relevant type of harm is established.

Contemporary sensitivities would probably require Shirt’s “reasonable man” to become explicitly gender neutral, but that is not a change in substance.

At 354 [69] in RTA v Dederer , Gummow J said: “What Shirt requires is a contextual and balanced assessment of the reasonable response to a foreseeable risk.”

Similarly, in Graham Barclay Oysters at 585 [106], McHugh J said: “The duty of care owed by a manufacturer or producer to a consumer is a duty to take reasonable care to avoid injury to the consumer. To formulate the duty in more specific terms invites error because it is likely to mix a question of law (whether a duty existed) with a question of fact (whether a breach occurred). If the duty is formulated in specific terms, the issue on breach is whether the duty has been performed in accordance with the terms of the duty as formulated. But, as Wyong Shire Council v Shirt at 47-48 shows, the question of breach is far more complex than an affirmative or negative answer to the question whether the defendant carried out the duty as formulated. It involves evaluating and weighing a number of competing considerations.” (citations omitted)

The concept was explained in a little more detail by Lord Reid in Overseas Tankship (UK) Ltd v Miller Steamship Co Pty ( Wagon Mound (No 2) ) [1967] 1 AC 617 at 642E-3A; (1966) 67 SR (NSW) 21 at 30; [1966] 1 NSWR 411 at 418. His Lordship, after referring to the decision in Bolton v Stone [1951] AC 850 said, in a passage quoted with approval by Gibbs J in Schiller v Mulgrave SC at 135: “But it does not follow that, no matter what the circumstances may be, it is justifiable to neglect a risk of such a small magnitude. A reasonable man would only neglect such a risk if he had some valid reason for doing so, eg, that it would involve considerable expense to eliminate the risk. He would weigh the risk against the difficulty of eliminating it ... In their Lordships’ judgment Bolton v Stone did not alter the general principle that a person must be regarded as negligent if he does not take steps to eliminate a risk which he knows or ought to know is a real risk and not a mere possibility which would never influence the mind of a reasonable man. What that decision did was to recognise and give effect to the qualification that it is justifiable not to take steps to eliminate a real risk if it is small and if the circumstances are such that a reasonable man, careful of the safety of his neighbour, would think it right to neglect it.” (citations omitted)

As I have said earlier, Mr Jamieson’s report does not provide a basis of expert evidence for the conclusion that screens should have been fitted at the time of construction of the bridge. Even accepting that the American publication was in the library of the DMR at that time, there is no evidence that, whatever the situation might have been in the USA, the problem of objects being dropped from overpasses onto freeways was a real one in Australia in the 1975-1978 period. One can say, in fairly abstract terms, that it is always predictable that someone might drop an object from a bridge, and the possibility of that happening is not far-fetched or fanciful (or not insignificant, if section 5B(1) Civil Liability Act imposes a test that is any different in substance: Drinkwater v Howarth [2006] NSWCA 222 at [13]-[16], [25]; Harriton v Stephens [2006] HCA 15; (2006) 226 CLR 52 at 94 [138]; Pollard v Trude [2008] QSC 119 at [39] (considering the equivalent section in Queensland: section 9 Civil Liability Act 2003 (Qld)), affirmed in Pollard v Trude [2008] QCA 421 at [17]; Review of the Law of Negligence Final Report , (September 2002), (the “ Ipp Report ”) at [7.15] (p 105)). However, before a roads authority is negligent for having failed to screen the bridge it must have failed to take a step that a reasonable roads authority would have taken. There is no basis for concluding that a reasonable roads authority in the 1975-1978 period in Australia, would have fenced a bridge like the present one. Thus, if the judge had made his finding of negligence on the basis of failure to install protective screening at the time of construction of the bridge, such a finding would in my view be erroneous, and I would set it aside, and on reconsidering the question of breach find it had not been established. Thus, the balance of this judgment can concentrate on whether there was negligence in the RTA failing to retro-fit protective screens to the Glenlee Bridge prior to Mr Evans’ death.

I have made no mention of section 42 Civil Liability Act in this discussion of breach of duty at the time of construction of the overpass because it played no part in the argument on this topic. Issue 2(b) – Breach of Duty in Not Retrofitting Screens?

The trial judge decided whether there had been a breach of duty, in not retrofitting screens, by reference to the common law unmodified by section 42 Civil Liability Act . He took that course because he found that section 42 was inapplicable. Of course, in any case to which section 42 is applicable, giving consideration to it is a necessary part of making any finding on the topic of breach of duty. However, it is convenient in the present case to take the same course the trial judge took and consider whether, under the common law, there was a breach of duty in not retrofitting screens. I will consider later, at para [378] ff, whether the judge was right in treating section 42 as inapplicable. I can say now that the judge was wrong in treating section 42 as inapplicable, but that the application of section 42 makes no difference, in this particular case, to the conclusion I would arrive at by applying the common law. Issue 2(b)(1) – The Factual Findings Challenged

In para [60] of the judgment below, the judge made various statements that he described as “findings as to the background facts” . They included the following: “(10) The [RTA’s] response [to the incidents of objects being dropped from overbridges in 1992 and 1993] that the provision of preventative screening was not justified at that time was an inappropriate and unreasonable response in light of the emerging incidents, particularly with the American experience; even more so was the [RTA’s] view that there was little it could do and it was a matter for the police. (11) The [RTA’s] statement that the enclosure of bridges with wire mesh was only practicable with pedestrian bridges and that the screening of vehicular bridges may impose problems was an unreasonable denial of the safety issues involved and contrary to what could have been done as shown by later measures to screen bridges. … (13) The [RTA’s] policy up to mid 1995 not to erect fencing or barriers on overbridges represented an unreasonable disregard for what was occurring in practice. (14) The erection of screening fences or barriers on overbridges, consistent with the American guide from its experience, was an appropriate means to address the safety risks involved with overbridges for users of the freeways; it would have been a reasonable measure, particularly in remote rural areas, so as to avoid or, at the least, make less likely, injury to those users. … (16) The [RTA’s] initial attention of screening State roads and not national roads such as the F5, but for which it still maintained responsibility, unreasonably ignored a real and potential risk of very busy freeways passing under bridges in remote areas. … (18) The [RTA’s] delay in effecting that decision to screen bridges in the 1996/97 year was relevantly unexplained, other than in terms of funding issues with increasing costs in doing so. However, the general allocation of $27.36 million in 1996/97 for safety and urgent minor works, which included bridge screening, was underspent by $3.26 million – there was no application to carry forward that amount and where in that year the cost of screening was $50,000 for each bridge at a total cost of $2.2 million. (19) The [RTA’s] methodology in determining a priority rating for the screening of bridges was flawed in that it gave a high weighting to pedestrian access and to a history of incidents/graffiti as distinct from one or two incidents; the matrix was thus biased in favour of bridges near schools, hotels or clubs, other pedestrian generators and with exposure to buildings as against bridges over high-speed, restricted access roads on the outer urban freeways – such as the Glenlee Bridge and characteristic of those bridges with the highest risk. (20) The resultant priority for the Glenlee Bridge so as to be due for screening in 1999/2000 was unreasonably delayed. That was particularly so where in 1996/97 the underspent funds were directed to noise abatement works rather than the safety issue of bridge screening. (21) The 1997/98 budget for safety and urgent minor works was $25.99 million but it was an amount indicated by the federal government from the prior year’s amount. There was no evidence why specific funding for bridge screening was not sought or why there was no application to use the $3.26 million underspent from the 1996/97 budget. And that was in a context where the [RTA] had bids from its regions for a total amount for 1997/98 for safety and urgent minor works of $41.0 million but with no apparent co-ordination of the need for bridge screening so that the $41.0 million was simply pared down to $25.99 million - that is indicative of a budgetary process not in accordance with the real and reasonable needs, particularly as to safety considerations.”

When the judge came, at [108]-[109] to consider expressly the question of breach of duty he said that it could be “dealt with quite shortly by reference to the factual findings I have made earlier . ”

Insofar as these paragraphs contain findings of unreasonableness, and assuming in favour of the judge that the type of unreasonableness he is talking about is a failure to take reasonable care to prevent personal injury to road users, the findings are made without going through the process of reasoning described by Mason J in Shirt at 47-8 (quoted at para [178] above).

As well, at [26] the judge listed 24 incidents of objects being thrown or dropped from overbridges in the Sydney region in the period 1992 to April 1998. That list was derived from a chronology prepared by the RTA. It does not include all of the incidents that occurred in NSW and that I have listed at para [24] above (which total at least 29 in number). However concerning those incidents that he listed, the judge said, at [27], “Of these twenty-four incidents, eight occurred in the Campbelltown area.” It is impossible to derive that conclusion from his Honour’s list. Rather, from that list one can conclude only that three incidents occurred in the Campbelltown area, one in February 1993 and two in January 1996. The judge later said, at [32], that the F5 was “also known as the M5” . That statement is wrong – the F5 and the M5 join each other, but are separate freeways. Mr Fishburn gave evidence that “it changes its name from M5 to F5 depending whether it's tolled or not.” The lists of incidents that the judge gave at [26] includes five incidents on the M5. It is his mistaking the M5 for the F5 that led, it seems, to his conclusion that 8 of the 24 incidents occurred in the Campbelltown area. That mistaking of the extent of the risk in the Campbelltown area would make the judge’s assessment of what was a reasonable response to the risk unsound. However, the judge’s failure to apply the Shirt process of reasoning is sufficient by itself to lead to his conclusions about unreasonableness not being upheld.

In these circumstances the Court must consider the question of whether there has been a breach of duty for itself. Reasonable Foreseeability

By 1995 it was not only reasonably foreseeable, but actually foreseen by the relevant officers in the RTA that motorists might be killed or injured by people dropping objects from overhead bridges on freeways. They expressly recognised that the people who dropped such objects were “criminals” , and applied their minds to what the RTA should do to lessen the risk of such criminal activity. Thus, whether there has been a breach of duty depends on what is the reasonable response to the risk. Characterisation of the Risk

In RTA v Dederer at 351 [59]-[60], Gummow J emphasised the importance of identifying the particular risk that was foreseeable. As his Honour said at 351 [59]: “It is only through the correct identification of the risk that one can assess what a reasonable response to that risk would be.”

At 351 [60], his Honour held that it had been an error for this Court in Dederer to characterise the risk as being “serious spinal injury flowing from the act of diving off the bridge” . Rather, “the true source of potential injury … arose not from the state of the bridge itself, but rather from the risk of impact upon jumping into potentially shallow water and shifting sands of the estuary” . His Honour saw this characterisation of the risk, at 351 [61] as important because it led one to concentrate not on the comparatively large numbers of people who had jumped or dived from the bridge, but on the fact that none of them had previously been injured.

In the present case, the relevant risk is that of a vehicle on the freeway being struck by a falling object.

The consequences of that risk materialising might be a bad fright for motorists involved, might be property damage, might be personal injury, or at worst might be death.

The evidence in this case about the occurrence of objects being dropped from an overpass onto a freeway shows that, in the period from 1990 to mid-1998 inclusive, there had been one death from such a cause in Australia. The ministerial briefing note of 24 January 1996 (para [46] above) reported that there were approximately five such incidents in the Sydney region per year, and that there had been seven incidents during the previous year, three of which resulted in injuries. On the information available to the RTA, most involved opportunistic rather than premeditated behaviour.

The RTA had, by the time of that briefing paper, considered various alternative ways of dealing with the problem, and found all except screening to be either counterproductive or impractical (para [48] above).

The significance of the incidents of objects being thrown from overpasses needs to be seen in the context of the overall picture of accidents on NSW roads.

RTA summary data for the year 1998 relating to all accidents in New South Wales showed that there were a total of 52,575 recorded accidents in the year, of which 56 related to “load or missile struck vehicle” . Mr Fishburn gave evidence, on which he was not challenged, “… that probably somewhere in the region of half that number would have occurred on an arterial road, remembering that that figure there is all over the State. And so this type of thing includes items falling off the backs of trucks; items on vehicles that fall off …”

As well, there were 158 incidents where a vehicle struck an object on the carriageway, which could include some objects that had come from overhead bridges.

Even if one makes the extravagantly conservative assumption that all of the 56+158=214 incidents where a load or missile struck a vehicle, or a vehicle struck an object on the carriageway, were ones connected with objects thrown from overpasses, that is still 0.4% of the total of 52,575 accidents occurring on NSW roads.

The 1998 NSW road accident figures included 491 fatal accidents, 19,667 “injury accidents” , and 32,417 “non-casualty accidents” . Those same statistics show that 556 people were killed, and 26,415 people were injured on NSW roads in the year. None of the people killed were killed through incidents involving objects thrown from overpasses.

The corresponding 1997 figures for accidents were 525 fatal accidents, 18,327 injury accidents, and 31,268 non-casualty accidents, to give a total of 50,120 recorded accidents. The corresponding 1997 figures for casualties were 576 people killed, and 24,454 people injured, to give a total of 25,030 casualties.

Of course, as well as the extremely small or nil proportion of injuries or deaths that had arisen from objects being dropped from overpasses on freeways, a reasonable roads authority would take into account that it had the capacity to achieve a reduction in the number of objects dropped from overpasses by screening the overpasses, while many road accidents would be ones that had not been contributed to by the condition of the road or its surroundings, in a way that was remediable by the RTA. It would also take into account that it was not only the presence of the freeway and the overpass that created the conditions for the risk materialising, but as well actual materialisation of the risk frequently involved deliberate conduct by people who the roads authority had no right and little practical ability to control. Indeed, those people had a statutory right to be on the overpass, at least for the purpose of passing over it. All the roads authority could do was to create an impediment to those people throwing objects over the edge by putting a physical obstacle in their way.

A reasonable road authority would also take into account that, while screening overpasses would create a very significant obstacle to objects being deliberately dropped onto freeways, it would not completely eliminate that risk. There had been at least one incident in which a tree planting stake had been dropped onto a vehicle after screens were erected (para [30] above) and, as demonstrated by experience on the M2 (para [30] above) and at the St Andrews Bridge on 9 August 1998 (para [102] above) screening of overpasses did not make it impossible to drop rocks from them. Experience shows that rocks could be lobbed onto freeways from the sides of the freeways, sometimes even by children (para [25] above).

As well, and particularly important, the cost of screening was significant. Even at the initial cost estimate of $50,000 per bridge, screening all 231 bridges where there was the potential to drop rocks would cost $11,550,000. At the ultimate cost of $75,000 per bridge, screening 231 bridges would cost $17,325,000.

On one version of an evaluation matrix for prioritising bridges that is of particular relevance in this case (see paras [414] and [430]) funding 22 bridges would have been just enough to achieve the screening of the Glenlee Bridge, if bridges were screened in that order of priority. Screening even 22 bridges would cost $1,100,000 at $50,000 per bridge, or $1,650,000 at $75,000 per bridge. The availability of funding was a matter of considerable importance to an assessment of what the reasonable response to the foreseen risk would have been. I will first consider what guidance previous cases provide concerning the role of funding limitations in a case like the present. Issue 2(2) – The Significance of RTA’s Financial Limitations to Breach of Duty

Mason J’s explanation in Shirt at 47-8 of how to decide whether there has been a breach of duty includes the “expense … of taking alleviating action and any other conflicting responsibilities which a defendant may have” among the elements to be taken into account in deciding what is the reasonable response to the risk. Brodie on Role of Financial Considerations in Deciding Breach

The role of financial considerations impacting on a roads authority in deciding whether it has breached its duty of care was considered to some extent in Brodie .

The joint judgment in Brodie said, at 559-60 [104]: “… financial considerations and budgetary imperatives may fall for consideration with other matters when determining what should have been done to discharge a duty of care: Pyrenees Shire Council v Day (1998) 192 CLR 330 at 394-395 [183]-[184]. … [The position] is that advocated in this Court a century ago. In Miller v McKeon (1905) 3 CLR 50 at 60, Griffith CJ said: ‘So the Government of a newly-settled country, which undertakes the first formation of a road, whether the soil has or has not been formally dedicated as a highway, is bound to use such care to avoid danger to persons using it as is reasonable under all the circumstances. These circumstances include the nature of the locality, the extent of the settlement, the probabilities as to the persons by whom the road is likely to be used, and the moneys available to the Government for the purpose.’ Each element in these sentences merits careful attention. Evidence respecting funding constraints and competing priorities will be admissible. See Hill v Commissioner for Main Roads (1989) 68 LGRA 173 at 181; Gloucester Shire Council v McLenaghan (2000) 109 LGERA 419 at 423; cf Woodward v Orara Shire Council (1948) 49 SR (NSW) 63 at 65-67.”

Their Honours continued, at 560 [105]: “The public resources in question are, as indicated earlier in these reasons, provided in part by government grants; the prospect of irate ratepayers left to shoulder the apprehended increased burden is conjectural. Further, it is implicit in the submissions for the interveners that highway authorities carry insurance in respect of their liability for misfeasance and other acts or omissions falling outside the ‘highway rule’. The Attorney-General for Victoria submitted that it should not be assumed that road authorities would be able through insurance to ‘transfer ... the financial burden of increased exposure to claims for compensation if their immunity for non-feasance is removed’. Nor should it be assumed that they will be unable to do so.”

At 560 [106], their Honours rejected an argument that the courts should respect the “political choice” involved in a statutory or government entity deciding matters of resource allocation. They said: “… citizens, corporations, governments and public authorities generally are obliged to order their affairs so as to meet the requirements of the rule of law in Australian civil society. Thus, it is no answer to a claim in tort against the Commonwealth under s 75(iii) of the Constitution that its wrongful acts or omissions were the product of a ‘policy decision’ taken by the Executive Government; still less that the action is ‘non-justiciable’ because a verdict against the Commonwealth will be adverse to that ‘policy decision’. Local authorities are in no preferred position. Yet it is submitted that those bodies which answer the description ‘highway authority’, distilled from the case law, merit and require a special consideration which only statute may displace. That submission should be rejected.”

In the context of considering when liability could be imposed on a roads authority for (amongst other things) failing to take preventative measures concerning a road, their Honours said, at 579 [155]: “The cost and practicality of an alternative and safer design, if one be available, may be weighed against the funds available to the construction authority. This may involve striking a balance between competing designs or methods of construction.”

At 580-1 [162] their Honours said: “The formulation of the duty of care includes consideration of competing or conflicting responsibilities of the authority. In the circumstances of a given case, it may be shown that it was reasonable for an authority to deal in a particular priority with repairs in various locations. The resources available to a road authority, including the availability of matériel and skilled labour, may dictate the pace at which repairs may be made and affect the order of priority in which they are to be made. It may be reasonable in the circumstances not to perform repairs at a certain site until a certain date, or to perform them after more pressing dangers are first addressed.”

The cases referred to at 560 [104] in Brodie are cited as authority for evidence about funding constraints and competing priorities being admissible. Of those cases, Hill v Commissioner for Main Roads (1989) 68 LGRA 173; 9 MVR 45; [1989] Aust Torts Reports ¶80-260 (68,781) was a decision in this Court that upheld an appeal against a verdict for the defendant concerning a motor accident. The roads authority had patched the road surface in question superficially, when a superior method of fixing the problem was available, and indeed was adopted two months after the accident. Samuels JA (with whom Kirby P and Priestley JA agreed) held that that amounted to an actionable misfeasance.

He went on (at LGRA 181; MVR 53; Aust Torts Reports 68,788-9): “… to consider whether there is sufficient substance in Mr Gross QC’s submissions concerning the availability of funds and the extent to which the respondent must be accorded some independent initiative in its selection of priorities to justify a new trial or some further investigation of the point. However, I am not satisfied that enough appears to compel such a course. I am well aware that the legal onus remains firmly on the shoulders of the appellant to establish all the ingredients of his cause of action. However, there was at least an evidentiary onus on the defendant to bring into contention the assertion that there were exculpatory economic circumstances which it might adopt as a shield. I do not think that there were …”

He described the serious inadequacy in the road, and continued: “… I see no reason to infer that the respondent had other priorities which were so compelling in their demands as to overcome what was plainly called for in this situation nor do I think that fairness requires any further investigation of this situation.”

That appears to be a case where the defendant had not provided an evidentiary basis for showing that on financial grounds it was reasonable for it to refrain from doing the work. It provides no assistance for dealing with the case (like the present) when there is an argument, supported by a mass of evidence, concerning financial limitations on the roads authority.

Gloucester Shire Council v McLenaghan [2000] NSWCA 208; (2000) 109 LGERA 419; 31 MVR 340 concerned a road accident held to have been caused when insufficient warning was given of the transition of the road surface from being sealed to being gravel. As Davies AJA (with whom Priestley JA agreed) said at [78] “The danger was reasonably foreseeable and readily remedied at little cost” , by installing appropriate signage.

The passage to which Brodie at 560 [104] refers is in the judgment of Priestley JA at LGERA 423 ([12]-[15]): “[12] The case itself provides a good example of a basic problem lying behind the rule in Buckle [ie, the non-feasance rule]. The extensive use of motor transport generates a need for good roads and highways and an expectation that they will be provided. This expectation translates itself into a tendency on the part of courts to think that liability for damage caused by defective roads and highways should be borne by highway authorities. Yet in many cases defects in roads and highways are the result of insufficient funding being available to the highway authorities. [13] The evidence in the present case shows the following: the population of the Gloucester Shire was approximately 4,900, so that the number of ratepayers would be very considerably lower; the council’s area was about 2,900 km 2 ; much of the road building and improvement in the Shire was paid for by special grants from the Commonwealth and the State; the council was earnest and persevering in its efforts to complete a fully sealed road between Gloucester and Walcha, but it was simply impossible to do so from its own funds and those otherwise made available to it, any more quickly than by the rather stately rate of progress shown in the evidence. [14] These features were emphasised in the council’s case, on the merits, and also were no doubt symptomatic of the policy background to the non-feasance rule. On the other side of the merits question was the fact that the council actively promoted the use of the road for tourist and commercial purposes, with a view to improving the economic life of the district. This was why the Gloucester/Walcha road was renamed Thunderbolt’s Way. The council was thus in the dilemma of wanting traffic on the road to increase but not having sufficient funds to bring it quickly into the state which was planned for it. [15] Cases more or less like the present one are continually occurring and cause acute problems both for damaged users of the roads and the highway authorities.” That passage recognises the problem that limited funds could provide for a roads authority, particularly a small country council, but does not tell how the law of negligence goes about taking that problem into account.

Woodward v Orara Shire Council (1948) 49 SR (NSW) 63 is a judgment of the Full Court of the NSW Supreme Court, delivered by Jordan CJ. The passage referred to at 560 [104] in Brodie (and, from the “cf” , apparently not endorsed) included a consideration of Miller v McKeon (1905) 3 CLR 50. At 66, Jordan CJ said that Griffith CJ’s phrase “and the moneys available to the Government for the purpose” “… has been assumed to make evidence of revenue necessarily admissible in all cases in which a local government body is charged with negligence in road making or road repair. I do not think that it was used with any such intention. It is difficult to understand how a jury could reasonably be asked, or even allowed, to assume to decide how much of its revenue a government ought to devote to road making or road improvements and in what localities the expenditure should be made. These are political questions which lie entirely outside the scope of any inquiry proper to be made by a jury, or for that matter by a court.”

The proposition that the majority judges in Brodie were laying down, and contrasted with what Jordan CJ had said, is that evidence respecting funding constraints and competing priorities will be admissible. However, they do not state how such evidence, once admitted, should be dealt with.

This consideration in Brodie of the role of financial considerations in deciding whether a roads authority has breached its duty of care is not complete. That happened because of procedural matters affecting how Brodie came before the High Court.

In Brodie , the plaintiff had been injured and his truck damaged when a bridge over which he was driving collapsed. The trial judge awarded the plaintiff damages, holding that the council that had control of the bridge owed the plaintiff a duty of care because its inadequate maintenance of the bridge was misfeasance rather than nonfeasance. The Court of Appeal reversed the trial judge, holding that such maintenance of the bridge as the council had carried out was not relevant to its structural integrity, and that the cause of the bridge’s collapse was the failure of the council to deal at all with those aspects of the bridge’s structure that caused the collapse. In consequence, this Court held that the council had an immunity from liability by reason of the nonfeasance rule.

As mentioned earlier, in the High Court in Brodie the focus of attention was on whether highway authorities should continue to have an immunity from liability for non-feasance. Having concluded that the highway authority in the instant case was under a duty of care, at 585 [181], Gaudron, McHugh and Gummow JJ expressed the view that the decision in favour of the plaintiffs was “supportable by application of the ordinary principles of negligence to the facts as found” . However, their Honours remitted the matter to the Court of Appeal so that the council could have the opportunity of challenging the facts as found. Kirby J, at 607 [249], agreed with that approach.

There was no detailed exposition of the reasoning by which their Honours concluded that the decision was “supportable by application of the ordinary principles of negligence to the facts as found” . The judgments delivered in Brodie in the Court of Appeal ( Singleton Shire Council v Brodie [1999] NSWCA 37) give no indication that financial limitations or competing budget priorities had been any part of the Council’s case as to why it had not been negligent. If that is so, the conclusion that the result arrived at by the trial judge was “supportable” is, with respect, readily understandable.

An illustration of the actual application to the facts of a particular case of general principles concerning matters to be “taken into account” in deciding if there had been a breach of duty of care can be very helpful in understanding those general principles. Several such illustrations can together provide an ostensive definition of the content of those general principles. However the circumstances in which Brodie came before the High Court had the effect that it does not give practical guidance on how to apply the general principles stated in it for deciding whether there has been a breach of the duty of care of a road authority. Thus, I turn to more general considerations. Decisions Courts are Incapable of Making? – Case Law

The cases show two separate strands of reasoning concerning the relevance of the financial resources of a statutory authority to whether it has acted negligently. One of them relates to decisions that courts are incapable of making. The other relates, in those areas where there is no inherent impossibility of courts making decisions, to the standard by reference to which reasonable conduct of a statutory authority will be assessed.

The strand of principle about some decisions regarding resource allocation being inherently unsuited for courts to determine has close connection to a distinction between “policy” and “operational” decisions of a statutory authority. The policy/operational distinction can be traced to East Suffolk Rivers Catchment Board v Kent [1941] AC 74, where Lord Romer at 102-3 quoted from du Parcq LJ in the Court of Appeal ( Kent v East Suffolk Rivers Catchment Board [1940] 1 KB 319 at 338): “… when Parliament has left it to a public authority to decide which of its powers it shall exercise, and when and to what extent it shall exercise them, there would be some inconvenience in submitting to the subsequent decision of a jury, or judge of fact, the question whether the authority had acted reasonably, a question involving the consideration of matters of policy and sometimes the striking of a just balance between the rival claims of efficiency and thrift.” This is in accordance with the remarks of Jordan CJ in Woodward v Orara SC , quoted earlier at para [224] above.

The distinction between policy and operational decisions was adopted by Lord Wilberforce (with whom Lords Diplock, Simon of Glaisdale, and Russell of Killowen agreed) in Anns v Merton London Borough Council [1978] AC 728.

Anns was a case involving an allegation of negligence on the part of a council in permitting a building to be constructed with defective foundations, and failing to carry out inspections that it had a statutory power to carry out “sufficiently carefully or at all” (at 750A-B). The English Court of Appeal had rejected a claim that the action was statute barred, and on an appeal to the House of Lords leave was granted to argue a question of law about “whether the council was under any duty of care to the plaintiffs at all” (at 751A). Thus, the speeches in the House proceeded on the hypothetical basis that the allegations in the statement of claim were true, and without any finding about whether any inspection had been made.

At 754C-E, Lord Wilberforce said: “Most, indeed probably all, statutes relating to public authorities or public bodies, contain in them a large area of policy. The courts call this ‘discretion’ meaning that the decision is one for the authority or body to make, and not for the courts. Many statutes also prescribe or at least presuppose the practical execution of policy decisions: a convenient description of this is to say that in addition to the area of policy or discretion, there is an operational area. Although this distinction between the policy area and the operational area is convenient, and illuminating, it is probably a distinction of degree; many ‘operational’ powers or duties have in them some element of ‘discretion.’ It can safely be said that the more ‘operational’ a power or duty may be, the easier it is to superimpose upon it a common law duty of care.”

It is, with the greatest respect, difficult to know what to make of this. The opening of the paragraph suggests that, because a decision in the policy area is one that is “not for the courts” , there must be a bright line between the policy area and the operational area. However, the end of the paragraph suggests that the distinction between the policy and the operational areas is one of degree, not of a bright line.

Lord Wilberforce went on at 754F-G to identify as part of the area of policy under the particular statute he was examining: “… the scale of resources which [the council] can make available in order to carry out its functions … – how many inspectors, with what expert qualifications, it should recruit, how often inspections are to be made, what tests are to be carried out …”

At 754G-H, he says (consistently with the distinction being a bright line one) that these decisions: “… must be for its [the council’s] decision. It is no accident that the Act is drafted in terms of functions and powers rather than in terms of positive duty … public authorities have to strike a balance between the claims of efficiency and thrift …; whether they get the balance right can only be decided through the ballot box, not in the courts.”

In Sutherland SC v Heyman at 469, Mason J adopted the policy/operational distinction, saying: “The distinction between policy and operational factors is not easy to formulate, but the dividing line between them will be observed if we recognize that a public authority is under no duty of care in relation to decisions which involve or are dictated by financial, economic, social or political factors or constraints. Thus budgetary allocations and the constraints which they entail in terms of allocation of resources cannot be made the subject of a duty of care. But it may be otherwise when the courts are called upon to apply a standard of care to action or inaction that is merely the product of administrative direction, expert or professional opinion, technical standards or general standards of reasonableness.”

In Sutherland SC v Heyman at 442, Gibbs CJ (with whom Wilson J agreed) said that the distinction between the policy and the operational area “is a logical and convenient one” . Deane J at 500 (who dissented as to the result) also recognised the distinction, at least where the “policy-making powers and functions” involved are “of a quasi-legislative character” .

In Stovin v Wise [1996] AC 923, the House of Lords considered a situation where a motorcyclist had been injured in a collision at an intersection at which the motorist’s view was obstructed by a bank on private land adjoining the highway. The council sought to persuade the landowner to sell the land to it so that the bank could be removed, but the landowner did not reply to the council’s letter, and the council did nothing further. The council had statutory powers to require the landowner to remove the bank. A claim of negligence against the council was rejected.

The opinion of the majority in the House of Lords was delivered by Lord Hoffmann (with whom Lords Goff of Chieveley and Jauncey of Tullichettle agreed, Lords Slynn of Hadley and Nicholls of Birkenhead dissenting) at 951 rejected the policy/operational distinction as being “an inadequate tool with which to discover whether it is appropriate to impose a duty of care or not” . He quoted with approval the remark by Lord Keith of Kinkel in Rowling v Takaro Properties Ltd [1988] AC 473 (UKPC) at 501B-D that the distinction: “… does not provide a touchstone of liability, but rather is expressive of the need to exclude altogether those cases in which the decision under attack is of such a kind that a question whether it has been made negligently is unsuitable for judicial resolution, of which notable examples are discretionary decisions on the allocation of scarce resources or the distribution of risks …. If this is right, classification of the relevant decision as a policy or planning decision in this sense may exclude liability; but a conclusion that it does not fall within that category does not, in their Lordships’ opinion, mean that a duty of care will necessarily exist.”

Lord Hoffmann said that the distinction was “often elusive” and continued, at 951F-G: “This is particularly true of powers to provide public benefits which involve the expenditure of money. Practically every decision about the provision of such benefits, no matter how trivial it may seem, affects the budget of the public authority in either timing or amount.”

He expressed the view, at 952D-F, that the non-feasance rule for highway authorities was, in terms of public finance, perfectly reasonable: “It is one thing to provide a service at the public expense. It is another to require the public to pay compensation when a failure to provide the service has resulted in loss. Apart from cases of reliance, which I shall consider later, the same loss would have been suffered if the service had not been provided in the first place. To require payment of compensation increases the burden on public funds. Before imposing such an additional burden, the courts should be satisfied that this is what Parliament intended.”

His Lordship held that an action for negligence consisting of the failure of a statutory authority to exercise a discretionary power would lie only when the statutory authority had been irrational in failing to exercise its statutory powers. (I give further consideration to that test at [339] ff below.) Presently relevant is that in the last paragraph of his speech at 958, Lord Hoffmann said, in a passage often considered since: “In my view the creation of a duty of care upon a highway authority, even on grounds of irrationality in failing to exercise a power, would inevitably expose the authority’s budgetary decisions to judicial inquiry. This would distort the priorities of local authorities, which would be bound to try to play safe by increasing their spending on road improvements rather than risk enormous liabilities for personal injury accidents. They will spend less on education or social services. I think that it is important, before extending the duty of care owed by public authorities, to consider the cost to the community of the defensive measures which they are likely to take in order to avoid liability.”

In Pyrenees SC v Day , Toohey J at 358 [68] regarded the policy/operational distinction as “not particularly appropriate or helpful in determining the present appeals” . That was because, as he explained at 362 [80], the decision involved in that particular case was clearly an operational decision. He did not reject the distinction as a matter of principle. Gummow J at 393 [182] held that the distinction “is not useful in this area” . Kirby J, at 426 [253] subpara 4, stated that: “Although the distinction is far from perfect, it has some validity” .

Amongst various factors against recognising a duty of care that Kirby J identified in Pyrenees at [253] were the comparatively small resources of the Shire Council, and the ability of the owners and occupiers to protect themselves by obtaining fire insurance. He also recognised that “In many cases, a court is not competent to review the decisions of public authorities as to their use of their resources.” On the facts of Pyrenees , however, providing more effective notice to the occupier would, he said, clearly fall on the operational side of the policy/operational divide. However, he accepted that “this conclusion leaves standing Lord Hoffmann’s comment that courts should be wary of undue intervention in decisions having significant implications for the budgets of public authorities” , and went on to quote from the speech of Lord Hoffmann in Stovin v Wise at 958.

In Romeo v Conservation Commission , not all judges gave explicit consideration to financial matters in deciding whether the statutory authority there involved had breached its duty of care.

Kirby J at 480-1 [129] said that courts must bear in mind the limited nature of resources available for public bodies, that allocation of resources is carried out by bodies accorded that function by law, and that expenditure of resources in one area necessarily diverts resources from other areas. He said these matters were factors “to be taken into account in making judgments which affect the operational priorities of a public authority” . He accepted that “so-called operational decisions” are subject to court assessment, but that even in relation to them: “… it is necessary to evaluate more than simply the cost of preventing the particular accident. Inherent in the suggestion of the obligation of prevention is the cost that would be incurred in the measures necessary to prevent all equivalent accidents of a like kind and risk.” (citation omitted)

At 484-5 [138]-[140], Kirby J rejected an argument that there was no duty of care because a decision on whether to provide any precautions against the risk of falling down the cliff “involved detailed evaluation of financial, economic, social or political factors”. Rather, at least concerning operational rather than policy matters, the: “… budgetary, political and other constraints within which such authorities must operate are factors to be taken into account in determining the scope of the duty of care and whether, in a particular case, it has been breached.” (citation omitted)

At 491-2 [166], Hayne J declined to decide on the validity of the policy/operational distinction, and on whether, if the distinction could be drawn, it offers a “touchstone of liability” . However, he observed that there “seems to be much force” in Lord Hoffmann’s remarks in Stovin v Wise about the difficulties of drawing such a distinction.

In Crimmins at 19 [27], Gaudron J said that: “… the nature or purpose of the powers and functions conferred [on a statutory body], or of some of them, may be such as to give rise to an inference that it was intended that the common law should be excluded either in whole or part. That is why distinctions are sometimes drawn between discretionary and non-discretionary powers, between policy and operational decisions and between powers and duties.” (footnotes omitted)

At 37 [87], McHugh J (with whom Gleeson CJ agreed) said of the policy/operational distinction: “It may be that functions and powers which can be described as part of the ‘core area’ of policy-making, or which are quasi-legislative or regulatory in nature, are not subject to a common law duty of care. Outside this narrowly defined policy exception, however, as Professor Todd has argued, it seems preferable to accommodate the distinction at the breach stage rather than the duty stage.” (footnotes omitted)

At 50 [131], McHugh J said: “The ‘policy/operational’ distinction has certain difficulties that attend it, but the nature of the other powers and functions exercised by the Authority with respect to safety clearly fall closer to the ‘operational’ end of the spectrum. Although they involve considerations of convenience, discretion and budgetary allocation, they are matters appropriately considered as part of the breach question.”

In his dissenting judgment at [292], Hayne J said of the policy/operational distinction that it “may not always be useful” .

In Graham Barclay Oysters Gleeson CJ referred at 554 [7] to Brodie (in which he had dissented) and said: “In that case three members of the majority expressly accepted that it may be proper and necessary for a court to decide whether the priorities of a local council in dealing with road repairs in various locations were reasonable: Brodie (2001) 206 CLR 512 at 580-581 [162]. The decision in the case required that view. Even so, the scope for judicial examination of the reasonableness of governmental spending priorities was not held to be, and cannot be, at large. Raising and spending money for road repairs involves setting priorities, not only between parts of the road system, but also between all the claims upon an authority’s resources, and between the interests of taxpayers and those of road users. My view remains that setting priorities by government for the raising of revenue and the allocation of resources is essentially a political matter, and that, if the reasonableness of such priorities is a justiciable issue, that can be so only within limits.”

At [12] he referred to the policy/operational distinction, saying: “That distinction was never rigorous, and its validity and utility have been questioned: eg, Pyrenees Shire Council v Day (1998) 192 CLR 330 at 393-394 [180]-[184]; Stovin v Wise [1996] AC 923 at 951-953. Even so, the idea behind it remains relevant in some cases, such as the present.”

After having set out Mason J’s remarks in Sutherland SC v Heyman at 469, Gleeson CJ went on to say at 557 [13]-[15]: “One of the reasons why matters of the first kind are inappropriate as subjects of curial judgment about reasonableness is that they involve competing public interests in circumstances where, as Lord Diplock put it, ‘there is no criterion by which a court can assess where the balance lies between the weight to be given to one interest and that to be given to another’: Home Office v Dorset Yacht Co [1970] AC 1004 at 1067. There are forms of governmental activity, which courts in the past endeavoured to describe by the term ‘operational’, where there is no reason for hesitating to assimilate the position of governments to that of citizens in imposing duties and standards of care. Such activity might involve budgetary considerations, but that does not prevent such assimilation. Individuals and corporations also have to watch their budgets, and decisions about what is reasonable may have to take account of that. As the other extreme, the reasonableness of legislative or quasi-legislative activity is generally non-justiciable. Here we are concerned with the problem of deciding, in a case where the government had certain powers, whether it is accountable, through the law of negligence, for not exercising its powers, or for not exercising them sufficiently. To apply that form of legal accountability requires the identification, not merely of a power, but also a duty; a duty of care owed to a citizen or a class of citizens. A conclusion that such a duty of care exists necessarily implies that the reasonableness or unreasonableness of the inaction of which complaint is made is a legitimate subject for curial decision. Such legitimacy involves questions of practicality and of appropriateness. There will be no duty of care to which a government is subject if, in a given case, there is no criterion by reference to which a court can determine the reasonableness of its conduct.”

All this provides no clear guidance as to how one reasons. It can be said, however, that: (a) there is significant, but not unanimous, judicial recognition that the distinction between policy and operational matters is of some use in deciding when a public authority has been negligent; (b) the weight of opinion seems to be against the policy/operational distinction being used as a determinant of when a duty of care is owed; (c) the relevance that the distinction has in the law of negligence is that matters of policy are ones that are unsuited for determination by courts; and (d) there has been repeated recognition that there are some decisions about whether a statutory authority has taken reasonable care that courts are unable to make: Woodward v Orara SC at 66, East Suffolk Rivers Catchment Board v Kent at 102-3, Anns at 754, Rowling v Takaro Properties at 501, Pyrenees SC v Day at [253], Graham Barclay Oysters at 554 [7], Home Office v Dorset Yacht Co at 1067, all quoted above. Standard of Reasonable Care for a Statutory Authority – Case Law

Pyrenees was a case where the omission that was held to be negligent was the failure of the council to follow up on and exercise its statutory powers concerning a requirement it had made of a building owner not to use a particular fireplace. That was a situation where the cost of taking the follow-up action was likely to have been comparatively small. At 392 [179], Gummow J noted that it was not a case where “at least stopping short of prosecution, the further pursuit by the Shire [of follow-up action to its initial notice] would have interfered with the budgetary priorities of the Shire, or distorted its priorities in the discharge of its statutory functions.” (citation omitted) On that ground he distinguished the last paragraph of the speech of Lord Hoffmann in Stovin v Wise that I have quoted at para [245] above.

Gummow J did not express any disagreement with that statement of Lord Hoffmann. It is to be noted, however, that Lord Hoffmann did not go so far as to say that exposing the authority’s budgetary decisions to judicial inquiry was a reason why a duty of care could never be imposed on a statutory authority – rather, the distortion of an authority’s provision of public services by taking defensive measures was a matter that was to be considered. That seems consistent with the view expressed by Gummow J at 394 [183] that “questions of resource allocation and diversion, and budgetary imperatives should fall for consideration along with other factual matters to be ‘balanced out’ when determining what should have been done to discharge a duty of care.” (citations omitted)

Gummow J at 394-5 [184]-[185] recognised that giving effect to these matters might result in the standard of care owed to a plaintiff by a statutory authority being “less than that which would be owed by a private party” . He quoted with approval the statement of Cory J (delivering the judgment of the majority) in Just v British Columbia [1989] 2 SCR 1228 at 1243-4 concerning how the standard of care which is owed to a plaintiff by a government agency may be less than that which would be owed by a private party: “Nevertheless the standard of care imposed upon the Crown may not be the same as that owed by an individual. An individual is expected to maintain his or her sidewalk or driveway reasonably, while a government agency such as the respondent may be responsible for the maintenance of hundreds of miles of highway. The frequency and the nature of inspection required of the individual may well be different from that required of the Crown. In each case the frequency and method must be reasonable in light of all the surrounding circumstances. The governmental agency should be entitled to demonstrate that balanced against the nature and quantity of the risk involved, its system of inspection was reasonable in light of all the circumstances including budgetary limits, the personnel and equipment available to it and that it had met the standard duty of care imposed upon it.”

If, when a private individual or corporation is carrying out some activity, taking reasonable care to avoid harm to people to whom a duty of care is owed would require expenditure to be made, but the particular individual cannot afford that expenditure, the poverty of the individual and significant conflicting responsibilities sometimes provide no escape from a finding of breach of duty if the individual nonetheless persists with the activity. For example, taking reasonable care for other motorists could require a car owner to spend money on the repair of his or her vehicle. If the money is not spent, and the individual continues to drive, and the lack of repair causes an accident, the individual cannot be heard to say even that his choice was between maintaining the vehicle and providing food or necessary medical attention for his family. The response of the law of negligence to that situation is “if you can’t afford to make your car safe to drive, don’t drive it.” That is because the standard is judged by reference to what a reasonable car owner would do, and the hypothetical reasonable car owner is not someone who lacks the resources to make the sort of expenditure that is part and parcel of owning a car.

The history of litigation about industrial safety conditions shows how failure to provide even costly equipment, systems, training or supervision for employees could be a failure to take reasonable care, because the reasonable employer in a particular line of business is not someone who lacks the resources to make the sort of expenditure that is involved in employing people to work in that line of business in conditions of reasonable safety.

By contrast, it is not open to a statutory authority that has responsibility for administering some field of endeavour conferred on it by statute, to withdraw from that field if it lacks resources to carry out some particular activity that is within its powers. It would ignore reality for a court to proceed on the basis that a statutory authority should be taken to have sufficient resources to carry out all its statutory duties, powers and discretions. An effect of this is that the standard by which one decides whether a statutory authority has acted negligently is not the same as that applicable to a private individual or corporation, but rather is the standard of what a reasonable authority, with its powers and resources, would have done in all the circumstances of the case: Crimmins at 21 [35] (per Gaudron J), 38 [90] (per McHugh J, with whom Gleeson CJ agreed). Decision Courts Unable to Make – General Considerations

There are two different ways in which a court might be unable to make a decision about what is involved in taking reasonable care in a particular situation. The first is where it can perceive that there is some public policy that prevents it from making any such decision. Such a situation could arise concerning an allegation of negligence in the exercise, or failure to exercise, a particular statutory discretion. There have been suggestions that, at the least, such a policy might be perceived in statutes that confer a statutory discretion of a quasi-legislative kind. I will not elaborate on that particular mode in which a decision about breach might be unsuited for being made by a court, because the present case is not one concerning which any such public policy can be perceived. A sufficient reason why that is so is that the present case is not one that involves a failure to exercise a statutory discretion at all – the complaint against the RTA is, simply, that it failed to act when it could have.

The second way in which there could be a decision that it is unsuited for courts to make is one where the unsuitability arises from the nature of the judgment that the court is called upon to make.

One of the driving concepts in the law of negligence is that of “reasonable care” . It comes from a recognition that one person’s actions or inactions can affect other people, and that sometimes the relationship between two people is such that the law imposes a duty on one of those people to take some steps to avoid harming the other. When such a duty is imposed, the extent of the steps that the law requires the first to take in order to avoid harm to the second is measured by the concept of reasonable care. That concept is an objective one in that, as Mason J explained in Shirt , it is dependent upon the steps that a reasonable person would take to avoid the relevant type of harm. That “reasonable person” is one who exists in a particular society, at a particular point in time.

In the first few decades after the modern law of negligence was formulated in Donoghue v Stevenson [1932] AC 562, the function of deciding whether reasonable care had been taken in a particular situation where the law held a duty of care existed was one entrusted to juries. What counted as reasonable care in a particular situation was the sort of thing that an ordinary member of the community, called to serve on a jury, could be expected to form a view about on the basis of his or her ordinary knowledge and experience. In this way, it had a basis in community standards about the extent to which one member of the community should avoid causing harm to another. The change from having questions of whether there had been a failure to exercise reasonable care decided by juries, to having such questions decided by judges, has not altered the nature of the question that is being asked.

There are numerous activities that governmental or statutory authorities carry out that are of a type concerning which there is no difficulty in deciding whether the governmental or statutory authority involved has acted in accordance with the way a reasonable member of the community carrying out that activity should act. The standards for a governmental or statutory entity taking reasonable care in driving a motor car, maintaining a motor car, employing labour, acting as a landlord of premises, or being the occupier of an office or shop to which members of the public come as licensees or invitees rather than as of right are exactly the same as those imposed on any other person carrying out those activities.

The task of deciding whether reasonable care has been taken can become more problematic concerning activities that are of a type that private individuals do not carry out. However, sometimes, an activity that only a statutory authority carries out can be sufficiently analogous to activities carried on by private individuals or corporations to enable the standards applicable to private individuals in that analogous activity to be drawn on and adopted to the statutory authority. Such a situation arose in Crimmins , where the statutory authority’s role as a co-ordinator of labour was analogous to that of an employer or co-ordinator of labour in private industry (I say this recognising that the debate in Crimmins in the High Court concerned existence of duty of care, not breach).

As Pyrenees shows, even in relation to activities that are peculiarly governmental and not analogous to anything a private individual or corporation does, like issuing a statutory notice requiring work to be done, it is sometimes possible for a judgment to be made about what the taking of reasonable care concerning the exercise of such a power requires.

The difficulty in giving content to the notion of reasonable care becomes greater when the allegation of negligence against a statutory authority is of a failure to act in some particular respect (rather than having embarked on some course of action and carried it out negligently), where that failure to act is one that has no analogue in the activities of private individuals, and where having acted would have involved significant expenditure. There are many situations where the making of a significant expenditure by a governmental or statutory authority involves choosing on which of several disparate types of activities the money will be spent on. Some such choices, such as whether money will be spent on roads rather than schools or hospitals, are typically made by budget legislation. Even within the one area of ministerial responsibility, there are some topics concerning which it will not be possible to say that the taking of reasonable care required the expenditure to be made on one type of activity rather than another. In the present case, the taking of reasonable care does not provide a workable criterion for choosing between whether the RTA should have spent money on, eg, road widening, erection of sound screens beside freeways, installing warning lights at railway level crossings, or screening freeway overpasses.

In an action for negligence, before a plaintiff can succeed one of the things the judge must decide is that the defendant has failed to take reasonable care. There will be cases where the judge does not come to such a decision because the judge is affirmatively satisfied that taking reasonable care required certain steps to be taken, and those steps were taken. However there will be other cases where the evidence is deficient, so that, even though the topic in question is one that, in a differently run case, the judge would have had no difficulty in deciding that the defendant did not take reasonable care, in the case being tried the plaintiff has simply failed to prove that the defendant did not take reasonable care. There will be other cases where the task of deciding that the defendant did not take reasonable care is one that the judge is incapable of carrying out because the subject matter is too far removed from the area in which the notion of “reasonable care” operates. In such a case, acknowledgment of the nature of the subject matter will result in the judge being unable to conclude that the defendant has failed to take reasonable care. All three cases are equally ones where the plaintiff fails, because the judge is not satisfied that the defendant failed to take reasonable care.

The task of operating public roads is no longer a peculiarly governmental one. One can take judicial notice that in the Sydney region there are several toll roads that are run by private operators. However, that fact does not make it any easier to decide what reasonable care on the part of the RTA would require concerning a safety matter like that involved in the present case. The evidence in the present case does not tell us whether the private toll road operators are legally free to raise their tolls to whatever levels the market will bear, whether they have any contractual or other obligations to keep on providing the road service, whether there are contractual or other obligations about the standards of safety or service they must provide, or whether the RTA has any continuing legal rights or powers concerning the privately operated roads. Those matters would be very relevant to what the taking of reasonable care by a private road operator required. And even if one knew these matters, there might still be some difficulty in deciding what reasonable care on the part of such a private operator required, and further difficulty in analogising that standard to the position of the RTA.

There is a complex collection of laws relating to the manner in which the NSW government raises and spends money. Section 39 Constitution Act 1902 establishes a Consolidated Fund, of which (unless other legislation provides otherwise) all public monies collected received or held by any person for or on behalf of the State form part. Section 45 Constitution Act enables the Consolidated Fund “to be appropriated to such specific purposes as may be prescribed by any Act in that behalf . ” As Twomey, The Constitution of New South Wales (2004), The Federation Press) says at 541: “There are two basic types of appropriation: annual appropriations and special appropriations ( Brown v West (1990) 169 CLR 195, per the court at 206). Annual appropriations form what is known as the ‘Budget’. Each year annual Appropriation Bills are enacted to provide for the expenditure of the government in the next financial year… ‘Special appropriations’ (also known as ‘ standing appropriations’) are permanent provisions (until altered by legislation) which provide for appropriation from time to time in specified circumstances. They act as a charge against the Consolidated Fund. The amount to be appropriated is calculable by reference to the requirements of the Act. They are used where it would be inappropriate to have the appropriation debated before the Parliament each year or where it is convenient that an appropriation continue to apply in relation to the subject in the long term.” See also Pape v Federal Commissioner of Taxation [2009] HCA 23; (2009) 83 ALJR 765; 227 ALR 1 at [64], [293].

An appropriation both authorises the expenditure of public money, and restricts the purpose for which expenditure is authorised: Victoria v The Commonwealth and Hayden (“ AAP Case ”) (1975) 134 CLR 338 at 392 (per Mason J); Commonwealth v Colonial Ammunition Co Ltd (1924) 34 CLR 198 at 224 (per Isaacs and Rich JJ); Twomey, op cit p 542. However, a contract by the State can be valid even if there has been no appropriation of money for the purpose of the State performing that contract – rather, the contract is regarded as subject to an implied condition that the relevant amount will be appropriated: Twomey op cit p 543 and cases there cited.

It may well be that, if a court holds that the Crown in right of the State, or some State governmental entity, was negligent in failing to act in some way that would have involved the expenditure of money, it is deciding that a reasonable government or statutory entity would have spent money in a way that has not been the subject of appropriation. Whether that is so is something a court often does not pause to enquire – as when Nagle v Rottnest Island Authority held, without any consideration of the state of the finances of the defendant authority, that the defendant authority should have erected a (fairly cheap) warning sign.

Indeed, whenever a court orders judgment against a governmental authority whose finances come from parliamentary appropriation it usually does not pause to enquire what is the situation about budget appropriation before pronouncing the judgment. Such judgments are usually made on the basis that the government will appropriate money to satisfy them: New South Wales v Bardolph (1934) 52 CLR 455 at 510 (per Dixon J); Alcock v Fergie (1867) 4 WW & a’B (L) 285 per Stawell CJ at 316. (A qualification is that in Bardolph at 516, Dixon J noted that this “does not mean that, if Parliament has by an expression of its will in a form which the Court is bound to notice, refused to provide funds for the purposes of the contract, it remains actionable under the Judiciary Act . ” )

In deciding whether a court is able to decide that a governmental authority had breached a duty of care it owes by not taking some course of action, what matters is not whether there has or has not been money appropriated for that activity, but the type of decision that would have been involved in deciding to spend money on that activity. Concerning comparatively cheap activities like the erection of a warning sign, or service of a statutory notice requiring some source of danger to be fixed, a court can sometimes say that it is able to make the judgement about whether a reasonable statutory authority, in the position of the defendant, would have spent the money. That decision of reasonableness has implicit in it all the pre-requisites for the statutory authority having been able to spend the money.

Concerning other failures of the defendant authority to act in some particular way, the court may not be in a position to decide how a reasonable authority would have acted. It is often only by considering a particular action, which a statutory authority is alleged to have breached its duty by failing to engage in, that a court can decide whether the failure to do that particular action is one concerning which the court is capable of deciding whether there has been a failure to exercise reasonable care. Before a court finds breach of duty, the court must make a positive decision that the defendant authority has failed to act as a reasonable authority in its position should have done. The further the disputed failure to act is from the area in which ordinary knowledge and experience provides a basis for making decisions, the harder it will be for the court to reach such a positive decision.

It might sometimes happen, even in relation to a decision that is clearly in the operational area, that an authority could not reasonably have been expected to get itself into the financial position to make the expenditure in question by the time the plaintiff suffered his or her injury. This could happen concerning dealing with a completely new type of risk that emerges, notwithstanding that the risk is comparatively cheap to deal with. There would be an onus on the authority of adducing evidence to show that that was the situation – a court would often presume that a reasonable authority would find the money to make a comparatively small expenditure.

The impossibility of a court making some decisions about whether a statutory authority has exercised reasonable care is sometimes relevant to the existence of a duty of care, and sometimes only to whether there has been a breach of a duty of care. Consider the situation where it can be predicted that actions that a statutory authority might take, or fail to take, in the course of the relationship that it has with some particular class of persons have the capacity to cause harm to that class of persons. That situation might also be one in which it can be predicted that all the actions that the statutory authority is likely to take, or fail to take, in the course of that relationship will be of a type concerning which the court is not in a position to judge whether the action or failure to act is conduct that a statutory authority taking reasonable care to avoid harm to people who could be affected would engage in. If so, that is a good reason for denying that the statutory authority owes a duty of care concerning whether it does, or does not, engage in that type of action.

(The nature of the relationship between the statutory authority and the class of persons who could be affected by the action, or failure to act, of the statutory authority might also provide a reason why no duty of care was owed, but that is a separate matter.)

However, if some of the actions, or failures to act, that it could be predicted the statutory authority might take are ones concerning which a court can make a decision about whether those actions, or failures to act, are ones that a statutory authority taking reasonable care to avoid harm to those who might be affected by its actions or lack of action would have engaged in, but others are not, that could be a reason for holding a duty of care does exist (if there were not some other reason for denying its existence), and leaving it to the breach stage of the analysis of negligence to apply the test of whether the concept of taking reasonable care is too hard to apply to the particular decision in question to enable a conclusion of breach of duty to be arrived at.

Brodie at 560 [106] rejected a submission that the fact that a decision of a public authority involved a “policy decision ” was a sufficient reason for the law of negligence not to apply to that decision. It is readily understandable how this could be so. A court would often have no difficulty in deciding that a statutory authority’s having made a policy decision that it would spend money on something other than maintaining its motor cars properly, or providing safe conditions of work for its employees, was of no relevance to whether the statutory authority had exercised reasonable care in those respects. However, there are occasions when the obligation of governmental authorities to comply with “the requirements of the rule of law in Australian civil society” can run up against inherent limitations in applying the law of negligence to particular activity of governmental organisations. There are some governmental actions, or failures to act, concerning which it is particularly difficult to apply the notion of “taking reasonable care” for those affected by the action or failure to act. The greater the difficulty there is in applying that notion, the harder it will be for a court to reach an affirmative conclusion that the governmental authority has failed to take reasonable care. The Decision Below Concerning Significance of Funding Constraints

I will consider first the situation concerning breach of duty under the general law, and later consider how the provision of section 42 Civil Liability Act affects the situation.

It will be recalled that from the time the problem of objects being dropped from overpasses first came under active consideration by the RTA it acted on the basis that the money to be spent on achieving screening would come from the Commonwealth Government. However, when the problem achieved a higher degree of urgency in August 1998 Mr Dunlop's instruction was to “do what is needed now and then find money from where ever . ”

Mr Onggo’s evidence was that, even if there was no allocated funding for a project, “if something needed to be done, it could be done” . Mr Fishburn’s evidence was that in his time with the RTA (up to June 1998) in relation to a federally funded road, “… if there arose a safety issue with respect to that piece of road then if the federal government would not come up with the funds the RTA would have to come up with the funds.”

The judge used this evidence of the basis for treating the evidence in the case about budgetary procedures as of little importance. He said, at [51]: “Although the allocation of federal funds was made and budgets set, it was shown from various pieces of evidence that if a need arose for certain work to be done the budgetary constraints were no impediment.”

One of the judge’s findings of fact at [60] (22) was: “If the need arose for certain works to be done then budgetary constraints on the defendant were no impediment as there could have been a re-allocation of available funds to the more necessary works.”

As part of his findings concerning breach of duty at [108] the judge said that in August 1998: “Mr Dunlop did not see funding as a problem as it would be obtained from ‘wherever’ and, indeed, the Minister forthwith announced state funding to screen the overbridges on the National F5 freeway.”

Part of that last finding is factually mistaken. Mr Scully, the relevant State Minister, made a media release on 3 September 1998 that included a statement that the State Government would spend $3.2 million to accelerate its program of screening bridges over high-speed motorways in the greater Sydney metropolitan area. It quoted Mr Scully as saying that he had decided to accelerate the work program to screen all the remaining over bridges at nearly twice the rate originally planned, and that an additional $1 million in state funds would be provided to allow for the acceleration of the existing $2.2 million program. In that announcement, Mr Scully was talking about both screening of overbridges on state funded roads, and screening of overbridges on federally funded roads. The media release did not say that any part of the extra $1 million of state funding would be spent on screening bridges on federally funded roads.

Mr Scully wrote to the relevant Federal Minister on 5 November 1998. He referred to a federal Budget advice that New South Wales would receive $24.62 million in 1998/99 for safety and urgent minor works on the National Highways. The letter enclosed a schedule that revised forward estimates for safety and urgent minor works on National Highways in New South Wales. It did so by revising the manner in which the $24.62 million would be allocated between different projects. That schedule showed that, instead of the $370,000 that had been estimated in the forward strategy for 1998/99 to be spent on overhead bridges protection screening on the Hume Highway, an amount of $630,000 was now proposed to be spent on the item in 1998/99, and a further $600,000 to be spent in 1999/2000. Similarly, the amount proposed to be spent on overhead bridges protection screening on the F3 was estimated to be $600,000 in 1998/99, instead of the $300,000 that had been proposed in the forward estimate for that year. Mr Scully’s letter expressly stated that the action of increasing those amounts “follows the tragic death of Mr Mark Evans on the Hume Highway last August.” The letter made a thinly veiled complaint by pointing to the way in which federal funding for safety and urgent minor works on National Highways in New South Wales had been cut from $27.36 million in 1996/97 to $24.62 million in 1998/99, and a further reduction had been indicated for 1999/00.

The Federal Minister replied on 18 December 1998, approving that program of works. In response to the thinly veiled complaint he said: “I too would like to have more funds available for roads in NSW as well as in the other States and territories. However, the reality is that the Government has since its election adopted a policy of fiscal restraint, the benefits of which the whole country is enjoying in the healthy state of the economy. NSW is in a favourable position in relation to road funding as it has a very high share of the total funds available, which it has maintained during the last few years. During the recent election campaign commitments were made to provide additional funding to NSW that will increase its share further.”

Of course, the way in which federal funding works, by reimbursing states for expenditure actually made, meant that in the first instance the State would be spending its own money (or using its own credit) in carrying out any work at all on federally funded roads. However that is a matter of timing of cash flow, not of which government ultimately pays. It is not established that, after the death of Mr Evans, the NSW government became willing to carry out bridge screening works on the F5 on the basis that it would be ultimately paying for those works from its own resources.

There are two reasons why the fact that an accelerated program for screening of overpasses was adopted following the incidents of 9 and 10 August 1998 is not indicative of a failure to exercise reasonable care in not having screened those bridges at a faster rate previously. The first is that section 5C Civil Liability Act provides: “In proceedings relating to liability for negligence: … (c) the subsequent taking of action that would (had the action being taken earlier) have avoided a risk of harm does not of itself give rise to or affect liability in respect of the risk and does not of itself constitute an admission of liability in connection with the risk.”

The second is that the incidents of 9 and 10 August demonstrated to Mr Dunlop that the risk at that location was greater than he had previously assessed it as being. Quite apart from section 5C, his response to a greater known risk after 9 and 10 August is no indication of what a reasonable response to a risk understood to be a lesser one would have been prior to 9 and 10 August.

If the judge was intending by his finding at [51] to convey that the officials of the RTA were free to ignore budgetary constraints to meet what they perceived to be a need for expenditure of money, that could not be accepted. When the court makes a finding that a public instrumentality has failed to take reasonable care, by failing to spend money in some particular way, it is not finding that the officers of that instrumentality ought to have ignored the budgetary context within which they worked. After all, the rules concerning the way in which public money is raised, appropriated, spent and then accounted for are themselves rules of law, and it would be a rare case indeed (if it could ever happen) when a court could properly make a finding that performance of one legal duty required a public official to ignore another legal duty that he or she was subject to. Rather, the finding is one that has implicit in it that the ordering of priorities for expenditure of money ought, if reasonable care was to be taken, have been different to what they actually were. In fairness to the judge, it should be said that it is not at all clear that he was intending to say that the public officials involved in the present case should have ignored the budgetary constraints on them and simply spent the money. Decision on Whether RTA Failed to Exercise Reasonable Care

I turn to consider whether in the present case I am satisfied that, by the decision-making processes adopted once incidents of objects being dropped from overpasses became known, and by prioritising its expenditure in the way it did, so that the Glenlee Bridge had not been fenced prior to 23 August 1998, the RTA has failed to exercise reasonable care.

A constant feature of the approach that the RTA adopted was that it would only spend such money on screening of overpasses on National Highways as the Federal government made available. Our attention has not been drawn to any legislation or intergovernmental agreement that would have prohibited the RTA from spending state-sourced money on such a project. As mentioned earlier, the evidence does not establish any limitations arising under the budgetary processes of the State government that would have prevented state-sourced money from being spent on such a project. Even so, it is a tolerably clear inference that a decision has been made, or course of conduct adopted, within the State government that funding of the National Highway is a matter for the Federal government. That is the sort of decision or course of conduct the departure from which would have implications for a vast array of state expenditure decisions. There is not the evidentiary basis in the present case to even attempt to make a decision on whether the RTA taking reasonable care to protect motorists from a risk of falling objects would have involved departure from such a decision or course of conduct. In any event, whether the RTA taking reasonable care to protect motorists from falling objects required departure from such a decision or course of conduct is the sort of decision that is outside the scope of matters that a jury could have decided. A judge is in no better position.

Some practical limitations on the RTA funding bridge screening from the federal allocation are that the Federal government not only decides what is the total amount of money it will make available for expenditure on National Highway roads in New South Wales in any year, but as well exercises minute control over the particular projects on which money that is made available is spent. A program of works is required to be submitted in advance of the Federal budget, and any money that is allocated in the Federal budget that is not actually spent during that budget year is not available to be spent in a subsequent year. That process means that a risk to road users, that is to be remedied through federal funds, must be identified by the RTA, and the cost of remedying it decided upon, by the time bids are put in for federal funds prior to a Federal budget. There is no “cushion” or “hollow log” amount that is allocated by the Federal government but not tied to a particular project, which can be resorted to if an emergency not allowed for in the budget bids arises in the course of a financial year.

Further, prior to putting in its bids for any financial year the RTA must decide on the relative importance of remedying any particular type of risk to road users by comparison with the importance of remedying other risks to road users, or the achieving of other benefits from road improvements. It is apparent from the documents the RTA prepares in the bidding process that the considerations it takes into account include not only reducing the risk of injury and death to road users, but also other matters. There is the achievement of economic benefits through matters such as reduced vehicle operating costs and faster travelling times. There is replacement of flood-prone structures with ones that are not flood-prone. There is improving pavement quality so that particular roads can withstand heavier vehicles. There is the need to comply with various regulations, and environmental initiatives such as providing noise protection. There are matters of amenity for motorists such as the provision of parking or rest areas, and the provision of facilities for people affected by roads other than in their capacity as motorists, such as pedestrian and bicycle facilities.

Even concerning matters relevant to safety of road users, there is quite a variety of different types of measures that need to be weighed up – road widening, road deviation, conversion of a single-carriageway road to a dual-carriageway road, provision of passing lanes, bridge widening and replacements, hazardous location treatments, shoulder sealing, railway crossing treatments, and various types of delineation and guidance enhancement. As well, an element in the RTA’s choice about in what locations or on what particular projects money would be spent was pure electoral politics, as is illustrated by the exchange of emails between Mr Manwaring and Mr Procter in July 1997 (paras [71]–[74] and its ultimate outcome in para [85] above). The evidence also shows how the RTA can take electoral politics into account along with safety considerations – the need to protect two railway crossings on the Newell Highway, that presented itself in mid 1997, was clearly a real safety issue, but as well there was a need to “appease the community” .

For a jury to have come to a decision that the RTA failed to take reasonable care to prevent harm to motorists, when it decided what priority to give to bridge screening by comparison with the other calls on the funds available from the federal government, is in the present case not a possible task. I hesitate to say that there would never be a case where a decision could be come to on such a topic, but if they exist such cases would be rare.

If one alters one’s focus from the broad picture to the particular, the RTA sought funding for the screening of the Glenlee Bridge through the budgetary process starting as early as 30 April 1996 (para [51] above). Funding for the Glenlee Bridge was included in the forward estimates for the 1997/98 year, that were submitted on 5 February 1997 (para [56] above). Even though there had been no application for funding for screening in the 1996/97 budget allocation, the Federal budget for 1996/97 was late in being delivered and in consequence NSW was not able to spend all of the amount that had been allocated to those safety and urgent minor works projects that had been approved for the 1996/97 year. When it became apparent that NSW would suffer a shortfall of this type, application was made in April 1997 to spend virtually the entire projected shortfall on other works during the 1996/97 year. The choice was made to seek approval to screen two bridges during the 1996/97 year, by advancing them in priority from the 1997/98 year. The Federal government approved that being done, on 9 May 1997.

The existence of carry-over work from the 1996/97 year cut into the work that could be done from the 1997/98 allocated amount of Federal funds. The result of a reconsideration within the RTA of how to spend money on safety and urgent minor works during 1997/98 (including the matters of electoral politics earlier referred to) was that the amount to be spent on bridge screening during that year shrank, by November 1997, to $370,000. That amount allowed for the screening of two bridges on the F5 during 1997/98.

By the end of the 1997/98 year, three bridges on the F5 had been screened, not including the Glenlee bridge, and two bridges on the F3 had been screened. Funding for the 1998/99 year had been obtained for a total of $670,000 to cover both the F3 and F5. Screening of the bridges was being done in accordance with the order of priority derived from the evaluation matrix that Mr Onggo had developed. The statutory prerequisite of performance of a REF for the Glenlee Bridge was complete. According to Mr Onggo’s evaluation matrix, the order of screening of bridges on the F3 and F5, as at 30 June 1998, had the bridges over the F3 and F5 that were still unscreened ranked in order so that six bridges on the F3 had higher priority than any on the F5, the Brooks Road bridge on the F5 was next, followed by another bridge on the F3, then the Narellan Road Mt Annan bridge on the F5, and then the Glenlee Bridge. The forward estimates for 1998/99 were to the effect that six bridges on the F5 would be screened during the year, and three on the F3. That suggests that Mr Onggo’s evaluation matrix was not the sole basis for decision about on which road bridges would be screened. Even so, spending available money that had been allocated to bridge screening would not have resulted in the Glenlee Bridge being screened before 23 August 1998.

In light of the nature of the risk, and of the other responsibilities of the RTA for expenditure of money available to it from the Federal government on the National Highway in New South Wales, I am unable to reach a conclusion that the RTA failed to take reasonable care in not having the Glenlee Bridge screened prior to 23 August 1998.

The judge’s reference at [60](18) of his reasons to the 1996/97 allocation for safety and urgent minor works having been underspent by $3.26m is explained by Mr Manwaring’s email of 4 July 1997 (para [71] above). In cross-examination, Mr Manwaring agreed that it would have been possible for the NSW government to request the Federal government to permit that sum of money to be spent in the following year on the projects for which it had been intended, but that that course was not followed. However, there was no particular of negligence alleging that the NSW government should have asked the Commonwealth government for more money at any time. In any event it is speculation whether, if the NSW government had requested to carry over the $3.26m to the 1997/98 year, the Federal government would have agreed to it. There is no basis for concluding that the Federal government probably would have agreed to it. Particularly is that so when one bears in mind that in November 1996 the Federal department had foreshadowed decreases in the amounts to be spent on safety and urgent minor works in each of the years to 1999/2000 (para [55] above), and that the amount it had actually approved for safety and urgent minor works in each of 1997/98 and 1998/99 were exactly in accordance with its predictions about the reduced amounts that would be spent (para [90] and [92]). The political doublespeak of referring to this as an “efficiency dividend” or “fiscal restraint” does not hide the reality that it was a funding cut, and outside the control of the RTA.

Even adopting this more limited focus on the facts of the case, in light of the nature of the risk to motorists, and of the other responsibilities of the RTA for expenditure of money available to it from the Federal Government on the National Highway in New South Wales, I am unable to reach a conclusion that the RTA failed to take reasonable care in not having the Glenlee Bridge screened prior to 23 August 1998. Issue 2(3) – Section 43A Civil Liability Act Available?

Mr Maconachie sought, on the appeal, to place reliance upon section 43A Civil Liability Act . It provides: “(1) This section applies to proceedings for civil liability to which this Part applies to the extent that the liability is based on a public or other authority’s exercise of, or failure to exercise, a special statutory power conferred on the authority. (2) A special statutory power is a power: (a) that is conferred by or under a statute, and (b) that is of a kind that persons generally are not authorised to exercise without specific statutory authority. (3) For the purposes of any such proceedings, any act or omission involving an exercise of, or failure to exercise, a special statutory power does not give rise to civil liability unless the act or omission was in the circumstances so unreasonable 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. (4) In the case of a special statutory power of a public or other authority to prohibit or regulate an activity, this section applies in addition to section 44.”

Section 43A was inserted by the Civil Liability Amendment Act 2003 , and commenced on 19 December 2003 – after the date of Mr Evans’ death. Schedule 1, clause 15 Civil Liability Act extends the operation of section 43A to “civil liability whether arising before or after 13 November 2003”. Whether Section 43A Should be Permitted to be Argued on the Appeal

Section 43A had not been pleaded, or relied on, in the court below. The appeal was argued on the basis that the Court would decide in the course of its reasons whether the RTA was at liberty to rely on section 43A. Thus, we heard argument concerning what the situation would be if section 43A were permitted to be relied on.

Rule 14.14 Uniform Civil Procedure Rules 2005 requires a defendant to plead specifically any matter: “(a) that, if not pleaded specifically, may take the opposite party by surprise, or (b) that the party alleges makes any claim … of the opposite party not maintainable.”

In Port Stephens Council v Theodorakakis [2006] NSWCA 70 at [15], Bryson JA (with whom Giles and Ipp JJA agreed) said: “If s 42 [ Civil Liability Act ] was to be relied on, I am of the view that it was a matter which ought to have been referred to in the claimant’s pleading, having regard to District Court Rules Pt 9 r 9(2); see now Uniform Civil Procedure Rules r 14.14. As well as indicating reliance on s 42, the pleading should have given particulars of the facts which were said to give rise to consideration of s 42. Unless reliance is notified in this way and with particularity, procedural unfairness will arise, particularly as, irrespective of whether the onus of proof of facts under s 42 lies upon a plaintiff or on a defendant, a public authority is in a far better position to know and to adduce evidence relating to the facts with which s 42 deals than any litigant contending with it is likely to be.”

In my view those comments apply equally to section 43A.

In Chilcotin Pty Ltd v Cenelage Pty Ltd [1999] NSWCA 11 at [15]-[18] Giles JA recognised that there are two separate principles concerning the raising of new points on appeal. The first is cast in negative terms, namely that a point not taken in the court below cannot be raised on appeal if evidence could have been given there which by any possibility could have prevented the point from succeeding, or where, if the point had been raised, the respondent might have conducted the case differently at trial. The second principle stems from a requirement that generally parties be bound by the course they adopted at trial. Pursuant to it, except in the most exceptional circumstances, a party who has lost in the court below cannot “raise a new argument which, whether deliberately or by inadvertence, he failed to put during the hearing when he had an opportunity to do so” : University of Wollongong v Metwally (No 2) (1985) 59 ALJR 481 at 483; 60 ALR 68 at 71 (per Gibbs CJ, Mason, Brennan, Deane and Dawson JJ). The two principles have been reiterated by Gleeson CJ, McHugh and Gummow JJ in Whisprun Pty Ltd v Dixon [2003] HCA 48; (2003) 77 ALJR 1598; 200 ALR 447 at [51].

Application of those tests require consideration of what is involved in applying section 43A by comparison to the common law of negligence. I start by considering the position the common law has reached concerning the relationship between statutory powers and duties of care. Common Law Re Relation of Duty of Care and Public Law Duties Re Exercise of Statutory Power

From early in the history of the imposition on statutory authorities of duties of care concerning exercises of their statutory powers it has been recognised that statutory authorities are under duties in public law concerning the exercise of their powers, and an attempt has been made to work out the connection between these public duties and any duty of care to which the authority is also subject.

In Anns , when considering the duty of a council concerning an inspection it had made (not a failure to make an inspection), Lord Wilberforce at 755F-G concluded that to succeed a plaintiff must prove “that action taken was not within the limits of a discretion bona fide exercised, before he can begin to rely upon a common law duty of care. But if he can do this, he should, in principle, be able to sue.”

Part of his Lordship’s reason for making this distinction arose from his consideration, at 755A-B, of an argument that because the local authority has a power to inspect, not a duty, then it cannot be liable for negligent inspection because “if it were to be held so liable … councils would simply decide against inspection.” His Lordship rejected that argument, saying that statutory authorities “must … make their discretionary decisions responsibly and for reasons which accord with the statutory purpose” .

In support of that proposition his Lordship cites Ayr Harbour Trustees v Oswald (1883) 8 App Cas 623 at 639. There, Lord Watson said that the statutory authority there in question: “… shall be vested with, and shall avail themselves of, these discretionary powers, whenever and as often as they may be of opinion that the public interest will be promoted by their exercise.”

The context in which Lord Watson uttered those remarks was a compensation to land case. The Harbour Trustees had resumed some of the land owned by Mr Oswald, and a question arose about the amount of compensation to which he was entitled by reason of the severing of the resumed land from his remaining land. The harbour authorities offered that the conveyance to them from Mr Oswald should be “qualified by a declaration” that they not erect buildings on a particular part of the land that would interfere with Mr Oswald’s access to the harbour. (The unusual form of restriction proposed to be included in the conveyance might be explained by the fact that the land in question was in Scotland.) Mr Oswald refused to agree to that proposal. For differing reasons (Lord Blackburn at 634, Lord Watson at 638) their Lordships decided the case on the basis of the situation that would apply if Mr Oswald had accepted that proposal. They held that the Harbour Trustees could not bind their successors as to the manner in which their successors would exercise their undoubted statutory powers to build whatever they liked wherever they liked on the land. It was as a reason for reaching that conclusion that Lord Watson made the statement that Lord Wilberforce quoted in Anns . That context is vastly removed from a context of deciding whether, and if so in what circumstances, a public authority can owe a duty of care in negligence concerning failure to exercise a power it has, or by reference to what standards one is to judge a breach of such a duty, if it exists.

Lord Wilberforce, immediately after the quotation from Ayr Harbour Trustees v Oswald , went on at 755C-D to say: “If they do not exercise their discretion in this way they can be challenged in the courts. Thus, to say that councils are under no duty to inspect, is not a sufficient statement of the position. They are under a duty to give proper consideration to the question whether they should inspect or not. Their immunity from attack, in the event of failure to inspect, in other words, though great is not absolute. And because it is not absolute, the necessary premise for the proposition ‘if no duty to inspect, then no duty to take care in inspection’ vanishes.”

In my respectful view, this reasoning is unpersuasive. Propositions that a statutory authority cannot by contract fetter the future exercise of a discretion that has been conferred on it by statute, and that it has a duty in public law to consider whether to exercise a statutory discretion, are in a different realm of discourse to whether a duty of care is owed in tort. The powers that are conferred upon a statutory body set the limits of how it is capable of validly acting. If it acts outside those powers, or purports to fetter them in advance, its action will be invalid. However, that a statutory body cannot validly fetter its powers in advance, and must as a matter of public law consider whether to exercise them, says nothing about how it ought act in relation to those powers, bearing in mind the reasonably foreseeable harm that might arise to particular persons, or groups of persons, from a failure to exercise those powers. If a statutory body fails to give proper consideration to the question of whether it should exercise a discretionary power, it would be open to a person with standing to seek an order in the nature of mandamus , requiring the statutory authority to turn its mind to the question of whether it should exercise its power. That the statutory body could be subjected to an order of mandamus says nothing about whether it owes, or has breached, a duty of care.

If one could find in the statute a parliamentary intention that a power should be exercised without regard to any duty of care that might arise under the general law, that intention would of course provide a reason why there was no duty of care – but in the present case there is no basis for finding such an intention. There is nothing unusual about a statutory authority being under duties that spring from two different sources, one in public law and the other in private law.

The balance of Lord Wilberforce’s judgment says nothing about the circumstances in which there could be negligence by reason of failure to exercise a statutory power, beyond the conclusion, at 760F-G: “… that the defendant council would not be guilty of a breach of duty in not carrying out inspection of the foundations of the block unless it were shown (a) not properly to have exercised its discretion as to the making of inspections, and (b) to have failed to exercise reasonable care in its acts or omissions to secure that the byelaws applicable to the foundations of the block were complied with””

In Anns , Lord Salmon differed from Lord Wilberforce. Lord Salmon recognised that a council could be subject to mandamus or certiorari if it decided to make no building inspections at all, but said, at 762E: “I doubt however whether this would confer a right on any individual to sue the council for damages in respect of its failure to have carried out an inspection.”

He was of the view, at 762G-H, that: “If there was no inspection of the foundations before they were covered up, the tenants’ claims would fail because the statute imposed no obligation upon the council to inspect the foundations of these maisonettes nor of any other particular building.”

In Heyman at 445, Gibbs CJ, in the course of considering Anns , said: “… public authorities … are liable for damage caused by a negligent failure to act when they are under a duty to act, or for a negligent failure to consider whether to exercise a power conferred on them with the intention that it should be exercised if and when the public interest requires it.”

While this statement recognises one circumstance in which there can be negligence concerning failure to exercise a statutory power, it does not purport to be exhaustive.

Mason J in Heyman at 457-8 rejected Lord Wilberforce’s account of the relationship between statutory powers and duties of care, saying: “Except in so far as a statute creates a civil cause of action for breach of duty, the distinction between a statutory power and a statutory duty, generally speaking, has limited relevance to civil liability arising out of performance or non-performance of statutory functions. When a statute sets up a public authority, the statute prescribes its functions so as to arm it with appropriate powers for the attainment of certain objects in the public interest. The authority is thereby given a capacity which it would otherwise lack, rather than a legal immunity in relation to what it does, though a grant of power may have this effect when the infliction of damage on others is the inevitable result of its exercise. In framing such a statute it is inconvenient to describe the intended activities of the authority in terms of a series of positive duties. It is preferable to express those activities as functions or powers so that the authority is free to make policy-making decisions and discretionary judgments with a view to attaining the statutory objects. Decisions and judgments of this kind will be involved as a preliminary to exercising, or declining to exercise, a power and, subsequently, in determining how it is to be exercised. Viewed in this light statutory powers are not in general mere powers which the authority has an option to exercise or not according to its unfettered choice. They are powers conferred for the purpose of attaining the statutory objects, sometimes generating a public expectation having regard to the purpose for which they are granted that they will be exercised. There is, accordingly, no reason why a public authority should not be subject to a common law duty of care in appropriate circumstances in relation to performing, or failing to perform, its functions, except in so far as its policy-making and, perhaps, its discretionary decisions are concerned. And, despite possible indications to the contrary in Anns v Merton London Borough Council at 755, 757-758, 760, there is no compelling reason for confining such a duty of care to situations in which a public authority or its officers are acting in excess of power or authority.”

Insofar as that passage contains the phrase “sometimes generating a public expectation having regard to the purpose for which they are granted that they will be exercised” it is not in accordance with later authority rejecting the notion of general reliance ( Pyrenees ), but the balance of the passage has not been overruled.

In Stovin v Wise , Lord Hoffmann, at 950, rejected a suggestion that the fact that a public body had a duty in public law to consider whether it should exercise its powers led to a conclusion that it owes a duty of care.

He saw the fact that a statutory authority had a statutory duty only by virtue of legislation conferring that duty as closely related to whether a duty of care existed concerning the exercise of that statutory duty.

He noted that a cause of action for breach of statutory duty existed only when the construction of the statute led to the conclusion that such an action was intended. An action for breach of statutory duty is, of course, dependent on legislative intention, while an action in negligence arises under the general law. However he said that any duty of care owed by a public authority in performing statutory duties “must be profoundly influenced by the statutory framework within which the acts complained of were done” (quoting from the speech of Lord Browne-Wilkinson in X-(Minors) v Bedfordshire County Council [1995] 2 AC 633 at 739C). He said, at 952H-B: “If such a duty does not give rise to a private right to sue for breach, it would be unusual if it nevertheless gave rise to a duty of care at common law which made the public authority liable to pay compensation for foreseeable loss caused by the duty not being performed. It will often be foreseeable that loss will result if, for example, a benefit or service is not provided. If the policy of the Act is not to create a statutory liability to pay compensation, the same policy should ordinarily exclude the existence of a common law duty of care.”

His Lordship clearly regarded the situation concerning a duty of care in a situation where there had been an omission to perform a mere statutory power as a fortiori . He recognised, at 953B-D, that, even though Parliament legislated to create a statutory power rather than a statutory duty, “It may nevertheless have contemplated that in circumstances in which it would be irrational not to exercise the power, a person who suffered loss because it had not been exercised, or not properly exercised, would be entitled to compensation. I therefore do not say that a statutory ‘may’ can never give rise to a common law duty of care … [b]ut the fact that Parliament has conferred a discretion must be some indication that the policy of the Act conferring the power was not to create a right to compensation. The need to have regard to the policy of the statute therefore means that exceptions will be rare.”

At 953D-E, he identified “the minimum preconditions for basing a duty of care upon the existence of a statutory power” as being: “… first, that it would in the circumstances have been irrational not to have exercised the power, so that there was in effect a public law duty to act, and secondly, that there are exceptional grounds for holding that the policy of the statute requires compensation to be paid to persons who suffer loss because the power was not exercised.”

In application of that test to the facts of the case before him, Lord Hoffmann, at 956-7, held that it was not established that it was irrational for the highway authority to not have moved faster in eliminating the danger at the intersection in question. As well, he held that, even if there had been irrationality in the council having failed to get the bank removed, there were no exceptional grounds that justified a duty of care. The obstruction was obvious, and there was no question of reliance on the council having improved the intersection. The statutory power to require an owner of land adjoining a highway to remove an object that obstructs the view of users of the highway was enacted at a time when there was no liability on a highway authority for breach of its statutory duty to maintain the highway, and hence it was “impossible to discern a legislative intent that there should be a duty of care in respect of the use of that power” (at 958A).

That view of things makes the existence of a duty of care for failure to exercise a statutory power dependent on matters of public law – on the inaction complained of being outside the scope of the power, and also on a finding that, as a matter of construction of the statute, a private law remedy is intended. One can readily agree that if there is a statutory intention that there be no private right of action that statutory intention must prevail, but it is quite another thing to say that there can be no private right of action unless one can find a positive intention in the legislation that there be a private right of action. Setting out on that course invites all the difficulties that there are in the tort of statutory duty about how one ascertains the “statutory intention ” that is said to be the basis of the tort, and adoption of the creative fiction that is often involved in that exercise: see Campbell, “Contribution, Contributory Negligence and Section 52 of the Trade Practices Act” (1993) 67 ALJ 87 at 95-108, & esp fn 109 at 108. It also removes this particular manifestation of the tort of negligence from the usual roots of the tort of negligence in judicial decisions identifying the circumstances in which the law will impose an obligation to take reasonable care, and the connection between “reasonable care” and the standards of behaviour of the community.

When the existence of duties of care owed by public bodies is well established, there is no reason to construe legislation as having proceeded on the basis that a particular public body will be subject to a duty of care concerning exercise of a power only if a positive intention can be found that it be so subject. One does not proceed on the basis that public officials are entitled to commit crimes in the course of exercising their statutory discretions unless there the legislation indicates otherwise, and it is hard to see why entitlement to act without reasonable care should be any different.

As well, it seems to me that applying public law standards of when a decision is unreasonable to a question of whether there has been a failure to act with reasonable care involves conflating two different senses of “reasonable” .

Actions, or failures to act, cannot be assessed as “reasonable” or “unreasonable” in the abstract – they are reasonable or unreasonable by reference to some particular thing: Certain Lloyds Underwriters v Giannopoulos [2009] NSWCA 56 at [115]. In the law of negligence the question of whether there has been a failure to take reasonable care is answered by reference to the law’s requirement that a person in a particular type of relationship to other people take reasonable care to avoid harm to those other people (ie the nature of duty of care owed), and the question of whether reasonable care has been taken is judged by reference to what a reasonable person would have done for the purpose of avoiding harm to the person to whom the duty is owed: cf Lloyds v Gianoppoulos at [105]-[106]. By contrast, a question about whether a statutory authority has acted in a way that is not a reasonable exercise of or failure to exercise its statutory power must be answered by a reference to the proper scope of that particular statutory power. That will depend upon the construction of the statutory instrument that confers the power, and the purpose for which the particular power was conferred. While there might be particular powers concerning which the potential for an exercise of the power to cause harm to some predictable person or group of people will be relevant to deciding the scope of the power, in very many cases matters other than the potential of a particular action or inaction to cause harm will enter into deciding whether that action or inaction is one that a reasonable statutory authority would not have engaged in.

It seems strange that a question of negligence – whether reasonable care has been taken to avoid causing harm to particular people – could be dependent on matters that have nothing to do with whether reasonable care has been taken. Of course, what is involved in taking reasonable care in any particular situation can be influenced by other responsibilities that a defendant has, but the focus of the enquiry is on whether reasonable care has, or has not, been taken in all the circumstances. The focus of the inquiry in deciding a question of Wednesbury unreasonableness is wider, and different.

As well, in the administrative law context, the Wednesbury test of invalidity presupposes that there was a range of decisions that a reasonable authority could have made in exercising validly its power to act or not act concerning some subject matter, and it is only if the action of the authority is outside that range of possible reasonable decisions that it is beyond power, and hence invalid. The question of whether there has been a breach of duty of care in the law of negligence operates differently to this. The test for appellate intervention in decisions about breach of a duty of care, when the primary facts are established, has been laid down in Warren v Coombes (1979) 142 CLR 531, and it concerns “whether the learned trial judge was correct in finding that the respondent was not negligent” (at 536), or “the proper inference to be drawn from facts … ” (551). Such tests can only be answered by a response that “the trial judge was correct” or “the trial judge was not correct” , or “the proper inference to be drawn is that the defendant failed to take reasonable care” , or “the proper inference to be drawn is that the defendant took reasonable care” . Similarly, a trial judge deciding a question of breach of duty of care is trying to find the (single) right answer to the question of whether there has been a breach of a duty of care, not to choose which answer is preferable out of several that are legitimately open. The question of whether there has been a breach of duty of care is, of its nature, one concerning which a court cannot recognize that there is a range of legitimately open answers to the question of whether the defendant failed to take reasonable care.

There are some indications in later High Court authority that the Stovin v Wise approach to the relationship between public duties and duties of care concerning failure to exercise statutory powers is not accepted by at least some judges. In Graham Barclay Oysters , Gleeson CJ said, at 562 [32]: “It is necessary to distinguish between a public duty, enforceable by mandamus, to give consideration to the exercise of a power, and a legal duty, owed to a citizen, to exercise the power.”

And at 563 [35]: “… the circumstance that, in the public interest, certain powers of regulation of activity within its area are vested by statute in the Council does not mean that the Council owes a legal duty to individuals or classes of person whose health may be affected, directly or indirectly, by decisions made as to the exercise of those powers.”

At 575 [80], McHugh J recognised that a statutory authority can sometimes be liable for negligence for failing to exercise a statutory power. He continued: “But it invites error to think that the common law has converted the discretionary statutory power into an affirmative duty to exercise the power. The common law cannot interfere with the exercise of the discretion and require the authority to enforce the power. To attempt to do so would bring the common law into conflict with the legislative intention that the exercise of the power be discretionary. The common law avoids the conflict by holding that in the circumstances the failure to exercise the power is a breach of a common law duty existing independently of the statute. The common law duty may or may not be an affirmative duty to take reasonable care to protect the plaintiff from harm. However, the existence of the statutory power does not create the common law duty although in some cases — particularly in reliance cases — it may be an important factor in finding that a duty of care was owed.”

In the absence of authority binding me to do otherwise, I would take the view that principled development of the common law is not in accordance with the Stovin v Wise approach to this topic. Construction of Section 43A

I now turn to compare the common law position with that arising under section 43A.

Considered as a piece of text, the preferable reading seems to be that section 43A(3) imposes a more stringent standard for the existence of liability arising from exercise or failure to exercise a “special statutory power” than had applied, in at least some cases, before the enactment of section 43A. If an authority was subject to a duty to exercise care in exercising, or failing to exercise a special statutory power, the question of whether there had been a breach of that duty would have been determined, before the enactment of section 43A, by reference to the court’s own finding about whether there had been a failure to take reasonable care in exercising, or failing to exercise, that power. The preferable reading of section 43A(3), considered as text, is that it either replaces or supplements that standard by a standard that in its wording is more akin to the standard used in administrative law to decide whether an exercise of power is a valid exercise.

I turn to the legislative history to see whether that possible reading is the correct one. Section 43A was previously introduced by the Civil Liability Amendment Act 2003 . Section 43 had been introduced into the Civil Liability Act by the Civil Liability (Personal Responsibility) Act 2002 . Section 43 had introduced a requirement that an act or an omission “was in the circumstances so unreasonable that no authority having the functions of the authority in question could properly consider the act or omission to be a reasonable exercise of its functions” , so far as an action for breach of statutory duty was concerned. Though the Second Reading Speech concerning the 2003 bill paraphrases without further explanation section 43A(3) it can be taken to be an implementation of the same policy as was involved in the incorporation of the self-same test in section 43.

The Ipp Report considered, in chapter 10, the application of the law of negligence to the liability of public authorities. That report was written after the High Court decision in Brodie . It noted, at [10.7] (p 152), that the requirement at 577-8 [151] of Brodie to have regard to “competing or conflicting responsibilities or commitments of the authority” in deciding whether a public authority took reasonable steps to reduce a risk “… has been that increasing amounts of time are spent in the course of trials considering whether the authority’s conduct in relation to the risk in question was reasonable given the other demands on the resources available to the authority.” The present case is a fine example.

At [10.21] (p 156), the Ipp Report rejected a suggestion that a public authority should have a defence in answer to any claim “for negligence in the performance of a statutory function” . The reasoning was: “Suppose, for example, that an employee of a public authority causes a road accident by driving negligently in the course of performing some statutory function of the authority. The mere fact that the accident occurred in the course of the performance of a statutory function should not displace the operation of the ordinary rules of liability and allow the policy defence to be pleaded.”

It can be seen that the definition of “special statutory power” seeks to meet that concern.

At [10.26] (p 157), the Ipp Report rejected a suggestion that some policy decisions such as “quasi-legislative” or “regulatory” decisions (expressions used, eg, in Sutherland SC v Heyman at 500 per Deane J, and Crimmins at 37 [87] per McHugh J) were “non justiciable” . The report continued (at [10.26]-[10.27]: “Rather, we think that Australian law should follow the lead of English law in this respect (see Stovin v Wise [1996] AC 923) by providing that in a claim for negligently-caused personal injury or death against a public functionary, where the alleged negligence consists of the exercise or non exercise of a public function, and the public functionary pleads that the failure to take precautions to avoid the relevant risk was the result of a decision about the allocation of scarce resources or was based on some other political or social consideration, liability can be imposed only if the decision was so unreasonable that no reasonable authority in the defendant’s position could have made it. This test of ‘unreasonableness’ is taken from public law where it is known as the test of ‘Wednesbury unreasonableness’ after the case in which Lord Greene MR invented it ( Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223; [1947] 2 All ER 680). The effect of the test is to lower the standard of care. It does not provide the defendant with an immunity against liability, but it does give the defendant more leeway for choice in deciding how to exercise its functions than would the normal definition of negligence (in terms of reasonable care).”

In the Second Reading Speech introducing the 2002 legislation, the Premier acknowledged that, “most of the changes implement or draw on the recommendations by the … Ipp report” (New South Wales Legislative Assembly, Parliamentary Debates (Hansard), 23 October 2002, 5764 at 5765). The Second Reading Speech also said (at 5767): “Public authorities carry out what is often a limitless task with necessarily limited resources. We must ensure, therefore, that it is not left to the courts to determine a public authority’s expenditure on its tasks. In keeping with this approach, the bill will also provide immunity for a public or other authority for breach of statutory duty, unless it has acted irrationally.”

It can confidently be said that the standard that section 43A imposes is not the same as that by which the reasonableness is assessed for the purposes of deciding whether there has been a breach of a particular duty of care. I say that because it is clear that by enacting section 43A the legislature was intending to alter what would otherwise be the law by which the negligence of public authorities was decided.

Consideration of the legislative history of section 43A confirms the reading of its text, that what was intended was to implement the sort of test that had been adopted as part of the English common law in Stovin v Wise . However, section 43A(3) does not purport to give an exhaustive account of when it is that there can be negligence in exercising or failing to exercise a special statutory power. What it does, by adopting a form of words that there is no civil liability unless … , is to state a precondition for the existence of civil liability in the sort of circumstances to which it is addressed. One would need to look to the pre-existing common law of negligence to ascertain when it was that there was a duty of care, and whether there had been what the common law would regard as a failure to exercise reasonable care. Section 43A(3) imposes an additional requirement, beyond those of the common law, before liability can be established.

There are some other questions about the construction and operation of s 43A that were raised by Allsop P in Precision Products (NSW) Pty Ltd v Hawkesbury City Council [2008] NSWCA 278 at [171]-[176] that are unnecessary to answer for present purposes. It is enough, for present purposes, that to permit section 43A to be raised now would involve the question of whether there had been a breach of duty being decided by a different standard to that which was applied at the trial. That is such a significant departure from the way the trial was conducted that it should not be permitted.

The present situation differs from that in Jaber v Rockdale City Council [2008] NSWCA 98; [2008] Aust Torts Reports ¶81-952 (61,700) at [20]-[21], where Tobias JA (with whom Handley AJA and I agreed) held that a trial judge had been correct in permitting reliance upon certain provisions of the Civil Liability Act when the facts that triggered the operation of those provisions had been pleaded, though the specific statutory provisions themselves, identified by section number, had not. Would Section 43A have Helped RTA?

There are two separate reasons why, even if section 43A had been permitted to be raised by the RTA on the appeal, it would not have improved the RTA’s position. The first is because the erection of protective screens on the bridge would not have involved the exercise of any “special statutory power” by the RTA. The second is because the plaintiff’s allegation in the present case was not that the RTA’s liability was based on the exercise of, or failure to exercise , any special statutory power conferred on the authority. No “Special Statutory Power”

As the Second Reading Speech on the introduction of the 2003 amending legislation explicitly recognised (New South Wales Legislative Assembly, Parliamentary Debates (Hansard), 13 November 2003, 4992), section 43A was a legislative reaction to the decision of Adams J in Presland v Hunter Area Health Service [2003] NSWSC 754. In Presland , damages were awarded to a mentally disordered person who the health service had failed to detain, and who, within hours of being released from a psychiatric hospital, killed someone concerning whom he had a delusion. That this is the basis for enactment of section 43A is not affected by the fact that Adams J’s decision was subsequently reversed on appeal: Hunter Area Health Service v Presland [2005] NSWCA 33; (2005) 63 NSWLR 22. The power of a medical superintendent to detain a mentally disordered person, specifically conferred by the Mental Health Act 1990 , is a clear example of a “special statutory power” .

When asked in argument to identify the special statutory power involved in this case, Mr Maconachie pointed to section 120 Roads Act 1993 , which provides: “(1) The Minister may order a roads authority to take such action, or to forbear from taking such action, with respect to the regulation of traffic on a public road as may be specified in the order. (2) For example, the Minister may order a roads authority to erect or to remove any notice or barrier. (3) Such an order may be given whether or not the roads authority is otherwise empowered to regulate traffic in the manner specified in the order.”

I do not accept that section 120 would be the source of the authority’s power to erect protective screens on the overpass. Section 120 would need to be interpreted by reference to the definition of “regulate travel” in the Dictionary to the Act, which provides: “ regulate traffic means restrict or prohibit the passage along a road of persons, vehicles or animals.”

Erection of protective screens does not “regulate traffic” within the meaning of that definition, as the very nature of the screens is that they are erected at the sides of the overpass, and in no way restrict or prohibit passage along either the freeway, or the overpass, of persons, vehicles or animals.

Further, in my view, the ownership of the bridge is quite sufficient to empower the RTA to erect screens on it. Its ownership of the bridge is not a “power … conferred by or under a statute” . Rather, it is a property right.

A citizen concerned about the lack of fencing on the bridge would not be legally free to take unilateral action to fix the problem himself, by erecting screens without permission from the RTA. However, one reason for this is that erecting the fencing would be a trespass on the RTA’s land. Another is that erecting it would require development consent, which could not be applied for without the consent of the land owner (at the time of the alleged negligence, this was provided for by section 77(1) Environmental Planning and Assessment Act 1979 ; see now Environmental Planning and Assessment Act section 78A(1) and Environmental Planning and Assessment Regulation 2000 clause 49(1)). It is not because the exercise of a statutory power, which the RTA has but ordinary citizens lack, is needed to achieve the fencing of the bridge.

Exercise of a statutory power of some sort would have been required to empower the RTA to deny members of the public their statutory right under section 5 Roads Act to pass along the Glenlee Bridge. (In making that remark I leave aside whatever distinction there might be between a statutory power simpliciter, and a special statutory power). However, it has not been shown that the process of erecting screens on the bridge would have required the bridge to be closed in a way that denied members of the public that statutory right. Liability Not “Based on Exercise of, or Failure to Exercise” Special Statutory Power

The Respondent’s case was not that the Minister was negligent in failing to give a direction under section 120, but rather that the RTA was negligent in failing, of its own initiative, to screen the bridge sooner.

There is a well recognised category of negligence that: “… when statutory powers are conferred they must be exercised with reasonable care, so that if those who exercise them could by reasonable precaution have prevented an injury which has been occasioned, and was likely to be occasioned, by their exercise, damages for negligence may be recovered”. ( Caledonian Collieries Ltd v Speirs (1957) 97 CLR 202 at 220 per Dixon CJ, McTiernan, Kitto and Taylor JJ.)

On the basis of Speirs , Mason J in Sutherland SC v Heyman at 458 said: “It is now well settled that a public authority may be subject to a common law duty of care when it exercises a statutory power or performs a statutory duty.”

Such a duty has also been recognised by McHugh J (with whom Gleeson CJ agreed) in Crimmins at [62] ff. However, a reading of the Statement of Claim shows that the plaintiff made no mention of any statutory power, or of negligence in the exercise of any such statutory power, in the way it put its case. Rather, the acts of negligence alleged were those that I have set out at para [107] above. When the way the plaintiff puts its case for the liability of the RTA makes no mention of any special statutory power of the RTA, I do not see how that alleged liability could be “based on” the RTA’s exercise of, or failure to exercise, any special statutory power. The distinction between a case asserting negligent exercise of a statutory power, and a case that a statutory authority had an affirmative obligation to take reasonable steps to prevent harm to a plaintiff is recognised by McHugh J in Crimmins at [62]-[70].

Even if it had been the case that construction of the screens had required temporary closure of the bridge, it would not have been negligence in exercising any power to close the bridge, or even delay in exercising a power to close the bridge, that is the basis of the plaintiff’s claim. Thus, any civil liability of the RTA in the present case would not be “based on … failure to exercise a special statutory power” and so the condition for the operation of section 43A would not arise.

In these circumstances, it is not necessary to give consideration to other cases that have concerned limitation of the liability of public authorities in exercise of their powers, such as Board of Fire Commissioners (NSW) v Ardouin (1961) 109 CLR 105; Hudson v Venderheld (1968) 118 CLR 171; Australian National Airlines Commission v Newman (1987) 162 CLR 466, or Puntoriero v Water Administration Ministerial Corporation [1999] HCA 45; (1999) 199 CLR 575. Nor is it necessary to consider whether the terms of section 43A require a distinction to be drawn between a “specific statutory authority” and a general statutory authority, and if so by reference to what criteria the distinction should be drawn. Some consideration of those matters is given in Watson, “Section 43A of the Civil Liability Act 2002 (NSW): Public law styled immunity for the negligence of public and other authorities?” (2007) 15 Torts Law Journal 153. Section 42 Civil Liability Act

The Respondent was asserting in the court below that the RTA had a “civil liability in tort” to Mr Evans, his widow and children, and the RTA did not contend that any such civil liability was of a type that was excluded from the operation of Part 5 of the Civil Liability Act by section 3B of that Act. Thus, section 40 Civil Liability Act had the effect that Part 5 of that Act (which runs from section 40 to section 46 inclusive) applied to that liability.

Section 41 Civil Liability Act 2002 provides: “In this Part: exercise a function includes perform a duty. function includes a power, authority or duty.”

Section 42 Civil Liability Act 2002 provides: “The following principles apply in determining whether a public or other authority has a duty of care or has breached a duty of care in proceedings for civil liability to which this Part applies: (a) the functions required to be exercised by the authority are limited by the financial and other resources that are reasonably available to the authority for the purpose of exercising those functions, (b) the general allocation of those resources by the authority is not open to challenge, (c) the functions required to be exercised by the authority are to be determined by reference to the broad range of its activities (and not merely by reference to the matter to which the proceedings relate), (d) the authority may rely on evidence of its compliance with the general procedures and applicable standards for the exercise of its functions as evidence of the proper exercise of its functions in the matter to which the proceedings relate.”

It is common ground that the RTA is a “public or other authority” within the meaning of this provision.

Section 42 was included in the Civil Liability Act by the Civil Liability Amendment (Personal Responsibility) Act 2002 . Pursuant to clause 6 of Schedule 1 of the Civil Liability Act , those amendments: “… extend to civil liability arising before the commencement of the amendments, but do not apply to or in respect of proceedings commenced in a court before that commencement.”

Section 42 commenced on 6 December 2002. The proceedings in the court below were commenced on 28 October 2004. Thus, section 42 applied in those proceedings.

The particulars that the RTA pleaded concerning its defence under section 42 were: “The highway on which Mr Evans was travelling, the Hume Highway, being part of the National Highway, was funded by the Commonwealth Government. Financial resources provided to the defendant were insufficient to achieve the erection of barriers or screens on each of many bridges over the highway by the time of Mr Evans’ death. It is not conceded, indeed it is denied, that the measures specified [in the particulars of negligence] in the Statement of Claim, or any of them, would have prevented the actions of the person [who threw the concrete as alleged in] the Statement of Claim. The defendant had allocated such resources as were available to it for the maintenance of the highway in accordance with its own determination of priorities. The defendant, at the time of its alleged breach of duty, had the broad range of duties of a roads authority and had attended to those duties. The defendant had complied with the general procedures and applicable standards for the exercise of its functions as a roads authority.”

In its Amended Defence in the court below, the RTA had also placed reliance upon section 45 Civil Liability Act . As the judge records at [112] in his judgment, during argument in the court below, section 45 was not pursued. Though some mention was made of section 45 in written submissions on the appeal, Mr Maconachie ultimately did not seek to press it.

The factual findings that the judge made at [60] of his judgment were made without reference to section 42 Civil Liability Act . While section 42 is a matter that a defendant must plead, once it has been pleaded (as it was in this case) the principles set out in section 42 must be taken into account in deciding whether there has been a breach of a duty of care. Though the judge did not say anything, in his reasons up to and including para [60] about how section 42 impacted on breach of duty of care, I would not regard that as a separate reason for overturning his conclusions of unreasonableness in para [60]. That is because, later in his reasons, the judge explained why he did not regard section 42 as applicable. However the adequacy of those reasons should themselves be examined.

The trial judge at [114] summarised the submission of counsel for the RTA at the trial: “Mr Perry took the plaintiff’s position, that had more money been allocated for bridge screening on the national highway there would have been a screen on the Glenlee Bridge prior to 23 August 1998 when the fatal incident occurred, as a direct challenge to the general allocation of the defendant’s resources and so specifically forbidden by s 42(b). Further, counsel referred to s 42(c) as requiring consideration of the broad range of the defendant’s activities and not merely bridge screening. And, so counsel submitted: ‘By application of s 42, then, the duty owed by RTA did not extend to reaching beyond the monies reasonably available to it to bring forward the screening of bridges at the expense of other projects. The allocation of resources is not open to challenge, and that challenge therefore fails.’”

The judge held (at [117]) that section 42 was inapplicable in the present case: “In my view, if s 42 was to be read strictly, as I apprehend Mr Perry did, then a public authority could simply plead para (b) in most, if not all, cases so as to avoid exposure to liability; that cannot be the intention as it would effectively grant immunity for a public authority from civil liability – all public authorities are granted resources and allocate them in satisfaction of their functions and consistent with their statutory powers. However, to so read the paragraph, I think, would be to deny a proper and responsible approach to its true meaning. In my view, the paragraph is to be seen in terms of a bar to the challenge of ‘the general allocation’ (my emphasis) of resources, that is, not the allocation in any particular case to the certain specified project. There is no challenge here to the defendant’s general allocation of its resources but only to the allocation of sufficient funding to screen overbridges from the safety and urgent minor works budget and, even then, only for the Glenlee Bridge. Reliance by the defendant on s 42 is rejected.”

Like section 43A, section 42 presupposes the existence of the law of negligence – it applies “in proceedings for civil liability to which this Part applies” , and, in the course of such proceedings, “in determining” whether a public or other authority has a duty of care, or “in determining” whether the public or other authority has breached a duty of care. Section 42 is in the nature of a supplement or corrective to the pre-existing law of negligence, not a replacement for it or for any part of it.

Both section 42(a) and (c) used the phrase “the functions required to be exercised by the authority” . As section 42 deals quite generally with the way one should proceed in deciding whether a public or other authority has a duty of care or has breached a duty of care, the force of “required” seems to me to be “required by the law of negligence” – ie, required by the legal standard of taking reasonable care not to harm in a relevant way a person to whom the public or other authority owes a duty of care. It does not refer to a requirement in the nature of a statutory duty.

The whole topic of section 42 concerns the manner in which the public or other authority exercises its “functions” . The definition of “function” in section 41 is only an inclusive one, but section 42 must be able to work in a way that makes sense by reference to at least situations where a function is a power, a function is an authority, and a function is a duty.

There may be room for an argument that the “ functions ” of a public authority, within the meaning of section 42, concern those activities or responsibilities that are conferred on or entrusted to it by legislation, and that those “ functions ” do not cover absolutely everything that the authority can do. However it is not necessary to decide whether that is so, because in the present case the RTA has a function of care, control and management of freeways conferred on it by the Roads Act . When we have not been pointed to any legislation that confers specific responsibilities on the RTA concerning the National Highway, we cannot conclude that its functions are any narrower than the care, control and management of all freeways in NSW. There may well be other cases where a “ function ” of an authority can be identified that is much narrower than this, but that will be a matter of construction of the particular legislation involved in that other case.

In the present case, the Federal Government exercised minute control over the items on which money for the National Highway was expended. However that does not lead to the consequence that, for example, the RTA had a function of screening two bridges over the F5 in a particular year. The practical reality of how the Federal Government exercised its control was that it would state how much money was available in a particular year for, for example, safety and urgent minor works, and it was largely a matter for the RTA to decide on which specific projects that money would be expended. The fact that its allocation amongst projects then needed to be approved by the Federal Government did not, on the evidence in this case, result in any alteration to the RTA’s initial proposals. When we have not had argument in the present case about the manner in which the system tied grants works, and in particular about its legislative basis, I would not be prepared to conclude that the provision of safety and urgent minor works on the National Highway was a separate function of the RTA, within the meaning of section 42.

The “resources reasonably available to the authority” would include resources that the authority in question already has and that it might reasonably be expected to expend in provision of its functions, and any that might reasonably be provided to it. If, for example, an authority had real estate that it used for its offices or works depots, they are resources of the authority, but not might not be reasonably available for the purpose of exercising any of the functions that are the subject of debate in litigation. The “resources reasonably available to the authority for the purpose of exercising its functions” would also include such amount as had been made available to that authority by appropriation (in the case of authorities funded from the state budget), or from other sources of income like levying of rates, or asset sales, or grants from some other level of government. The wording of section 42(a) seems to leave open an argument about whether more money than was actually made available to the authority from external sources was “reasonably available” to it.

Insofar as the NSW government had decided that it would devote none of its own money to the National Highway, and would rely on Federal funds for that sort of expenditure, it would only be if it could be said that more money than the Federal Government actually made available for expenditure on the National Highway was “reasonably available” to the RTA that the possibility of spending such money could enter into a question of whether the RTA had been negligent in the present case. For the reasons I have earlier given, there is no reason to conclude that more money was “reasonably available”.

The effect of section 42(a) in the present case is that what the RTA can be required by the law of negligence to do is limited by the financial and other resources that are reasonably available to the RTA for the purpose of carrying out the care, control and management of freeways and any other roads that are under its care, control and management. Its budget for that purpose is so large that any expenditure that would have been involved in earlier screening of the Glenlee Bridge, or indeed in earlier screening of all overpasses on freeways would have been well within its budget.

When section 42(b) uses the expression “those resources” it is referring back to section 42(a). What section 42(b) requires not to be challenged, in the present case, is the “ general allocation ” by the RTA of those resources that are reasonably available to the RTA for the purpose of the care control and management of freeways and other roads under its care control and management.

There is an important difference in prepositions between section 42(a) and section 42(b). Section 42(a) is concerned with the resources reasonably available to the authority, while section 42(b) is concerned with the allocation of those resources by the authority. In other words, section 42(b) starts from the position that certain resources are reasonably available to the authority, and considers the allocation that is made by the authority of those resources.

Effect must be given to the word “general” in section 42(b). It seems to be drawing a distinction between the general and the specific. It will be a matter that needs to be decided concerning any particular set of resources that is allocated to a public authority, whether a particular decision about allocation of those resources by the authority is regarded as a decision about the general allocation of resources, or a decision about the specific allocation of resources.

The force of the words “is not open to challenge” in section 42(b) is to prohibit a particular manner of contending that a public or other authority is under a duty of care, or has breached a duty of care. Thus, in a case like the present, which concerns an allegation of breach of duty of care, application of section 42(b) needs to be carried out bearing in mind each particular manner in which it is alleged a duty of care has been breached.

In State of New South Wales v Ball [2007] NSWCA 71; (2007) 69 NSWLR 463 at 466-7 [13]-[18], Ipp JA (with whom McColl JA and Young CJ in Eq agreed) was able to decide, on the strength of the particulars in a statement of claim, that at least some aspects of the claim a plaintiff wished to make involved a challenge to the general allocation of resources affecting the plaintiff’s work as a police officer. In consequence, his Honour struck out those allegations from the statement of claim. Sometimes it might happen that the particulars are narrowly drafted, or that a defendant chooses not to make such a strike out application, but once the case is heard it is apparent that the substance of a complaint of negligence that plaintiff is making involves a challenge to the general allocation of resources by a public authority.

In the present case, if one allegation had been that the RTA misapplied well-established principles and made careless factual errors in the way it prioritised overpasses for screening, and that a principled and careful prioritisation process would have put the Glenlee Bridge close enough to the top of the priority list to have been screened before 23 August 1998 with the money that the RTA actually chose to spend on bridge screening, the challenge that was being made would have been to the allocation of resources that the RTA had actually allocated to bridge screening. I do not think that such a challenge would be one to the general allocation of the resources reasonably available to the RTA for the purpose of exercising its functions.

It is elementary that deciding any allegation of breach of duty of care involves considering what is the reasonable response to the risk. In Romeo , the injury to the plaintiff could have been prevented if the authority in question had adopted the very cheap expedient of placing a couple of star pickets on either side of the track that the plaintiff went along before falling over the cliff, and connecting those star pickets with a few pieces of wire. However the judgments in the High Court made clear that approaching the matter in that way was not the relevant way, because the risk in question was one of a person falling over a cliff anywhere along the substantial length of coastline that the authority controlled. What needed to be considered was the reasonable response to that risk.

Similarly, in the present case, if the allegation had been that the RTA did not take seriously enough the risk of objects being dropped from overpasses, and should have spent more money on remedying that risk than on, for instance, providing warning lights and protective barriers at railway level crossings, the challenge would be to the allocation by the RTA of money to screening overpasses as opposed to other road safety measures. One would need to decide whether a challenge of that type was to the “general allocation” by the RTA of the resources reasonably available to it for the purposes of exercising its functions.

Concerning allegations that are permissible in accordance with section 42(b), whether such an allegation is accepted is a matter of application of the common law plus principles of the Civil Liability Act other than section 42(b) and any other relevant legislation.

In light of the conclusion I have come to about breach of duty under the general law, it is not necessary to answer that question. It can be said, however, that it could in principle be wrong to apply section 42(b) by saying that all that is being challenged is the failure to do the comparatively small amount of work that would have been needed to prevent the particular injury that the particular plaintiff suffered. To approach the matter in this way pays insufficient attention to the detail of the argument that the plaintiff put as to the particular ways in which the defendant has failed to exercise reasonable care. It is in principle quite possible for one way in which it is alleged a public or other authority has failed to exercise reasonable care, resulting in a particular plaintiff being injured, to involve alleging that the taking of reasonable care would have required the authority to make a different general allocation of resources than that which it in fact made, while another way of alleging that the public or other authority has failed to exercise reasonable care, resulting in the same injury, does not involve alleging that the taking of reasonable care would have required the authority to make a different general allocation of resources to that which it in fact made.

The reasoning process I have engaged in earlier in this judgment – considering the question of whether the RTA’s taking reasonable care required it to screen the Glenlee Bridge earlier – has been carried out by taking into account the broad range of the RTA’s activities. At least in the present case, section 42(c) adds nothing to the common law.

Section 42(d) was not invoked in the present case.

In all these circumstances I conclude that section 42 does not affect the conclusion to which I have arrived by the application of common law principles. Issue 2(4) – Negligence in the Evaluation Matrix The RTA’s Risk Analysis Methodology

The methodology that Mr Onggo had developed by the end of 1996 was a more elaborate version of the risk analysis methodology that the working party had developed by the end of 1995, and was produced with the assistance of a technique called the Kepner-Tregoe method of problem-solving and decision-making. It involved identifying a total of 11 different factors relevant to the risk of an object being dropped from an overpass. Those factors were: whether there was pedestrian access; the type of road under the bridge; whether it was near a school; whether it was near a hotel or club; whether it was near a youth attraction; whether it was near other pedestrian generators such as shopping centres, bus and train stations or “high resident” ; the lighting conditions; the exposure to buildings; the exposure to traffic; whether there was a history of incidents and/or signs of graffiti; and whether there were any loose objects nearby.

Concerning each of those risk factors, a priority rating was allocated, attributing a numerical value to the importance of the risk factor. Concerning each of the risk factors, various measures of the extent to which that factor was present were identified, and a numerical value given to each of those measures of extent. For example, concerning the risk factor of pedestrian access, measures of extent of “no access” (given a rating zero), “shared with motorist” (given a rating of 8) and “pedestrians only” (given a rating of 10) were identified. All the bridges being ranked in order of priority were inspected individually, and evaluated against the risk assessment criteria. In the course of evaluating a particular bridge, if there was pedestrian access shared with a motorist, a score of 80 (obtained from multiplying the priority rating of 10, and the measure of the extent to which the factor was present of 8) would be allocated to the risk factor concerning pedestrian access. That process was repeated in relation to each of the risk factors, and an overall score obtained for the particular bridge. The bridges so evaluated were then listed in order in an evaluation matrix, so that the bridge with the highest assessed risk was at the top of the list, and the bridge with the lowest assessed risk was at the bottom.

In developing the evaluation matrix, Mr Onggo made enquiries of the Victorian and Queensland road authorities, and the New South Wales railway authority, to find out whether they had any method of prioritising screening of bridges, but they had no such methods. The evidence has not identified any publications in the professional literature of any such methods, or established that there were any people who had expertise concerning that topic. While the 1969 American publication does not go so far as to state a method for prioritising screening, it identifies some risk factors (para [149] above), the substance of which have been taken account of in Mr Onggo’s methodology.

Various overbridges that the RTA was considering screening were evaluated in accordance with that risk assessment, and an order of priority was produced for the screening of those bridges. In one version of the evaluation matrix dated 14 March 1997 the Glenlee Bridge was accorded priority number 36 out of a total of 77 bridges. In another version of the evaluation matrix dated 25 May 1998 – the last produced before Mr Evans’ death – it was accorded priority number 42 out of 84 bridges.

In assessment of the Glenlee Bridge, it had been rated as having no pedestrian access. Mr Onggo said this was because the way the methodology worked was that “pedestrian access” required there to be a walkway or footpath for pedestrians, not merely that pedestrians were physically capable of crossing the bridge by walking on the road. He explained that the risk being evaluated by the risk assessment survey was the risk of spur of the moment or opportunistic launching of objects from bridges, not planned ones.

Another feature of the application of the evaluation matrix to the Glenlee Bridge was that it was given a score of zero for “history of incidents/graffiti” . It was given that score because Mr Onggo was not aware of the previous incidents that had occurred when objects were dropped from the bridge. If the rating attributed to the Glenlee Bridge were adjusted to take account of there having been previous incidents, and no other change to the evaluation matrix were made, the Glenlee Bridge would have advanced to number 22 in priority. That calculation was done on the basis that Mr Onggo had adopted when compiling the matrix, that prior incidents without graffiti would result in a score of 6. He accepted in cross-examination that if, contrary to the methodology he had actually adopted, prior history resulted in a score of 10 on that variable, and no other change was made to the evaluation matrix, the Glenlee Bridge would have risen in priority to be between 16 and 17. The Additional Particular of Negligence

In the course of the appeal, the Respondent sought to add a further particular of negligence, namely: “(h) Failing to properly determine the rating of the Glenlee Bridge in its risk assessment, in that: (i) It failed to categorise the bridge as being a bridge affording shared access to motorists and pedestrians; (ii) It failed to characterise the bridge’s comparative isolation as indicative of a higher risk of objects being thrown or dropped from it; (iii) It failed to appreciate and have regard to the history of objects being dropped from bridges in the Campbelltown area and the Glenlee Bridge in particular.”

The plaintiff had obtained a large volume of material from the RTA on subpoena prior to the hearing. It had supplied the risk assessment summary and evaluation matrix by which the screening of bridges was prioritised to its expert, Mr Jamieson. He annexed them to his first report. In the body of his report, Mr Jamieson identified the 11 individual risk factors, and commented: “It is not known on what basis these individual ‘risk factors’ were identified, noting that the frequency of such behaviour is relatively rare. That is, there was probably no opportunity for the RTA contractors to analyse any significant number of past incidents to determine a data-based list of potential factors. Therefore the list was possibly derived from subjective assessments. … The key issue arising from this methodology concerns the scoring of each individual site – and its resultant prioritising. For example, a site can score ‘20’ because of its proximity to a school and ‘youth attraction’. Such a site may eventually score higher than an isolated site, and its ultimate priority for rectification would be high. This example ‘risk factor’ involving schools presupposes a process or sequence of events that involves school children throwing objects off bridges while walking to or from school in broad daylight. While the detail of previous incidents is not known, subjectively, it is considered unlikely that such a possible scenario should receive any significant weight. An alternative, more likely process leading to concrete being dropped from a bridge, would be one which occurs at locations away from prying eyes, ie under cover of darkness at remote and isolated locations. While it is beyond the scope of this current report to propose an alternative risk assessment, with an alternative list of potential risk factors, it would appear on analysing the present material (particularly documents C5 and C6) that the prioritising of these bridge rectifications would not seem to be based on any “formal” Risk Engineering principals [sic] (at least not those presented in the literature). Rather, the prioritising seemed to be based on a more subjective process. If a more formal risk analysis process had been applied, the identified ‘risk factors’ probably would have been different. The review of the papers shown in Appendix C, suggests that the Glenlee Bridge probably should have been given a higher score than it eventually received (handicapped as it was for not being near a school). It follows that a higher score would have increased its rectification priority. It should be emphasised that these comments are not taken in ‘20-20 hindsight’, but rather from a risk engineering viewpoint which should take into account risk factors arising from a formal analysis of the phenomenon involved.”

Comparatively early in the first day of the trial Mr Jamieson’s report was tendered. Counsel for the RTA objected to the expressions of opinion relating to the evaluation matrix. In the course of dealing with that objection, the judge enquired whether he was going to be getting evidence from the defendant from engineers about risk assessment. Counsel for the RTA said that Mr Onggo would be called as a witness and would explain how the order of priority was established. The judge allowed Mr Jamieson’s evidence.

Mr Jamieson was cross-examined about his opinion that the Glenlee Bridge should have been placed higher in the order of priority of bridges. He agreed that seeking to prioritise bridges by reference to risk factors was “a rational practice” . His cross-examination on that topic included: “A. My – admittedly – unsophisticated view, and commonsense could be a wise – just because it was near a school didn’t necessarily make it a higher risk location given the school kids walk backwards and forwards at school in broad daylight presumably around traffic. Q. Isn’t it the case that we are just in the area of speculation? Could I not put that to you in response. There are different types of rock dropping; all of them potentially lethal. A brick only needs to hit a windscreen to cause a windscreen to shatter, to cause a fatal accident. And so if we have one form of accident which is the type that occurred in this particular case, where bricks are loaded – rocks are loaded into the back of a vehicle and driven some distance to a site, one might well say of the second part – that is the part we’ve just described – that’s more likely to happen at an isolated location. And the reasons you’d say that would be that the felons would know they had a vehicle; they had the rocks in the boot of the vehicle; they would need to go to an isolated place and they would be better able to escape detection as they went through the process of parking the vehicle, opening the boot, taking the rocks out, moving to the freeway and dropping them. So for that type of criminal the isolated location is attractive, I take it; that it would be consistent with what you said. A. Of course it is, but I’m – yet again, these are commonsensical matters I would have thought. … Q. My submission to you which I’d like you to agree with is that it’s a commonsense matter about which Mr Sheldon or his Honour or myself could form a view as to the correctness or otherwise of what I’ve just put too you. A. Well, I can agree with you, and that’s why the risk matrix as I reviewed perplexed me so much.”

The cross-examiner put to Mr Jamieson that, by comparison with that situation, a child showing off to friends on the way home from school could just as easily toss an object over the side of the bridge: “Q. If the child had a box in its hand, a lunchbox or any form of heavy object – a cricket ball that he’d taken from school – there’s your object, and there’s your numbers. A. Of course you’re right. Q. Mr Jamieson, there are your numbers to lead the author of this document to say, ‘Let’s lift the profile – let’s lift in priority places that are near schools.’ Would there not be rational ground for saying, ‘We’ve got to take this into account in determining where we give the weight’? A. I’d agree with you so far as to say yes, schools should be in the equation in a matrix, but to give them a score of 10 to me gives them an unfair weight for potential exposure to risk. Q. You see, when it comes to your expressing that view, ‘I think isolated is more risky than populated’, if I can put it in those categories, your own view about it, but not an engineer’s view. A. Well, both decisions – sorry, both points of view could be considered as non-enduring view.”

In the course of cross-examination, Mr Onggo had put to him the propositions that his evaluation matrix had proceeded erroneously by failing to recognise that the Glenlee Bridge had a history of incidents, and by failing to categorise it as a bridge with pedestrian access. In substance he accepted the first of these propositions, but not the second – see para [413]-[414] above. He also had put to him that the evaluation matrix itself was flawed because it failed to recognise that a bridge with very little traffic on it and away from the public view would be the very place that people minded to drop rocks onto vehicles on freeways would choose, rather than somewhere that was near a public facility like a school or hotel. He did not accept that proposition.

Counsel addressed on those topics. The judge made findings, in para [60], (19), and (20) of his reasons (set out at para [189] above) concerning the inadequacy of the methodology.

In my view, the additional particular of negligence was litigated at the trial. The Respondent should be permitted to amend its Statement of Claim to add that particular of negligence, and to rely on it in the appeal. The Judge’s Findings about the Evaluation Process

The judge said, at [46], that the evaluation matrix: “… gave the Glenlee Bridge a priority rating of 42 out of 84 bridges for screening as at the last modification on 25 May 1998. It may be interpolated, on my calculations, that if the Glenlee Bridge (as suggested by the plaintiff) had been given appropriate scores for the two criteria [of no pedestrian access and no history of prior incidents] rather than zero then its priority rating would have been raised to at least position 10 on the evaluation matrix.”

He continued, at [46]: “It would seem, perhaps curiously in light of Mr Dunlop’s earlier view that the highest risks occurred with bridges over high-speed, restricted access roads on the outer urban freeways – very descriptive of the Glenlee Bridge – that those bridges with pedestrian access near schools, hotels or clubs, other pedestrian generators and with exposure to buildings rated a higher priority for screening on Mr Onggo’s evaluation matrix.” That is a comment, not a finding of negligence.

It was only after making the factual findings in [60] that the judge turned to any comparison of expert evidence in the case. He accepted Mr Jamieson’s criticism of the risk analysis and priority rating exercise “as not being based on any formal risk engineering principles but more on a subjective process by the rater.” His only reference to the RTA’s expert engineering evidence was (at [64]): “Mr Fishburn presented a very detailed and comprehensive report dated 1 May 2007 in which he attended to financial restrictions placed on the defendant in maintaining and upgrading the F5, the system in place prior to August 1998 for the construction of screens over the F5, the priority system and criteria used, any undue delay in screening and the defendant’s reaction to the dropping of objects from the Glenlee Bridge on 10 August 1998. Mr Fishburn reasoned that the defendant’s risk management technique ‘was a very good method to use’ and that generally its response to the problem was reasonable for a prudent road authority to take. It will be apparent from my findings that I do not accept Mr Fishburn’s views.” Decision About Negligence in the Evaluation Process

I do not accept that it has been shown that it was more likely than not that the evaluation matrix “was flawed” . Mr Jamieson’s evidence did not explain how such views as he expressed were based upon any established body of knowledge, or on any experience that he had. He ultimately accepted that his view that isolated locations were more risky than populated, and the opposing view that a higher risk should be attributed to areas near schools, could each be open.

I do not accept that it has been established that there was any failure to proceed as a reasonable authority would have proceeded by not treating the Glenlee Bridge as one that had pedestrian access. There was no expert evidence to that effect. The fact that a bridge carries a road over a freeway will always bring about some risk that an object might be dropped from the bridge. That a bridge is included in the evaluation matrix at all recognises that that risk exists. What Mr Onggo’s methodology did, by allocating points to the fact that there was a walkway or footpath for pedestrians, was to evaluate the extra risk that would arise from pedestrians crossing the bridge in the course of their ordinary activities, as they would have a greater tendency to do if the bridge were solely a pedestrian bridge, or a bridge that made special provision for pedestrians to cross it by having a pedestrian walkway on it.

Mr Onggo was wrong in assigning a score of zero to the Glenlee Bridge for “history of incidents/graffiti” . At least a significant part of the history of incidents at that bridge was already contained in the RTA’s files, as the evidence tendered in this case demonstrated. No satisfactory explanation has been given of how it came about that the data that was relevant to the exercise Mr Onggo was undertaking was not collected from the RTA’s own files.

Mr Onggo had embarked upon a task of applying the methodology he and other RTA officers had worked out. This is not a case where negligence is alleged for failure to do something that the RTA could have done, but did not embark on the doing of. Mr Onggo was carrying out this task for the specific purpose of deciding the order in which bridges should be screened, because he and his superiors appreciated that funds were not available to screen all of them at much the same time, that there was a risk of injury or death to motorists, and that the bridges differed in the extent to which they manifested that risk. While the purpose of his task was to provide a rational means of allocating scarce funds, there was no question of competing priorities being involved in the actual carrying out of that task – it was a tedious, but still comparatively simple, matter of data collection and categorisation, followed by arithmetical calculation. It was a task that was clearly within the operational area of the RTA’s activities. For the RTA to fail to take into account, in circumstances that it did not try to explain or justify, data that it already had in its own files amounts in my view to a failure to take reasonable care to protect motorists from the very risk that was the reason why Mr Onggo was doing the task.

As mentioned at para [414] above, if the rating attributed to the Glenlee Bridge were adjusted to take account of there having been previous incidents, and no other change to the evaluation matrix was made, the Glenlee Bridge would have advanced to number 22 in priority.

Mr Maconachie submits that notionally placing the Glenlee Bridge at number 22 in priority through such a process of reasoning would be excessively simplistic. Rather, he submits, once the error in application of Mr Onggo’s methodology was shown, it would be necessary to rework the entire evaluation matrix, to make sure that other bridges were not also advanced in priority in the table. That, he submits, has not been done.

I recognise that such a submission has some theoretical appeal. Before accepting it, one would need to consider how it accords with shifting onuses of adducing evidence. I will not pause to do so. I take that course because, even if the Glenlee Bridge was advanced in priority to number 22 in the evaluation matrix of 25 May 1998, there would still be nine, or possibly ten, federally funded bridges, on the F3 and F5, that were above it in priority.

I express this uncertainty about the number of bridges that would be above it because on a full version of Mr Onggo’s evaluation matrix of 25 May 1998 something identified as “Rail Overbridge, Wahroonga” that is said to be over the F3 is accorded a very high priority for screening, but that bridge does not appear on a list of the current status of bridge screening on National Highways according to the evaluation matrix (which included completed bridges) dated 23 September 1999.

Even leaving that rail bridge out of account, if the Glenlee Bridge were advanced to number 22 in priority there would still be seven bridges on the F3, and two on the F5, above it in priority. The evaluation matrix as at 25 May 1998 does not include any of the bridges that had actually been screened on the F5 or the F3 by 30 June 1998. Thus, even if the error in application of Mr Onggo’s matrix that arose from his not taking into account the history of incidents at the Glenlee Bridge were to have been corrected, it would still not have resulted in the Glenlee Bridge having been screened before 23 August 1998. Mr Onggo’s mistake is an act of negligence that has not caused Mr Evans’ death. ISSUE 3 – CAUSATION OF DAMAGE

In light of my findings about breach of duty, there is no occasion to discuss this issue. Orders

I propose the following orders: 1. Appeal allowed 2. Set aside the judgment and orders in the court below 3. In lieu thereof, order that the proceedings be dismissed with costs 4. Order Respondent to pay Appellant’s costs of the appeal.

SACKVILLE AJA : I have had the privilege of reading Campbell JA’s comprehensive reasons for judgment. I agree with the conclusions reached by his Honour and, in general, with his Honour’s reasoning. However, I wish to add the following comments.

The respondent pleaded that the appellant (“ RTA ”) had care and control of and was responsible for the maintenance of the Hume Highway and of the Glenlee Bridge. The respondent also alleged that RTA owed a duty to take reasonable care for the safety of road users. The RTA in its defence admitted the substance of these allegations, but denied that the scope of the duty extended to preventing the commission of criminal offences by third parties.

I agree with Campbell JA that, for the reasons his Honour gives (at [141]-[142]), the RTA’s duty did not exclude taking reasonable care to protect motorists using the F5 Freeway from the criminal actions of third parties.

I also agree with Campbell JA (at [165]) that, insofar as the primary Judge found that the RTA breached its duty of care to Mr Evans by failing to install protective screening at the time the Glenlee Bridge was constructed, his Honour was in error in making that finding.

The primary Judge found (Judgment, at [109]) that the RTA’s failure over many years before October 1998 to fence or screen the Glenlee Bridge “ represente[d] a failure to take reasonable care ”. Campbell JA has explained (at [192]) that the primary Judge’s finding rested on a flawed assessment of the extent of the risk from objects thrown or falling from overpasses in the Campbelltown area. In addition, Campbell JA concludes that the primary Judge erred in failing to apply the process of reasoning articulated by Mason J in Council of the Shire of Wyong v Shirt [1980] HCA 12; 146 CLR 40, at 47-48. (The relevant passage from the judgment of Mason J is reproduced in Campbell JA’s judgment at [178].)

I agree with Campbell JA that the primary Judge’s analysis of the matters to be taken into account in determining whether RTA breached its duty of care was flawed. However, I would prefer to say that the flaw in the primary Judge’s analysis was his failure to apply s 5B of the Civil Liability Act 2002 (NSW) (“ Civil Liability Act ”), rather than a failure to apply the process of reasoning set out in Wyong Shire Council v Shirt . (Section 5B of the Civil Liability Act is set out in Campbell JA’s reasons (at [172])).

It is true that, as Ipp JA pointed out in Waverley Council v Ferreira [2005] NSWCA 418, Aust Torts Reports 81-818, at [45], the matters set out in s 5B(2) of the Civil Liability Act are, in substance, a reiteration of Mason J’s oft cited remarks in Wyong Shire Council v Shirt , at 47-48; see, too, Review of the Law of Negligence: Final Report (September 2002) (“ Ipp Report ”), pars 7.5–7.18. It is also true that s 5B operates against the backdrop of the law of negligence. Thus s 5B does not itself impose an obligation on a person to exercise reasonable care, but sets out in subsection (1) requirements that must be satisfied before a person can be found to be “ negligent in failing to take precautions against a risk of harm ”: D Villa, Annotated Civil Liability Act (2002) (NSW) (2004) at [1A.5B.050]; Penrith Rugby Club Ltd v Elliott [2009] NSWCA 247, at [22]. Section 5B(2) specifies the matters that the Court is to consider (among other relevant things) in determining whether a reasonable person would have taken precautions against a risk of harm. The calculus provided by s 5B(2) requires the matters specified in pars (a) and (b) (“ the probability that harm would occur if care were not taken ” and “ the likely seriousness of the harm ”) to be weighed against the matters specified in pars (c) and (d) (“ the burden of taking precautions to avoid the harm ” and the “ social use of the risk-creating activity ”): Ipp Report , at [7.9].

While s 5B of the Civil Liability Act incorporates concepts derived from the common law, the primary Judge was required by statute to satisfy himself that each of the matters specified in s 5B(1) was satisfied before he could find that the RTA had been negligent in failing to take precautions against the risk of objects being thrown or falling from the Glenlee Bridge onto vehicles travelling along the F5 Freeway. In order for the primary Judge to find that the criterion stated in s 5B(1)(c) was satisfied (“ in the circumstances, a reasonable person in [the RTA’s] position would have taken these precautions ”), his Honour had to weigh the competing considerations identified in s 5B(2).

The primary Judge did not refer to s 5B(2) of the Civil Liability Act , presumably because the provision did not play a prominent part in the argument at trial. While the absence of any reference to s 5B in the primary judgment is not necessarily determinative of whether his Honour undertook the task mandated by the legislation, in my opinion he did not do so. In particular, in determining that the RTA had breached its duty of care, the primary Judge did not weigh the probability that the harm would occur if care were not taken against the burden of taking precautions to avoid the risk of harm. Section 5C(a) of the Civil Liability Act makes it clear that the burden of taking precautions to avoid the risk of harm includes: “the burden of taking precautions to avoid similar risks of harm for which the person may be responsible.” These provisions required the primary Judge to undertake a careful analysis of the RTA’s responsibilities to avoid other risks of harm to road users and to take account of the competing demands on the limited (if substantial) resources of the RTA.

Sections 5B and 5C of the Civil Liability Act may not require a Court to take any different approach to determining whether a highway authority has been negligent than is required under common law principles. Thus in Brodie v Singleton Shire Council [2001] HCA 29; 206 CLR 512, a case from New South Wales decided before the passage of the Civil Liability Act , Gaudron, McHugh and Gummow JJ pointed out (at 579 [158]) that the rejection of the so-called “ immunity ” for highway authorities did not mean that bodies such as the RTA were obliged in all cases to exercise their powers to repair roads or to ensure that they were kept in repair. Their Honours observed (at 580-581 [162]) that: “The formulation of the duty of care includes consideration of competing or conflicting responsibilities of the authority. In the circumstances of a given case, it may be shown that it was reasonable for an authority to deal in a particular priority with repairs in various locations. The resources available to a road authority, including the availability of materiel and skilled labour, may dictate the pace at which repairs may be made and affect the order of priority in which they are to be made. It may be reasonable in the circumstances not to perform repairs at a certain site until a certain date, or to perform them after more pressing dangers are first addressed.” This statement of principle would apply equally to the amelioration of risks such as those presented to road users by objects being thrown or falling from overpasses.

While there may be little or no difference in substance between the statutory directions contained in ss 5B and 5C of the Civil Liability Act and common law principles, it is to be remembered that the very point of the statutory provisions is to: “encourage judges to address their minds more directly to the issue of whether it would be reasonable to require precautions to be taken against a particular risk.” Ipp Report at [7.17]. In my view, the legislation must be the starting point for both trial judges and appellate courts dealing with cases such as the present.

Campbell JA has examined in detail the competing demands placed on the resources available to the RTA to minimise the risk of injury or death to road users posed by the propensity of some people to throw objects from overpasses into the path of oncoming vehicles on freeways and other roads. There is no doubt that the risk of injury or death from activities of this kind was not only foreseeable but actually foreseen some time before the tragic incident that caused Mr Evan’s death. It was also foreseen that the Glenlee Bridge, along with many other overpasses, presented a significant risk of injury to road users requiring attention from the RTA. The chances of injury or death occurring at any particular location, if preventative measures were not taken, were very low. But if harm did result from objects being thrown or falling onto the carriageway from an overpass, the harm was likely to be very serious.

Subject to the effect of ss 42 and 43A of the Civil Liability Act , I see no compelling reason in the present case, whether deriving from distinctions sometimes drawn between policy and operational matters or otherwise, for the Court to shy away from undertaking the assessment contemplated by s 5B, in particular weighting up the matters identified in s 5B(2). This requires the Court to determine whether a reasonable person in the position of the RTA would have fenced the Glenlee Bridge notwithstanding competing claims on its resources to address similar risks of serious injury elsewhere. In making this determination, the Court needs to consider whether the RTA’s ordering of priorities was a departure from standards to be expected from a reasonable person in the RTA’s position. The Court should also take account of the opportunities reasonably available to the RTA to gain additional funding from the Commonwealth or other sources for the purpose of addressing particularly acute risks of which it was aware or should have been aware.

Campbell JA’s analysis of the facts in this case seems to me to show that the RTA appreciated in a reasonably timely fashion the nature and magnitude of the risk to road users, including the risk to users of the F5 from the Glenlee Bridge, by reason of objects being thrown or falling from overpasses. The RTA’s response to the risk, having regard to the burden of taking precautions to alleviate the risk of harm to all road users from similar sources, was not shown to be unreasonable. The RTA adopted a rational and apparently systematic (although not perfect) approach to assessing priorities for the erection of protective fencing on the basis of the magnitude of risk. It acted on that assessment within the limits of available resources. The evidence does not demonstrate that a reasonable authority in the position of the RTA would have sought additional funding from the Commonwealth or that, if it did, such funding would have been made available for fencing the Glenlee Bridge before the incident that led to Mr Evans’ death.

I do not think it is necessary to consider whether Mr Onggo’s apparent error in assigning a score of zero to the Glenlee Bridge in relation to its “ history of incidents/graffiti ” would have established, of itself, that the RTA was negligent. As Campbell JA points out (at [434]), even if Mr Onggo had assigned the correct score for Glenlee Bridge in preparing his matrix, it still would not have prevented the incident which caused Mr Evan’s death.

I agree with the orders proposed by Campbell JA.