Transport for NSW v Hunt Leather Pty Ltd; Hunt Leather Pty Ltd v Transport for NSW

1. [1]

   THE COURT: This appeal and cross-appeal, which were heard concurrently, are brought from decisions determining liability and damages in respect of the four lead plaintiffs in representative proceedings pursuant to Part 10 of the Civil Procedure Act 2005 (NSW) arising out of the construction of the Sydney Light Rail. The appellant, Transport for NSW, a corporation constituted by s 3C of the Transport Administration Act 1988 (NSW), which statute also refers to it as “TfNSW”, was the sole defendant. TfNSW planned the construction of the Sydney Light Rail, but did not itself undertake any work.

2. [2]

   The Sydney Light Rail is a light rail system between Circular Quay and Moore Park, from which separate branches continue to Randwick and Kingsford. Most of the route is along public roads, including George St in the Sydney CBD and Anzac Parade in Kensington. (A map of the route is found at [20] of the main judgment: Hunt Leather Pty Ltd v Transport for NSW [2023] NSWSC 840.) Hunt Leather Pty Ltd operated two retail stores along George St, one inside the Queen Victoria Building and the other in leased premises facing George St in the Strand Arcade. Ancio Investments Pty Ltd was the trustee of a trading trust which operated various restaurant businesses from leased premises on Anzac Parade in Kensington.

3. [3]

   The group members were defined by reference to causes of action in private nuisance and public nuisance. The action for public nuisance failed and there is no appeal from that part of the judgment. The second and fourth plaintiffs, Ms Sophie Hunt and Mr Nicholas Zisti, were not occupiers of land and were only plaintiffs for the purposes of pursuing claims for public nuisance, but their companies, Hunt Leather Pty Ltd and Ancio Investments Pty Ltd, obtained judgments in their favour based on private nuisance for very substantial reasons for judgment occupying 1140 paragraphs over 292 pages following a trial lasting almost six weeks in November and December 2022. Hunt Leather’s success was based solely on its Strand Arcade premises, and there is no appeal from the finding by the primary judge at [887] and [911] that there was no substantial interference to its premises in the Queen Victoria Building.

4. [4]

   By a separate judgment following a further hearing in December 2023, the primary judge dismissed Hunt Leather’s and Ancio’s claim that the damages should include a 40% commission they were liable to pay to a litigation funder: Hunt Leather Pty Ltd v Transport for NSW (No 4) [2024] NSWSC 140. Those companies have cross-appealed from that decision. Their cross-appeal is put on the basis that the commission should have been incorporated into the damages ordered, in effect, a form of grossing up; it was not said that the commission was recoverable pursuant to any costs order (cf Anderson v Canaccord Genuity Financial Ltd (No 2) [2024] NSWCA 161 at [106]-[158]). The result was judgments in their favour against TfNSW in the amounts of $3,693,164 and $317,773 respectively. The cross-appeal was expanded to include a separate challenge to one aspect of a subsequent costs order, made on 24 June 2024, by which TfNSW was ordered to pay 65% of the plaintiffs’ costs in respect of the claims for private nuisance: Hunt Leather Pty Ltd v Transport for NSW (No 5) [2024] NSWSC 776. Because that order was made only a week prior to the hearing of the appeal, it was developed by submissions supplied in accordance with a timetable imposed at the conclusion of the hearing. The last submissions were supplied on 26 August 2024.

5. [5]

   Although there has been a final determination of the claims by Hunt Leather and Ancio against TfNSW, the claims of other group members have not been determined. TfNSW’s appeal was brought purportedly as of right. At the time the notice of appeal was filed, the appeal was from an order made on 3 August 2023 that there be “Judgment in favour of the first and third plaintiffs for sums to be quantified”. The order, which determined liability but not quantum, was plainly interlocutory for the purposes of s 101(2)(e) of the Supreme Court Act 1970 (NSW): see Computer Edge Pty Ltd v Apple Computer Inc [1984] HCA 47; 54 ALR 767 at 768 and Narellan Franchise Pty Ltd v RBME Pty Ltd [2023] NSWCA 139 at [16]-[17]. However, the judgments have now been quantified and as between Hunt Leather, Ancio and TfNSW all matters in dispute have been determined. Consistently with the analysis in Queensland Bulk Water Supply Authority t/as Seqwater v Rodriguez & Sons Pty Ltd [2021] NSWCA 206; 393 ALR 162 at [14]-[20], the appeal from those judgments is as of right, and thus the notice of appeal correctly anticipated the position which obtained by the time the appeal was heard.

Background

1. [6]

   The following paragraphs are uncontroversial.

2. [7]

   TfNSW did not own the roads along which the Sydney Light Rail was built. Nor did it perform any construction work. As will be seen, those matters do not exclude TfNSW from being liable in private nuisance to occupiers affected by the construction.

3. [8]

   Construction would inevitably occur along the entirety of the route, so as to prepare foundations, lay tracks, install electrical infrastructure and signalling, and so on. The construction occupied much of the land dedicated as a road including the footpath, including by the erection of fencing and hoarding. Footpaths are ordinarily part of the land dedicated as a road; see for example Brodie v Singleton Shire Council (2001) 206 CLR 512; [2001] HCA 29 at [50]-[51], and what matters, as Dixon J observed in Buckle v Bayswater Road Board (1936) 57 CLR 259 at 286; [1936] HCA 65, is that there is a public right of passage. The occupation of the road (including the footpath) was capable of constituting a public nuisance, but it was also capable of constituting a private nuisance, insofar as it interfered with the enjoyment of neighbouring land, notably, by reducing passing traffic.

4. [9]

   Pursuant to agreements with City of Sydney Council, Centennial Park and Moore Park Trust and Randwick City Council, TfNSW was granted a licence to enter and occupy the land (principally, the road surface and footpaths within the road reserve), which were owned by those bodies, and was empowered to grant licences to anyone else to do so for the purposes of constructing the Sydney Light Rail. Pursuant to cl 12.1 of the Project Deed, TfNSW granted OpCo a “non-exclusive Construction Site Licence” permitting it and its contractors to use and occupy the land.

5. [10]

   The main planning of the Sydney Light Rail commenced in 2012, and TfNSW determined that it would be built and operated through a “public and private partnership”. The label is at best uninformative and at worst misleading – as is perhaps most obvious from the fact that the sole defendant was TfNSW, with nothing to suggest any joint liability with or contribution from some private “partner”.

6. [11]

   What in fact happened in point of law was based on a Project Deed and a Design and Construct Contract. The parties to the Project Deed were TfNSW and a partnership of three ALTRAC companies each of which was a trustee of trusts known, respectively, as the ALTRAC Light Rail Trust 1, 2 and 3. The primary judge explained at [23] that he would refer to the partnership, interchangeably, as “ALTRAC” and “OpCo”, and these reasons will use the same nomenclature. The parties to the Design and Construct Contract were that partnership (ie ALTRAC or OpCo) and each of Alston Transport Australia Pty Ltd and Acciona Infrastructure Australia Pty Ltd, which two companies were defined as the “D&C Contractor”. None of those details greatly matters for the purposes of this litigation. The essence of the bargain was that a Spanish infrastructure company and a French rolling stock manufacturer would construct the light rail system which would then be operated privately for a period of 15 years. The Project Deed provided in cl 1.6 that the “Project Agreements” (which included the Project Deed, and the Design and Construct Contract) did not incorporate any of the principles or guidance published by either the Commonwealth under its National Public Private Partnership Policy and Guidelines or the State under its NSW Public Private Partnerships Guidelines (2012).

7. [12]

   The D&C Contractor was not to occupy the entirety of the site throughout the construction period. Instead, construction would occur in stages. It is obvious that staging the construction reduced the ability of the D&C Contractor to be flexible in a range of matters (including the timing of each work and the storage of machines and materials along the route) and may well have increased the cost and time taken for construction. It is also obvious that to the extent that disruption to motorists, pedestrians and, most relevantly, the occupiers of neighbouring premises along the route, was caused by the construction activities and the erection of fencing and hoarding preventing ordinary access to the road and footpath, that disruption would be expected to be reduced in any particular area by the construction being staged.

8. [13]

   Those considerations were reflected in a development agreement between TfNSW and City of Sydney Council dated 18 December 2013, which recited that the parties sought to minimise the impacts of construction on the community, businesses and residences in the local government area, including by TfNSW’s agreement that staging of the works would be developed to demonstrate a balance between efficient construction methods and to minimise impacts on business operations and resident amenities.

9. [14]

   Attached to the Project Deed as Schedule A10 was an “Initial Delivery Program” or “IDP”. The Schedule was not included in the appeal books. It appears to have been a lengthy print out obtained from an .XER file, which is used in project management software. As will be seen below, on the third day of the hearing the Court was given a print out of the “amended IDP” which is critical to the approach taken by the primary judge. The print out has 56 pages, and each page has over 100 rows of entries. Many of the rows contain headings, but the large majority describe specific activities in a particular place and timeframe along the route of the Sydney Light Rail.

10. [15]

    The IDP reflected a division of the construction route into stages, which were called “fee zones”. Hunt Leather’s Strand Arcade store was in fee zone 5 and Ancio’s restaurants were in fee zone 29. The IDP identified a wide range of activities which would be undertaken in each fee zone and the order in which they would be undertaken, and stated the time during which the D&C Contractor would occupy each fee zone. There were thousands of activities in the IDP.

11. [16]

    In May 2015, TfNSW issued a media release to the effect that the Sydney Light Rail construction would begin at Moore Park in September 2015, with construction on George Street beginning in October 2015. The document advised that the light rail route had been divided into 31 individual zones to minimise, as much as possible, the impacts of the construction on discrete areas, and gave a list and map of the 31 fee zones, noting that construction would take place in fee zone 5 from 23 October 2015 until 12 August 2016, and in fee zone 29 from 19 February 2016 to 6 September 2016. The document stated, under a bold heading: “Major construction begins September 2015 – Major construction expected to complete mid 2018”:

12. [17]

    The times published in the media release approximately reflect the programming incorporated within the IDP. There are differences, but they are minor. The material supplied to this Court does not appear to enable an understanding of why, to take one example, the document recorded the commencement of occupation of fee zone 5 to be 23 October 2015 while the corresponding date in the IDP was 15 October 2015.

13. [18]

    As matters turned out, the times stated in the media release for the completion of the Sydney Light Rail were understated by around a year, but the periods stated for the occupation of each fee zone proved to be extremely optimistic. The statement that ALTRAC could not alter the duration within any individual zone was wrong.

14. [19]

    The Project Deed made it quite clear that the dates in the IDP were not contractual promises. Clause 17.2 required OpCo to use its best endeavours to achieve completion by 16 March 2019, and to “diligently progress the Delivery Activities”. Clause 17.3 required OpCo periodically to update the Delivery Program, which started with the IDP in Schedule A10 attached to the deed, and which said explicitly in cl 17.3(c) that “Nothing in the Delivery Program will bind TfNSW or otherwise affect the time for the performance of TfNSW’s obligations under this deed”.

15. [20]

    The reasons for the substantial delay between actual construction compared to the times identified in the IDP and the press release were not fully explained. The primary judge found that a contributing factor was the discovery of utilities along the route.

16. [21]

    At [111]-[202] the primary judge described the way in which the Sydney Light Rail was planned. Prominent in that process was the so-called “utilities risk”. The primary judge said of this:

17. [22]

    His Honour said:

18. [23]

    At a high level, it was common ground that in order to establish actionable private nuisance sounding in damages, it is necessary but not sufficient for a plaintiff to establish a substantial interference with the enjoyment of land. It was also common ground that there were some substantial interferences with enjoyment which were not actionable. The primary judge’s reasoning attended, appropriately, to these nuances. His Honour accepted that on any view the occupiers had no basis to complain of some measure of disruption caused by the construction of the Sydney Light Rail, but formed a view that there came a time after which TfNSW was liable for damage caused by the ongoing disruption.

19. [24]

    The primary judge did not use the IDP itself in determining the extent to which there was actionable nuisance. Instead, his Honour deployed an “amended IDP” which had been propounded by the plaintiffs and determined that insofar as occupation of the fee zones had exceeded the times in the amended IDP, it was unreasonable and therefore actionable. According to the amended IDP, work in fee zone 5 should have started on 15 October 2015 and finished on 1 November 2016. It did not finish until 3 December 2017. In fee zone 29, the work should have started on 12 December 2016 and finished on 3 September 2017. It did not finish until 28 February 2019.

20. [25]

    The author of the amended IDP was Mr Griffith. The primary judge recorded that Mr Griffith was a “planning and programming” expert, who was not originally called to provide an opinion about the quality of the defendant’s utilities investigations or management, but nonetheless ended up doing so (at [486]). There was a large issue at trial concerning Mr Griffith’s opinions, much of which is incorporated below:

21. [26]

    The primary judge explained the preparation by Mr Griffith of the amended IDP at [495]:

22. [27]

    We shall return to the details underlying Mr Griffith’s calculations and the criticisms of them. For present purposes, it suffices to note that the effect of the amended IDP was to extend the allowance for completion of work in the fee zones by an additional three or four months in fee zones 5 and 29, as reflected in the following table reproduced at [935]:

23. [28]

    The primary judge’s reasoning based on the amended IDP was explained:

24. [29]

    There was an error in the dates applicable for Ancio’s claim in fee zone 29, which was corrected by his Honour in a subsequent judgment under the slip rule (the application was not opposed by TfNSW): Hunt Leather Pty Ltd v Transport for NSW (No 3) [2023] NSWSC 1598 at [12]. For reasons which will become apparent, it is convenient to reproduce [9]-[12] of that judgment:

25. [30]

    The finding that the commencement of the interference to Mr Zisti’s restaurant was 7 May 2016 was at [888]-[909], where his Honour concluded that the interference continued between 7 May 2016 and 28 February 2019.

26. [31]

    It is apparent from the above how his Honour proceeded. Because the only integer that mattered for the purposes of calculating damages was the point at which the interference became, on the approach taken by his Honour, no longer reasonable, his Honour only corrected the date of 3 September 2017, replacing it with 27 January 2017. However, as the passage at [9]-[10] of the later judgment makes plain, the starting date for the actual occupation (which on the approach taken by his Honour was not tortious) was 7 May 2016, not 12 December 2016, despite that date not being corrected under the slip rule. It is also clear that the important date of 27 January 2017 was determined by using the methodology in the main judgment, which admitted of a period under the corrected IDP of non-tortious occupation of the period from 12 December 2016 until 3 September 2017 (which is 265 days), and shifted that period so that it commenced on 7 May 2016 (7 May 2016 until 27 January 2017 is also 265 days).

27. [32]

    How then did Mr Griffith calculate his amended IDP? Attachment 6 to his report was not included in the appeal books, but at the Court’s request was provided on the last day of the hearing of the appeal. In that document he explained, in paragraph 625, that on his calculations the “IDP Occupation” of fee zone 5 of 281 days had, when corrected, increased to 349 days. In the case of fee zone 29, IDP Occupation of 211 days had increased to 251 days. In paragraph 626 he explained why. In both cases, 19 days were attributable to “Increased Earthworks Contingency” and “Inclement Weather Allowance”. In fee zone 5, there were also 36 days for “Early Works Design Links” and 13 days for “Increased Float between Utility Treatments”. In fee zone 29, there were 25 days for “Increased Allowance for Treatment of Unknown Utilities”, 8 days for “Increased Float between Utility Treatments” and a reduction of 12 days for “Start date delay mitigation” (not all of those delays extended the time cumulatively). Each of those categories was explained in paragraphs 628-665 in his report. For example, Mr Griffith was critical of the fact that the IDP did not include any “Float”, which was essentially an allowance for various contingencies (“In my opinion, a contractor should reasonably have included some contingency in the IDP for the higher risk Activities such as Utility Treatments and Civil Works”). Mr Griffith also considered that there should have been a contingency for Unknown Utilities (“A Contractor acting reasonably would have made an allowance for [the delay in completing Unknown Utilities] in the IDP”).

28. [33]

    Mr Griffith produced a table in section 2.3 of his report headed “Corrected IDP” for the purpose of correcting the IDP so as “to represent a reasonable baseline against which to assess delay to Fee Zone Completion”, based on the links between the design submissions and construction, the lack of general contingency, lack of contingency for Unknown Utilities, lack of float between Utility Treatment Activities in successive fee zones, and lack of inclement weather allowance. The result was that the “IDP occupation” for fee zones 5 and 29 increased from 281 and 211 days, respectively, to 349 and 251 days, respectively.

29. [34]

    It is worth explaining two aspects of the details of Mr Griffith’s reasoning. The largest contributor to the increased times is an allowance for “utility treatment activities”, and the most straightforward component of the increase was the incorporation of an allowance for inclement weather.

30. [35]

    Mr Griffith’s treatment of unknown utilities was at paragraphs 633-639 of Attachment 6. In fee zone 5, he stated that there were 67 known utilities, and 55 “unknown utilities”. He said that the unknown utilities “in my opinion would have been allowed for had they been known at the time of Contract”. He used simple linear extrapolation. He noted that 28 workdays were allocated to utility treatment in the IDP, and that 55 was 82% of 67, and so increased 28 by 82% yielding 51 workdays. In fee zone 29, where there were 106 known utilities and 103 unknown utilities and 18 days allocated to utility treatment in the IDP, he noted that 103 was 97% of 106, and increased 18 by 97% yielding 36 workdays.

31. [36]

    Mr Griffith’s treatment of inclement weather allowance was at paragraphs 655-665 of Attachment 6. He simply allowed 4.5% of the workday duration of each activity in the fee zone rounded to the nearest whole day, noting that 4.5% “reflects the typical impact from weather that I have tended to see on projects carried out in the Sydney area”. The overall workday allowances for fee zones 5 and 29 were 208 and 251 days respectively, to which were added 9 and 11 days.

32. [37]

    The details underlying Mr Griffith’s reasoning do not for present purposes matter, save to confirm the basis on which they were derived. It is clear, as TfNSW submitted, without contradiction from the respondents, that the amendments reflected deficiencies, in Mr Griffith’s opinion, based on looking at project documents and programs produced at or around the commencement of the project. As it was put orally:

33. [38]

    That is to say, the amendments were what Mr Griffith regarded as estimates which incorporated appropriate allowances based on what was then known, i.e. prior to commencement of construction. The amended IDP was not intended to be an assessment, based on what in fact occurred each day along the route (including in relation to the discovery and treatment of previously unknown utilities), of the time after which continued work became unreasonable (it was also described, accurately but somewhat cryptically, as an “ex post facto a priori prediction”).

34. [39]

    As the judge correctly recorded at [58]:

35. [40]

    Two important things emerge from an understanding of the way in which the amended IDP was derived.

36. [41]

    The first is that it assumed that there was complete or substantially complete identification of all utilities along the route before construction commenced. That was confirmed in the following exchange:

37. [42]

    There was no evidence about how long it would have taken to obtain such knowledge, or how disruptive doing so would have been to the occupiers of neighbouring premises.

38. [43]

    Indeed, Mr Griffith and Mr McIntyre (an expert called by TfNSW) agreed that “the time it took to deal with the utilities” was always likely to have “had a similar effect on the overall completion of the project as actually occurred”.

39. [44]

    The second is that, at best, Mr Griffith’s amended IDP reflected an opinion of how long it ought reasonably to take to perform all of the activities in the staged construction of the Sydney Light Rail, assuming full knowledge of utilities (including subterranean utilities) in advance. But it is one thing to say in advance that a particular time period is a reasonable time period within which construction should occur, and another thing to say that a builder’s occupation of the roadway and footpaths for any time longer than a particular time period was in fact unreasonable. Mr Griffith’s amended IDP did not purport to delineate the times after which continued occupation of the roadway and footpaths would be unreasonable, whether as a matter of prediction or with the benefit of the delays that actually occurred, and the additional time taken to treat the additional utilities that were discovered in the course of construction.

40. [45]

    It boils down to this. The Initial Delivery Program attached to the Project Deed contained a timeline for the construction of the Sydney Light Rail. It did not of itself have any contractual force. To the extent that it could be regarded as a prediction of the time for construction, it was deficient in obvious respects (such as the absence of any contingency for weather), and there were obligations upon the parties to update it from time to time as construction proceeded. It is true that the approximate timing of the occupancy in each fee zone reflected in the IDP was communicated to the public and to the nearby occupiers who would be affected. Mr Griffith prepared an amended IDP which corrected the deficiencies he perceived, with the result that the occupancy of each fee zone was extended, but was still very substantially less than the actual occupancy that occurred.

41. [46]

    TfNSW was not sued for negligently compiling the IDP. Nor was it sued for misleading and deceptive conduct in issuing the press release which stated the durations for which ALTRAC would occupy each fee zone. Instead, TfNSW was sued for nuisance. The primary judge determined that TfNSW was liable for the time it exceeded the amended IDP in each fee zone. But that timing was based merely on Mr Griffith’s corrections of the methodology in the IDP.

42. [47]

    TfNSW did not adduce evidence of what occurred to delay the completion of the project and why. Nor did the plaintiffs. As will be seen, a significant aspect of the trial was the failure by both sides to establish why the occupation of each fee zone in fact took as long as it did.

43. [48]

    The critical issue in this appeal is the determination that TfNSW was liable for damages in private nuisance insofar as the contractors occupying the fee zones exceeded the times in the amended IDP. Before addressing that in detail, it is as well to mention the basis on which TfNSW was found liable.

44. [49]

    TfNSW was not a landowner. Nor did it do any relevant construction work. Nor was it said to be vicariously liable for the conduct of any contractors. Instead, it was found to be personally liable for creating the state of affairs which led to the nuisance: at [939]-[946]. The essence of the problem was that the disincentives in the Project Deed for delayed occupation of a fee zone were insufficient. Simplifying the position considerably, the total amount of daily fees for occupying fee zones beyond the “Base Fee Zone Occupation Period” was capped at $7.5 million, and even that amount might in certain circumstances be reduced by the $3 million “Base Fee Zone Credit” (the details are given at [224]-[225] of the reasons of the primary judge). (For completeness, it may be noted that the “Base Fee Zone Occupation Period” was calculated by reference to Schedule B9, as opposed to the IDP in Schedule A10.)

45. [50]

    The primary judge concluded:

46. [51]

    That approach accords with authority. It is sufficient if the defendant “created, adopted or continued” the nuisance: Hargrave v Goldman (1963) 110 CLR 40 at 62; [1963] HCA 56. Torette House Pty Ltd v Berkman (1939) 39 SR (NSW) 156 arose from a directed verdict for the defendant in an action for negligence and nuisance. A contracted plumber had gone onto the premises and mistakenly turned on a stopcock which caused water to flow into the plaintiff’s neighbouring land. It was admitted that the defendant did not create it, and was not shown to have known of its existence or allowed it to continue: at 160. However, Jordan CJ was at pains to emphasise that “[a] person who procures the doing of an act is liable for its actual consequences and for anything necessarily involved in its being done whomsoever he may have procured to do it. He is liable for the acts of any agent of his acting within the scope of his employment” at 170; see also at 165. The High Court dismissed an appeal: (1940) 62 CLR 637; [1940] HCA 1, with each of Latham CJ, Starke and Dixon JJ confirming there could be no liability for nuisance from the unanticipated and faultless act of the owner: at 646, 651, 659. This is different from the present facts, and indeed Latham CJ noted at 646 that “the defendant did not employ the plumber to do any act of which the nuisance was the necessary or a natural consequence”.

47. [52]

    As the Supreme Court of the United Kingdom recently explained, this form of liability is sometimes described as “continuing a nuisance”, and is to be distinguished from a continuing nuisance; the latter term denoting a continuing state of affairs: The Manchester Ship Canal Company Ltd v United Utilities Water Ltd (No 2) [2024] UKSC 22 at [6]-[13]. In order to avoid the verbal similarity between those quite different notions, it is clearer to refer to a person who created, adopted or continued a nuisance.

48. [53]

    There was no challenge to the reasoning that TfNSW created or procured the conduct which constituted the alleged nuisance.

49. [54]

    Reference has already been made to the “utilities risk”: the risk of delay and additional work arising from the need to treat utilities lying beneath the route of the proposed light rail as well as the risk of discovery of previously unknown utilities under the route which would need to be dealt with. The primary judge referred to there being two levels at which the utilities risk arose: in part because of the absence of agreements with utility providers such as Ausgrid, and in part because of the problems which might arise from the discovery of unknown utilities along the route.

50. [55]

    His Honour said:

51. [56]

    The Project Deed addressed the utilities risk in the Design and Construct contract, as summarised by the primary judge at [216]-[228], including, importantly, cl 12.3. The clause is long and reproduced in its entirety by the primary judge at [224]. His Honour summarised its effect at [225]:

52. [57]

    How could TfNSW have obtained better knowledge of unknown utilities? TfNSW’s basic submission was simple: “to achieve greater certainty about what utilities were under George Street, for example, in effect cannot be done other than by digging up George Street and having a look”.

53. [58]

    The primary judge recorded at [168] that, by 10 March 2014, TfNSW had undertaken surveys along the Sydney Light Rail route comprising 105,032 electronic files, all to be issued to the proponents for tender on secure hard drives, and at [178] that on 12 June 2014, TfNSW published further survey data comprising 1,855 electronic files issued to the proponents for tender. This work was done, at least substantially, by a trenching contractor, pursuant to a contract in September 2013 which required the contractor to: (a) carry out a desktop study of the “Dial Before You Dig” documentation; (b) coordinate with others to determine the exact location and depth of trenches required; (c) conduct field verification of existing infrastructure and services (known and unknown) and propose trench locations; (d) facilitate interface agreements with technical advisers and utility providers to optimise access to utilities; and (e) obtain all of the necessary approvals, including from responsible authorities, and comply with their conditions.

54. [59]

    Mr Griffith assumed full knowledge of utilities in his amended IDP. He confirmed in cross-examination that he did not say what additional surveying should have been undertaken:

55. [60]

    Further, his view that insufficient surveying had been undertaken was a product merely of the fact that so many unknown utilities were encountered during construction:

56. [61]

    The principal debate between experts at trial on the extent to which investigations of unknown utilities could or should have been conducted before construction commenced was between two engineers, Messrs Sampson and Szmalko. The primary judge summarised Mr Sampson’s evidence at length, at [428]-[472], with general approval. His Honour said that much of what he said “makes sense” and that on whether TfNSW’s conduct was unreasonable for the purposes of s 43A of the Civil Liability Act, which was addressed in detail in his second report, he preferred the evidence of Mr Sampson: at [466]. It is unnecessary to summarise Mr Sampson’s evidence on s 43A unreasonableness, but the following matters emerge from his first report.

57. [62]

    At [431] the primary judge noted that Mr Sampson had said that in 2013 and 2014, it was only possible to have utility assets at 100% quality level A if the entire length of the asset had been sighted and surveyed, which did not happen in the case of the Sydney Light Rail. He said that Stage 1 of the Gold Coast Light Rail had seen the risk of unknown utilities being left to the contractor, which might also have been the original intention in planning the Sydney Light Rail.

58. [63]

    The primary judge summarised at [193] a “Consolidated Clarification: Returnable 8 – Options” paper dated 1 September 2014 as follows:

59. [64]

    The references to “level B” and “level A” were to the following states of information (reproduced by the primary judge at [335] from Mr Szmalko’s report):

60. [65]

    Mr Sampson said that the level of investigation which had been undertaken exceeded Mr Szmalko’s generalised quality level estimates. However, he said that it would always be necessary for the contractor to undertake additional quality level A investigations for its design and construction needs:

61. [66]

    Mr Sampson said that undertaking sufficient quality level A subsurface investigations to satisfy the project needs by the Request for Proposal stage would have been an extremely invasive process, with considerable time required and a large stakeholder impact. The full investigation process would have required significant amounts of CBD roads and verges to be closed. He also said that it would have been quite impracticable to consider an option that involved fully excavating ground materials to expose the utilities for the entire SLR route prior to commencement of the D&C contract (or at any time).

62. [67]

    The primary judge summarised Mr Sampson’s conclusion about pre-construction investigations at [435]:

63. [68]

    The primary judge also gave the following summary of Mr Sampson’s evidence:

64. [69]

    The primary judge concluded with the following summary of the experts’ evidence on this issue:

65. [70]

    The trial judge rejected the respondents’ case that there should have been a different delivery model: at [780]-[790]. In large measure this turned on the finding that utility owners such as Ausgrid would not have committed to any treatments in advance of a completed detailed design (as opposed to a concept level design). There was no challenge to that rejection. His Honour continued at [790]:

66. [71]

    The primary judge rejected TfNSW’s submission that the utilities risk was allocated to OpCo, holding that cl 12.3 provided no disincentive for overstaying in the fee zones. There was no challenge to that finding. The primary judge concluded at [804]-[805] that the risk of delayed completion of the works in each fee zone caused by the discovery of unknown utilities and changes to known utilities continued, and despite TfNSW’s initial plans that OpCo assume that risk, the result was “an arrangement which imposed little of that risk on the D&C contractor”, and in turn “the business owners were exposed to that risk and the risk came home”.

67. [72]

    Under the heading “Was the interference with the plaintiffs’ businesses substantial and unreasonable?”, the primary judge concluded that it was at [910]-[923].

68. [73]

    His Honour said:

69. [74]

    The primary judge said at [914] that the period of the interference must be an important factor in assessing reasonableness, and noted that it was lengthy and much longer than TfNSW had assured businesses would occur. His Honour said that other relevant factors included the nature and purpose of the activities, the relationship between the parties, including the obligations of the landowner from which the nuisance emanates, the period during which the interference was substantial, the benefit of the activities to the public, whether the landowner took care to avoid unnecessary interference, whether there were self-help measures available to the claimants, and the extent to which the defendant might have known or anticipated that the interference would impact on the financial interests of the adjoining landowners. Many of those factors pointed in favour of a longer period of time for construction. The essence of the evaluative decision made by his Honour was as follows:

70. [75]

    That conclusion is, necessarily and explicitly, based on a finding that the IDP or the amended IDP represented a reasonable estimate of the time within which the work could be done.

71. [76]

    The primary judge also rejected TfNSW’s submission, reiterated in this Court in support of ground 2 of its appeal, that “if it was reasonable to undertake the works, it did not become unreasonable during the course of the works”. His Honour said:

The notice of appeal

1. [77]

   The notice of appeal contained 11 grounds. Parts or all of grounds 1, 2(b), 3, 5, 6, 7 and 8 turn on questions of fact. Grounds 2(a) and 4 turn on questions of law, as do the grounds concerning s 43A.

2. [78]

   Ground 1 challenged the finding that Hunt Leather and Ancio suffered an interference which was both substantial and unreasonable for which TfNSW was responsible, primarily because it was not open to his Honour to have regard to the IDP (or the amended IDP) as a reasonable estimate. Ground 2(b) was that the primary judge erred in finding that TfNSW did not use reasonable care “to protect the interests of the business owners” along the Sydney Light Rail route. Ground 3 was that the primary judge failed to give adequate reasons for finding that construction activities took substantially longer than planned. Ground 5 maintained that the primary judge erred in concluding that the use of the roads along the route for the purpose of constructing the Sydney Light Rail was an “exceptional” rather than a “common and ordinary” use of the roads. Ground 6 was that the primary judge erred in finding that TfNSW failed to show that any nuisance caused by the construction of the SLR for which TfNSW would otherwise be liable was inevitable. Ground 8 was that the primary judge erred in finding that the construction activities took substantially longer than planned because of the discovery of previously unknown utilities and the time taken to reach agreement with utility providers, as well as “the issues surrounding Modification 25”, and also erred in finding that there was a sufficient evidentiary basis to conclude that the nuisance commenced on 1 November 2016 and 3 September 2017 in respect of Hunt Leather and Ancio respectively, based on the amended IDP.

3. [79]

   Ground 2(a) was that the primary judge failed to treat as determinative whether TfNSW had taken reasonable care in connection with the construction activity. This ground amounted to the proposition that because the construction of the SLR was a means to an end, and was a legitimate endeavour, it could not constitute an actionable nuisance, at least so long as TfNSW was exercising reasonable care. Ground 4 complained of error in finding that TfNSW bore the onus of proving that it took all reasonable care in connection with the construction along the route of the SLR project.

4. [80]

   Ground 7 concerned the calculation of damages, and challenged the allowance in the calculation of damages given for a 12 month “recovery period” after the cessation of the occupation of the fee zones.

5. [81]

   Grounds 9, 10 and 11 concerned the defence under s 43A of the Civil Liability Act.

6. [82]

   The law of nuisance, perhaps more than most areas of law, gives rise to difficult questions of principle. In part this reflects the fact that there are relatively few decisions of the High Court, and some of those (including Hazelwood v Webber (1934) 52 CLR 268; [1934] HCA 62 and Torette House Pty Ltd v Berkman (1940) 62 CLR 637; [1940] HCA 1) largely concern the rule in Rylands v Fletcher, and arguably require revisiting in light of the High Court’s decision in Burnie Port Authority v General Jones Pty Ltd (1994) 179 CLR 520 at 555-556; [1994] HCA 13 to the effect that the rule should mostly be seen as “absorbed by the principles of ordinary negligence”, subject to the qualification that “there may remain cases in which it is preferable to see a defendant’s liability in a Rylands v Fletcher situation as lying in nuisance (or even trespass) and not in negligence”. In part it is a consequence of the divided decision of the High Court in Hargrave v Goldman (1963) 110 CLR 40; [1963] HCA 56, where the point of division was whether liability could extend to nuisance (as Taylor and Owen JJ considered) or whether the defendant could only be liable in negligence, the view advanced by Windeyer J. In part it is a consequence of the fact, noted by the primary judge, that some aspects of the Australian law of nuisance have diverged from that in England and Wales. In part it reflects the division seen in the judgments in Fearn v Board of Trustees of the Tate Gallery [2024] AC 1; [2023] UKSC 4, a decision pressed upon this Court by the respondents, to which we shall return. In part it is a consequence of statute, because many of the nineteenth and early twentieth century decisions decided as actions for private nuisance would today be determined using statutes, notably, planning and environmental legislation. In part it may reflect the proposition, urged upon this Court by TfNSW, that there is not a single law of nuisance, but instead there are separate categories of conduct capable of resulting in a substantial and unreasonable interference to the enjoyment of land, one of which is construction works, to which it was contended special rules apply.

7. [83]

   This Court had the advantage of receiving submissions from both sides of a very high quality, which extended to many points of principle. That said, this Court’s function is to resolve the appeal, rather than attempting to “discern and apply some other all-embracing criterion of liability in the common law”; cf Mann v Paterson Constructions Pty Ltd (2019) 267 CLR 560; [2019] HCA 32 at [76]. In accordance with the approach there stated by Gageler J, we shall apply “the standard common law judicial technique of deciding no more than what needs to be decided”. In large measure, that turns on the resolution of issues of fact, as is clear from the foregoing summary of the grounds of appeal.

The main issue of fact

1. [84]

   The primary judge found that insofar as construction took longer than the timeframes identified in the amended IDP, it was actionable nuisance. Ground 1 challenged the factual premise of that reasoning. For the reasons which have been anticipated in the foregoing, this ground is made out.

2. [85]

   The IDP could not on its face be regarded as anything like a reasonable estimate of construction time in any particular fee zone. It made no allowance for inclement weather. It made no allowance for the discovery of unknown utilities and their treatment. It made no allowance for various other contingencies. It may well have been the basis for the media release, but if so it was not a reasonable basis for any estimate of construction time in any particular fee zone.

3. [86]

   The “amended IDP” increased the time each fee zone would be occupied by construction for a few months, to address deficiencies apparent to Mr Griffith on the face of the original IDP, such as lack of allowance for bad weather and certain contingencies. However, the amended IDP proceeded on the basis that no construction would commence until there was complete knowledge of the thousands of sub-surface utilities along the route.

4. [87]

   It was not established that it was possible, and if so how long it would take and whether that too would amount to a substantial interference with the enjoyment of the plaintiffs’ land, to obtain complete knowledge of the unknown utilities. These points are developed below.

5. [88]

   Fee zone 5 was a single block – George St between King and Market Streets. So too were many other of the fee zones in the Sydney CBD (fee zone 6 was George St between Market and Park Streets; fee zone 7 was George St between Park and Bathurst Streets, fee zone 8 was George St between Bathurst and Liverpool Streets, and so on). It is in the order of 230-250m long. Before construction work commenced, 67 utilities were identified, and a further 55 utilities were identified during construction. That is an average of a utility every couple of metres.

6. [89]

   The utilities identified in advance of construction along the Sydney Light Rail route were extremely numerous. Mr Griffith’s report stated in Attachment 6:

7. [90]

   In addition to the thousands of utilities identified prior to construction commencing, a very large number of hitherto unknown utilities were uncovered during construction. Paragraph 633 in Attachment 6 of Mr Griffith’s report identified that there were 55 unknown utilities in fee zone 5 (compared to 67 known before construction), 46 in fee zone 6 (compared to 37), 48 in fee zone 28 (compared to 135) and 103 in fee zone 29 (compared to 106). That is to say, in fee zones 5 and 29, there were approximately as many utilities discovered during construction as had been located in advance of construction, and in fee zone 6 (the block of George St between Market and Park Streets), another 46 – more than the 37 which had been detected prior to construction – were discovered during construction.

8. [91]

   It is perfectly plain that the utilities were densely located, at least in the inner CBD fee zones such as in fee zone 5 and nearby. There was a utility which required treatment every few metres, with dozens on each city block.

9. [92]

   In those circumstances, the idea that a series of investigative trenches could be dug so as to identify all of these utilities (for example, a series of night works, with the road and footpath being temporarily restored in time for morning traffic) seems implausible. Yet the premise of the reasoning advanced by the plaintiffs and adopted by the primary judge was that the occupation of the fee zones became tortious after the expiry of the times in the amended IDP, with those times being based upon complete knowledge of the utilities prior to construction commencing. If that were to be the basis of liability, then the plaintiffs needed to establish:

   1. (1)

      first, that it was possible in some rational way to obtain complete knowledge of the utilities prior to construction, and

   2. (2)

      secondly, assuming it were possible to do so, how much interference would that investigation cause.

10. [93]

    It would in theory be possible to obtain complete knowledge of the utilities by digging up the entirety of George St in each fee zone, uncovering and identifying the utilities. That could not sensibly be done, however, without blocking the street for some extended period of time; at least, there was no evidence that this could be done in any feasible timeframe, and it seems inherently problematic. It also seems most unlikely that that could be done by a series of night works, with temporary resurfacing in time for the road to be reopened the following day. And it would be absurd to dig up the entire road and footpath in order to discover all the utilities, then to resurface it, only to proceed later to undertake the construction works, in the hope that the construction would conform to the timeframes in the amended IDP.

11. [94]

    Perhaps there are sensible solutions to those problems. It is easy to see that if more pre-construction work were undertaken, many more of the unknown utilities could be identified. It is also easy to see that, since the utilities were so numerous, it would be unlikely without digging up the entirety of the roadway to identify all of them, and that delays would still be caused subsequently by those that were only uncovered during construction. And it is easy to see that there would be substantial interference with the enjoyment of the group members’ land by the more extensive pre-construction activities required to obtain a greater understanding of the location of unknown utilities. These considerations make all the more conspicuous the absence from the evidence adduced by the plaintiffs of any explanation of how in some rational way without causing further substantial interference with occupiers’ enjoyment of their property could complete knowledge of the utilities be obtained in advance of construction.

12. [95]

    Enough has been said to demonstrate that (a) there was no evidence for a critical element of the plaintiffs’ case for damages based on the occupation of fee zones for times extending longer than the amended IDP, and (b) it is far from self-evident that there was a form of pre-construction investigation which would have reduced the interference with the plaintiffs’ enjoyment of their land.

13. [96]

    Separately from the above, let it be assumed that there was complete knowledge of the utilities, and it was reasonable to complete construction in the timeframes stated in the amended IDP. That does not establish that there was actionable nuisance for every week or month that any time zone was occupied for longer than the time period stated in the amended IDP. Even with complete knowledge of the utilities, the Sydney Light Rail was a very complex project. The amended IDP contained thousands of activities, many of them interdependent. The whole project was to take years, and the amended IDP had fee zone 5 occupied for more than a year. It cannot be the law that construction authorised by statute becomes actionable nuisance if it takes a month or two months or three months longer than scheduled. But the plaintiffs did not demonstrate when the time taken for construction became actionable nuisance in light of the number of sub-surface utilities that needed to be treated for the purposes of construction. Whether the time in fact taken in light of that universe of practical issues and challenges was unreasonable was not the subject of evidence or analysis.

14. [97]

    This is dispositive of the appeal. This is not a case where the Court should not let mere difficulty in determining damages stand in the way of an assessment of loss. This is a case where the plaintiffs have failed to prove a critical integer of their case. The result is that the appeal must be allowed, the orders made at first instance set aside, and judgment entered in favour of TfNSW.

Ground 2(b) – did the primary judge err in finding that TfNSW did not use reasonable care

1. [98]

   On the view taken by the primary judge, this issue did not arise. His Honour introduced this short section of his reasons as follows:

2. [99]

   The primary judge also noted that, as the issue was pressed by TfNSW, he would consider it. That approach accorded with the general principle that a trial judge should determine all issues in order to assist the appeal process and obviate the need for a retrial, and was especially apt in circumstances where there was a long and factually intensive trial in an area where the law is not free from doubt: see the authorities considered in PPK Willoughby Pty Ltd v Baird [2021] NSWCA 312 at [15] and Shinetec (Australia) Pty Ltd v The Gosford Pty Ltd [2024] NSWCA 174 at [129]-[131].

3. [100]

   The primary judge was also conscious, because of the way the issue arose, that certain difficulties stood in the way of the determination of this issue. The plaintiffs had disavowed establishing a failure to take reasonable care, and their case was advanced accordingly. His Honour noted at [815] that:

4. [101]

   Thereafter, his Honour explained why he was unsatisfied that TfNSW had established that it had used reasonable care. His Honour’s reasons were attuned to the way in which the issue arose, and (contrary to the way this ground is formulated) fall short of a finding that TfNSW did not use reasonable care. There is of course a difference between the two. The circumstances are unusual, because normally a plaintiff will seek to establish a failure to take reasonable care, and it is sufficient for a defendant to resist that finding, rather than to seek and obtain a finding that it did exercise reasonable care.

5. [102]

   In this Court, TfNSW complained that the primary judge did not engage sufficiently with the facts. It submitted that his Honour’s statements were “high-level and conclusory pronouncements that gloss over the complex factual matters that needed to be decided before the primary judge could conclude there was a failure to exercise reasonable care”. The criticism is more than a little unfair, in respect of a finding which did not arise on the view his Honour took and did not arise on the approach adopted by the plaintiffs.

6. [103]

   The circumstances are unusual. TfNSW asked the Court to make a negative finding that it did not fail to use reasonable care, when the converse positive finding was disavowed by the plaintiffs. TfNSW advanced submissions based on the need to identify the risk of harm, to apply the so-called “calculus” in Wyong Shire Council v Shirt (1980) 146 CLR 40 at 47; [1980] HCA 12, and to make specific findings of fact, which submissions would be well-founded if liability in this case was based on a finding of breach of duty. However, that was not how this litigation was conducted. In particular, the primary judge was not asked to find, and did not purport to find, how any failure to take reasonable care caused damage to the plaintiffs. There is an inconsistency in TfNSW on the one hand insisting that the Court make a finding on an issue, and on the other hand complaining that when it declined to find that TfNSW had not failed to use reasonable care, the Court did not make precise findings about risk of harm and the steps which a reasonable person would have taken when TfNSW did not itself identify what had occurred.

7. [104]

   Another way of looking at this is as follows. The primary judge was acutely aware of the fact that neither side had given an account of the course of construction of the Sydney Light Rail, metre by metre or fee zone by fee zone. It was not suggested that there was any inaccuracy in the following account given early on in his Honour’s reasons:

8. [105]

   Litigants are of course free to make forensic decisions to have issues determined on inferences and onus. But TfNSW cannot have it both ways. TfNSW cannot on the one hand choose to leave unexplained why the construction activities took so long, and on the other hand complain when the primary judge declined to make a negative finding that TfNSW did not fail to exercise reasonable care. To the contrary, given the evidentiary background, it is difficult to see how any other result was possible.

9. [106]

   This sub-ground is not made out.

Grounds 3 and 8 – adequacy of reasons and cause of delay

1. [107]

   These grounds were addressed concurrently. Insofar as these grounds are based on the difficulties of using the amended IDP to determine when the actionable nuisance began, they have been addressed above.

2. [108]

   Insofar as ground 3 complains of a failure to give reasons, nothing turns on it. TfNSW did not seek a retrial. Whether or not there was a failure to give reasons, this Court will conduct a real review of the trial. Even if there was a failure to give reasons, if the plaintiffs were correctly found to be entitled to judgment, the appeal will be dismissed, and if they were not entitled to judgment, the appeal will be allowed. These points were made in Housman v Camuglia (2021) 104 NSWLR 615; [2021] NSWCA 106 at [33]-[34] by reference to earlier authority:

3. [109]

   The substance of the remainder of these grounds challenges the identification by the primary judge of the cause of the delay. TfNSW contends that the primary judge erred in finding that the construction of the Sydney Light Rail took substantially longer than had been planned because of the discovery of unknown utilities and the need to treat them (and also the issue relating to “Modification 25”, which involved the relocation of one of the stops).

4. [110]

   The primary judge found at [779] that the construction activities in Fee Zones 5 and 29 took substantially longer than planned because of the discovery and identification of so many previously unknown utilities and the time taken to reach agreement with utility providers and treat the utilities, and that in respect of Fee Zone 6, and to a much lesser extent Fee Zone 5, a substantial delay arose from the issues surrounding Modification 25. That finding was based on the opinions expressed by Mr Griffith, the evidence of Mr Sampson that the discovery of unknown utilities would have made a “huge difference” to the construction timing, and his Honour’s review of about 120 contemporaneous documents which referred to utilities in the context of problems and delays. Of the latter, his Honour referred in [777] by way of example to the D&C Contractor’s progress report for July 2016 which referred to the inevitability of fee zone slippage for a number of reasons, including a “significant increase in unforeseen obstacles in the ground”. In [778] his Honour summarised a further 16 documents which referred to the discovery of utilities and difficulties in their being treated by the utility providers.

5. [111]

   TfNSW’s complaint in this Court is the same as was made at trial. It said that it was not possible to allocate causal significance to the discovery of unknown utilities without identifying when each new utility was discovered, what the utility provider required in respect of that utility, what treatment was required, and how long that treatment took. The primary judge recorded this submission at [768].

6. [112]

   The primary judge also recorded at [774] that:

7. [113]

   That paragraph is plainly correct.

8. [114]

   We see no error in the primary judge relying on the evidence identified by his Honour to draw the conclusion that the discovery of hitherto unknown utilities was a substantial contributor to the time taken to construct the Sydney Light Rail.

Ground 5 – “exceptional” as opposed to “common and ordinary” use

1. [115]

   This ground maintained that the primary judge erred in finding that the use of roads for the purpose of construction was “exceptional”. It is a relatively minor ground, described in TfNSW’s submissions as one of the “miscellaneous errors”.

2. [116]

   The primary judge said:

3. [117]

   This ground’s relative insignificance is reflected in the observation made by the primary judge in the immediately following paragraph, that “such a finding does not have the significance that the plaintiffs suggest”, and the statements that in his Honour’s view, it was “merely a factor to be assessed in considering the reasonableness of the interference” and “does not lead to the imposition of a form of strict liability”. His Honour was there referring to the submission advanced by the plaintiffs, and recorded at [648] that “the use of the roads constituting the light rail route was not common and ordinary and, as such, it is only necessary to establish that the interference with the plaintiffs’ land was substantial in order to succeed”.

4. [118]

   Fearn v Board of Trustees of the Tate Gallery [2024] AC 1; [2023] UKSC 4 holds that a “common and ordinary use”, which is “conveniently done”, is not actionable, even if causing material discomfort or inconvenience, while conversely uses which are not “common and ordinary” result in liability even if the defendant conducted itself reasonably. These converse propositions find support in Lord Leggatt’s reasons at [25]-[28]. However, as the primary judge noted at [650], Lord Sales, dissenting, favoured a broader approach, holding that while whether the defendant’s conduct was common or ordinary according to the standards of the locality was “highly relevant” (at [235]) or “highly significant” (at [236]), it was not sufficient for a plaintiff merely to point to a substantial interference with the enjoyment of land by reason of conduct which was not a common or ordinary use of the land, and instead other facts came into play in assessing the conduct against an objective standard of reasonableness: at [234]-[237]. The primary judge said at [651] that it seemed to him that Lord Sales’ observations were more consistent with the law as it had developed in Australia.

5. [119]

   On this point we agree with the primary judge. To take but one example, Gartner v Kidman (1962) 108 CLR 12; [1962] HCA 27 concerned whether a landowner onto whose land water flowed as a result of artificial works was entitled to perform works which caused damage to the upper owner. Most of the High Court’s reasons concern whether there was a natural watercourse and whether the law governing riparian owners applied. It was held by Windeyer J, with whom Dixon CJ agreed, that the position was as follows (at 49):

6. [120]

   What was there said about the inchoateness of reasonableness reflected the following passage at 47:

7. [121]

   This Court is bound by Gartner v Kidman. Nor does that decision stand alone. The examples given by Jordan CJ in Don Brass Foundry Pty Ltd v Stead (1948) 48 SR (NSW) 482 at 486-487 illustrate the role played by “reasonableness” as a restricting element in nuisance, a point noted in Gales Holdings Pty Ltd v Tweed Shire Council (2013) 85 NSWLR 514; [2013] NSWCA 382 at [279]. In light of those authorities, we do not think it is open to adopt the approach taken by Lord Leggatt insofar as it dilutes the role of reasonableness.

8. [122]

   This was a case where the roadway and footpath were occupied for the purpose of construction, in part to do work, in part to store equipment and materials, but mostly because it was not possible to complete the works because of other steps which had first to be taken (such as treating thousands of utilities). That was not an ordinary use of the land dedicated as a road. The essential aspect of land being a road is that there be a right of public passage. The essential aspect of the nuisance of which Hunt Leather and Ancio complain is that the construction activities prevented public access to parts of the road and footpath, which substantially interfered with their enjoyment of their land.

9. [123]

   TfNSW criticises the reliance by the primary judge on the facts that it was not a roads authority and that historically there were once trams which ran on George St. It may be agreed that those considerations have little force. TfNSW’s real point is based on what was said by Lord Leggatt in Fearn: that “[t]he right to build (and demolish) structures is fundamental to the common and ordinary use of land, involving as it does the basic freedom to decide whether and how to occupy the space comprising the property”. But there is no free-standing right to build a structure, even on one’s own land. Nor is there a free-standing right to demolish a structure, even on one’s own land. In most cases it is an offence to do so without development consent: Environmental Planning and Assessment Act 1979 (NSW), s 4.2 (formerly s 76A). That statute has long identified the threefold distinction between development which may be carried out with consent, development which may be carried out without consent, and development which is prohibited. So too has zoning in instruments made pursuant to that statute, as Spigelman CJ’s reference to “the conventional threefold structure” at the outset of his reasons in Woolworths Ltd v Pallas Newco Pty Ltd (2004) 61 NSWLR 707; [2004] NSWCA 422 at [2] illustrates.

10. [124]

    Similarly, in the 21st century, many of the activities which gave rise to claims of nuisance in the cases, including the emanation of noise, smoke, dust and odours from certain uses of land, are at least in large measure regulated under planning laws, mining legislation and environmental legislation. Noisy works when demolishing or constructing the equivalent of a large department store like Selfridges if they took place in 21st century Sydney would unquestionably be subject to detailed conditions of consent regulating when such works could take place.

11. [125]

    We do not accept TfNSW’s submissions that the primary judge erred in finding that the use of the fee zones was exceptional.

12. [126]

    A final point may be noted in response to this ground. The significance of whether a use is “common or ordinary”, as opposed to “exceptional” appears to be a product of decisions made at a time when the rule in Rylands v Fletcher was regarded as sui generis. As Windeyer J observed in Gartner v Kidman (1962) 108 CLR 12 at 44; [1962] HCA 27, “the notion of a natural use of land, and the distinction between a natural and a non-natural use seems to have come into the law at this point from Lord Cairns’ judgment in Rylands v Fletcher”, and Lord Cairns appears to have used “ordinary” and “natural” uses of land interchangeably: Rylands v Fletcher (1868) LR 3 HL 330 at 338. It is far from clear that that question of characterisation continues to serve the purpose it once served – as a threshold element of liability – rather than as an ingredient in the assessment of reasonableness in the manner indicated above.

Ground 6 – was any nuisance caused by TfNSW inevitable?

1. [127]

   This ground proceeded on the basis that the primary judge had correctly found at [828] that “if [TfNSW] can establish that the interference with the plaintiffs’ property (which might otherwise give rise to an action in nuisance) was an inevitable consequence of carrying out the work, then the plaintiffs would not succeed”. We did not understand the respondents to contend to the contrary.

2. [128]

   It was accepted that TfNSW was authorised by statute to construct the Sydney Light Rail, and as McLure P explained in Southern Properties (WA) Pty Ltd v Executive Director of the Department of Conservation and Land Management (2012) 42 WAR 287; [2012] WASCA 79 at [122]-[123]:

3. [129]

   That understanding of “inevitability” was recently adopted by the Supreme Court of the United Kingdom in The Manchester Ship Canal Company Ltd v United Utilities Water Ltd (No 2) at [18]-[20] and [94].

4. [130]

   The primary judge addressed this, as follows:

5. [131]

   TfNSW devoted a mere two paragraphs of its submissions in chief to this ground, and a single paragraph of its submissions in reply. Its complaint was that the observations by the primary judge “ignore the fact that the appellant’s planning, and in particular the IDP, were based on the information that was available at the time”, that perfect foresight in such matters is unrealistic, and that the plaintiffs failed to prove any alternative model for the project that would have produced a better outcome. TfNSW said that there was a need to apply a “common sense appreciation which cannot be rigidly defined, of practical feasibility, in view of situation and expense” (being the language used in Manchester Corporation v Farnsworth [1930] AC 171 at 183), in light of which this Court should find that the primary judge erred in finding that TfNSW did not show that any nuisance was inevitable.

6. [132]

   We do not accept this ground, for reasons which can be stated concisely. There was no examination of the events which occurred each day on the ground while the Sydney Light Rail was being constructed. There is no proper basis for a conclusion that there was nothing which TfNSW could have done to permit the occupation of fee zone 5 being, say, three months or six months shorter than it was. It is to be borne in mind that this ground is propounded as an answer to the entirety of the appeal.

7. [133]

   More generally, in circumstances where there was no detailed examination of the particular decisions made by contractors to order the various construction activities and to maintain occupation of the fee zones, it is tolerably clear that TfNSW was not able to demonstrate that the delay that occurred was in fact inevitable.

8. [134]

   Like so much in this trial, this ground reflects the parties’ forensic decisions. It is one thing not to accept the plaintiffs’ submission that the amended IDP was a guide to the time after which the occupation of the fee zones because actionable. It is another thing to accept TfNSW’s submission that the time which was taken was inevitable.

Grounds 2(a) and 4 – reasonable care and onus

1. [135]

   Ground 2(a) challenges the rejection by the primary judge of TfNSW’s submission that because construction was a legitimate activity, done for the end of establishing a light rail service, then there could be no nuisance if TFNSW showed that it was using reasonable care. This ground does not strictly arise for two reasons: first, because of TfNSW’s success based on the insufficiency of the amended IDP as a basis for the judgment against it, and secondly because of the rejection of ground 2(b) above.

2. [136]

   Ground 4 is related. It is that the primary judge erred in finding that TfNSW bore the onus of establishing that it took reasonable care in connection with the construction of the Sydney Light Rail. It is convenient to address both grounds together, and to do so concisely since they are non-dispositive.

3. [137]

   The first submission is contrary to authority. When determining whether a claim of private nuisance has been made out, it is necessary to ask whether a particular use is “reasonable according to the ordinary uses of mankind living in society, or more correctly in a particular society”: Elston v Dore (1982) 149 CLR 480 at 488; [1982] HCA 71, citing Sedleigh Denfield v O’Callaghan [1940] AC 880 at 903-904. It does not follow from the fact that a defendant has established that a use is being undertaken with reasonable care that there can be no nuisance. As much was rejected by Sir George Jessel MR in Broder v Saillard (1876) 2 Ch D 692 at 701-702:

4. [138]

   In Munro v Southern Dairies Ltd [1955] VLR 332 at 336, Sholl J said:

5. [139]

   Sholl J mentioned with approval the point made in the headnote to the report of Painter v Reed [1930] SASR 295, which reads:

6. [140]

   In Rapier v London Tramways Co [1893] 2 Ch 588, the defendant was authorised by statute to lay tracks for a railway. To that end, it established stables in a residential area to house carriage-horses. The plaintiffs complained of the noise and smell from the stables. The defendant was held liable in nuisance although it proved that it took reasonable steps to mitigate the nuisance. It was said at 600:

7. [141]

   Similarly, in Pwllbach Colliery Co Ltd v Woodman [1915] AC 634 at 638, Earl Loreburn said:

8. [142]

   Similarly, Sir John Salmond wrote in his Law of Torts (1912) at 199:

9. [143]

   Professor Nolan reiterated these points in D Nolan, “‘A Tort Against Land’: Private Nuisance as a Property Tort” in D Nolan and A Robertson (eds) Rights and Private Law (Hart Publishing 2011) 459 at 468:

10. [144]

    As Lord Hoffmann indicated in Southwark London Borough Council v Tanner [2001] 1 AC 1 at 20; [1999] UKHL 40, “it may in one sense be quite reasonable to burn bricks in the vicinity of convenient deposits of clay but unreasonable to inflict the consequences upon the occupants of nearby houses”. The two are analytically distinct questions, and it is only the latter that is determinative. Lord Hoffmann’s distinction recalled that made some 160 years ago in Bamford v Turnley (1860) 3 B & S 62; 122 ER 25 when, in disapproving the jury direction made in Hole v Barlow (1858) 4 CB NS 334; 140 ER 1113 that there was no nuisance to neighbours if a brick kiln was being conducted in a reasonable and proper manner, Erle CJ, Williams J, Keating J and Wilde B said that the jury should not be asked “whether the causing of the nuisance was a reasonable use of the land”.

11. [145]

    Lord Simonds said that “if a man commits a legal nuisance it is no answer to his injured neighbour that he took the utmost care not to commit it”: Read v J Lyons & Co Ltd [1947] AC 156 (HL) at 183. Lord Millett said that “the use of the word ‘reasonable’ in this context is apt to be misunderstood. It is no answer to an action for nuisance to say that the defendant is only making reasonable use of his land”: Southwark at 20. Much the same point was made more recently by Lord Sales in Fearn at [165]: “liability turns on the issue of reasonableness as between the two parties located in a particular locality, not on the reasonableness of the defendant’s use in the abstract”.

12. [146]

    TfNSW’s proposition does not accord with principle. The reasonableness of a use is conceptually distinct from the reasonableness of a defendant’s conduct. That is the point made by, among others, Lords Hoffmann, Millett and Sales in the passages cited in the previous paragraphs. Use of the term “reasonable user” as a substitute for unreasonable interference is apt to conflate those two points.

13. [147]

    It may also be seen by examples. A landowner who uses land as an airbase in a residential area, albeit having taken all reasonable steps to minimise noise and disruption according to industry standards, is likely still liable in nuisance because the question is whether the nature of the interference itself is unreasonable in the locality: see J Murphy, The Law of Nuisance (Oxford University Press, 2010) pp 11-12, referring to Dennis v Ministry of Defence [2003] EWHC 793; [2003] EGLR 121, where noise from Harrier jump jets was held to be a nuisance.

14. [148]

    Another way of making this point is to observe that it is plain that a defendant who has undertaken or authorised a use of land which will cause substantial interference with the enjoyment of a neighbour’s land will owe that neighbour a duty of care. TfNSW’s submission posits that the defendant has failed to take reasonable care. In those circumstances, the defendant would be prima facie liable in negligence. Alternatively, if TfNSW’s submission were sound, a defendant who could establish that it performed construction work in a way which was not negligent could not be liable in nuisance. But there is no reason to think that the boundaries of two separate torts, which serve different ends and are prompted by different legal policies, are coterminous in the way acceptance of this ground would entail. After all, “the common law is not antipathetic to concurrent liability”: Henderson v Merrett Syndicates Ltd [1995] 2 AC 145 at 193. It would be profoundly anhistoric to confine torts so that they ceased to overlap, a point elaborated in Perera v Genworth Financial Mortgage Insurance Pty Ltd (2017) 94 NSWLR 83; [2017] NSWCA 19 at [44].

15. [149]

    This appears to have been the primary judge’s view at [814]. When the primary judge said that “the defendant bears the onus of proving that it took all reasonable care” his Honour appears to have meant no more than that proof of reasonable care (or lack thereof) is not an essential or necessary part of the plaintiffs’ case but was nonetheless relevant to the inquiry, and if TfNSW wished to raise it as an argument or factor that points against the unreasonableness of the interference, it was for it to do so. We respectfully agree.

16. [150]

    As part of this point, TfNSW contended that nuisance cases based on construction fell into a different category from those decisions based on water, or fire, or noise, or smells, or other continuing uses of land. It is true that the assessment of reasonableness in the sense mentioned above is different in the case of construction, which will be for a finite period, and directed to an end, as opposed to cases where the use which gives rise to the interference is an end in itself. It is also true that cases dealing with rights to water and ancient lights proceeded from different principles, based on proprietary rights and an inexact analogy with servitudes. And it may be acknowledged that TfNSW’s submission is consistent with the title and structure of Garrett’s works “The Law of Nuisances” (William Clowes & Sons, 1890, 2nd ed 1897; 3rd ed 1908). But we do not consider that there is a special category that provides additional protection to defendants who perform, or authorise the performance of, construction work, as opposed to other activities which interfere with the enjoyment of land. For one, “construction” works is an indeterminate category. Is the disturbance caused by truck movements during the one-off removal of the overburden of an open-cut mine in its “construction” phase different from the disturbance caused by truck movements once mining has commenced? For another, the submission sits ill with the trend in the last few decades and centuries of breaking down particular categories (such as the liability of landowners, or the liability of common carriers) or the rule in Rylands v Fletcher, or the ignis suus rule, and replacing them by more principled rules.

17. [151]

    Finally, in relation to onus, the question arose in unusual circumstances where TfNSW asked the judge to make a finding which the plaintiffs disavowed to be part of their case. But in any event, the approach taken by the primary judge accords with Andreae v Selfridge & Co [1938] 1 Ch 1, where a neighbouring landowner was found liable in nuisance for the noisy works undertaken by its contractors in the course of works on the Selfridge’s site in central London. The works caused disruption to the plaintiff’s hotel, especially when they were undertaken at night, and also during a demolition phase, where a large amount of dust and grit was carried onto her land. The reasons given by Sir Wilfrid Greene MR focus on the onus of proof, which he was satisfied Selfridge had failed to discharge. His Lordship said:

18. [152]

    After emphasising it was a matter of fact and degree, he said that “[t]he use of reasonable care and skill in connection with matters of this kind may take various forms”, but concluded that there had been a “reprehensible lack of regard for the duty which it owes to its neighbours”. (If similar issues recurred today, they would involve an examination of the terms of development consent.)

19. [153]

    These grounds do not arise, but if they did arise they are not made out.

Ground 7 – damages – timing issues

1. [154]

   This ground turns on an element of the claims for damages. For the reasons already given, it does not arise. It was addressed only briefly in oral submissions. It will be addressed concisely in these reasons.

2. [155]

   The primary judge assessed damages based not merely upon the period when the substantial interference with the enjoyment of the plaintiffs’ land was taking place, but also, in the case of Hunt Leather, for a further 12 month “recovery period”.

3. [156]

   The assessment of damages took place in circumstances where it was accepted that that construction would in the best case occupy some months, and that the turnover and potential profitability of a neighbouring occupier’s business would not bounce back to pre-construction levels on the day after that construction ceased. There is a measure of artificiality in this aspect of the claim, because the position is different following the completion of construction from the position which obtained before construction started. Formerly, each group member with premises on George St faced a busy street congested with traffic, which was replaced with a mostly pedestrianised street with a light rail service. But that may be put to one side for the purpose of addressing this ground.

4. [157]

   TfNSW appeared to submit in writing that since there would always be a recovery period, whether the construction took place in a way that amounted to actionable nuisance or otherwise, no recovery period was applicable for the purposes of the calculation of damages. That is a little too simplistic. The certainty of some recovery period does not mean that the recovery period will be the same after a period of construction which has taken so long that it amounts to actionable nuisance compared with a much shorter period of construction. The real question posed by the concept of a “recovery period” as a component of damages is whether, by reason of the time the roadway and footpaths were occupied by construction exceeding the period identified as non-actionable, there was some additional recovery period over and above the inevitable recovery period which would arise even if the construction took place in a way which did not give rise to any nuisance.

5. [158]

   The only answer to this ground advanced in the respondents’ written submissions was that leave should not be granted to advance this ground, which is contrary to what was put at trial. There, senior counsel for TfNSW squarely accepted that there would be a recovery period of some length. That submission reflects a response to the more extreme submission advanced in writing by TfNSW, but is no answer to the point made in the previous paragraph.

6. [159]

   There is force in the respondents’ submission that the trial appears to have been conducted on the basis that a recovery period was concededly appropriate for Hunt Leather, and the competing periods were 18 months and 12 months (reflected in the accounting calculations propounded by each side). But the principal significance of this ground in the event that it arose was that the recovery period for each group member must be particular to the particular member. That would seem to be so, at least in principle. For example, there is every reason to think that the recovery period for a newsagency business operating on high volumes and low margins would be different from the recovery period for a high-end luxury goods retailer. It was a matter for each occupier to establish loss, and that would depend on the particular characteristics of the businesses run by each member, one of which would be the time for pre-interference turnover to resume.

7. [160]

   As noted above, this ground does not arise. Nothing more need be said.

Grounds 9, 10, 11 – section 43A of the Civil Liability Act

1. [161]

   TfNSW relied on s 43A of the Civil Liability Act 2002 (NSW), which provides:

2. [162]

   The phrase “civil liability to which this Part applies” refers to Part 5 of the Act and, particularly, to section 40(1) of the Act, which indicates that “This Part applies to civil liability in tort”. It is uncontroversial that TfNSW is a “public or other authority” for the purposes of section 43A(1).

3. [163]

   The primary judge considered that s 43A had no application, because s 43A did not operate as a defence, and there was no occasion for applying the attenuated standard of care: at [705]-[706]. The primary judge went on to consider the position if, contrary to his view, s 43A could apply. In that case, it would be necessary for TfNSW to establish the preconditions to the section, namely, that liability was “based on” the exercise of, or failure to exercise, a particular power, and that the power was a “special statutory power”.

4. [164]

   At [709], the primary judge distinguished between “a liability based on the exercise of (or failure to exercise) a special statutory power and an act or omission involving an exercise of (or failure to exercise) such a power” (emphasis added). That distinction was significant to his Honour’s reasoning. As was the case in Gales Holdings Pty Ltd v Tweed Shire Council (2013) 85 NSWLR 514 at [197], it is not sufficient to point to an element of public power in the facts constituting the nuisance; what his Honour required was that the complaint be directed to the public authority’s choice to exercise or not exercise that power.

5. [165]

   In response, TfNSW said that s 43A was never intended to be so confined. The appellant took issue with the primary judge’s understanding of a key passage in Basten JA’s judgment in Curtis v Harden Shire Council (2014) 88 NSWLR 10; [2014] NSWCA 314 at [244], where his Honour appeared to distinguish between liability “based on” the exercise of a statutory power and liability merely “involving” such an exercise. To the contrary, TfNSW submitted that Basten JA cast the question as whether the liability is one “involving an exercise of, or failure to exercise, a special statutory power”. On that view, the relevant question was whether there is a “sufficient connection between the liability of the appellant (properly characterised) and its exercise of a special statutory power”. But the respondents drew attention to decisions which seemed in terms to draw the distinction to which the primary judge adverted, including in Weber v Greater Hume Shire Council (2019) 100 NSWLR 1; [2019] NSWCA 74 at [47], which distinguished between activities that could be “undertaken in accordance with the general law” as opposed to special statutory powers “reliant for their lawfulness on a statutory power”.

6. [166]

   The primary judge also considered that the powers supporting TfNSW’s construction works did not answer the description of “special statutory powers”. The appellant in this Court, and in the Court below, identified the relevant statutory power as section 104O of the Transport Administration Act 1988 (NSW), which reads:

7. [167]

   However, the primary judge accepted the respondents’ submission that, because s 104O is contingent on the declaration of a “route” under s 104N(2), and that declaration was not made by regulations until 15 September 2015, the appellant could not have been exercising the statutory power under s 104O before that time. Rather, the primary judge held that TfNSW was exercising the more general powers conferred on the appellant under cl 9 of Sch 1 of the Act. That provision empowered the appellant to “make or enter into contracts or arrangements with any person in connection with the exercise of [its] functions”.

8. [168]

   Section 104N(1) and (2) provided:

9. [169]

   The primary judge reasoned that power in section 104O could not meet the description in s 43A(2) of a power “of a kind that persons generally are not authorised to exercise without specific statutory authority”, distinguishing between “any statutory power” and “special statutory powers”.

10. [170]

    TfNSW disagreed with the primary judge’s construction of s 104N. The appellant submitted that there was no limitation in the definition of a “light rail system” that the system must be one in respect of which a route has been formally declared under s 104N(2). On its construction, s 104N merely refers to a power or permission in subsection (2) to declare a route; the latter is not a necessary condition for the former. TfNSW submitted that this construction was reinforced by the fact that the term “develop” in s 104O(1) is defined broadly in s 104L to include preparatory works, and that it would be absurd to hold that any construction or financing activities undertaken prior to the declaration of a route fell outside the scope of s 104O.

11. [171]

    TfNSW also submitted that the primary judge asked the wrong question – whether the power created or altered people’s legal status or rights or obligations without their consent. The appellant noted that in Curtis, Basten JA treated the power to erect signs near a road as a special statutory power for the purposes of section 43A. TfNSW also noted that language in the case law concerning the distinction between ordinary and special statutory powers for the purposes of s 43A is qualified and uncertain – eg in Roads & Traffic Authority of NSW v Refrigerated Roadways Pty Ltd (2009) 77 NSWLR 360; [2009] NSWCA 263, Campbell JA at [370] expressly left aside “whatever distinction there might be between a statutory power simpliciter, and a special statutory power”.

12. [172]

    TfNSW also submitted that the legislative history and purpose of s 43A indicates a broader mischief than attenuating the standard of care for claims in negligence – it is to “reduce the more general risk that decision-makers would act conservatively when trying to decide how to exercise their powers in the public interest”. Were Parliament minded to limit s 43A in the way contemplated by the primary judge, the appellant noted that a constructional choice plainly open to Parliament was to narrow the application of Part 5 of the Act to “civil liability in negligence”, which was not done.

13. [173]

    The respondents drew attention to cases which seem to confine the operation of s 43A in the way the primary judge contemplated, including in Precision Products (NSW) Pty Ltd v Hawkesbury City Council (2008) 74 NSWLR 102; [2008] NSWCA 278, where Allsop P said at [172] (Beazley and McColl JJA agreeing) that “it is plain that the drafter of section 43A was attempting to ameliorate the rigours of the law of negligence”. TfNSW’s response was to submit that s 43A was never intended to be confined only to claims in negligence. Because, so it was said, the primary judge erred in holding that the taking of reasonable care is not an essential question when determining liability for nuisance, the appellant submitted that that error carried over into the primary judge’s application of s 43A.

14. [174]

    Once again, these grounds do not arise in light of what has been said above as to the principal challenge to the finding of fact, and we address these grounds concisely.

15. [175]

    First, as the primary judge noted at [692], s 43A does not provide, strictly, for a “defence”. “That characterisation obscures the true operation of the section”, as Basten JA said writing for this Court in Della Franca v Lorenzato; Burwood Council v Lorenzato [2021] NSWCA 321 at [8], [107], [147]. Rather, s 43A(1) imposes an altered standard of care, attenuating the standard of care required to be proved when establishing the “civil liability in tort” to which s 40(1) refers.

16. [176]

    Secondly, it is no part of a claim for nuisance to show that the defendant failed to take reasonable care. The fact that a defendant who creates, adopts or continues a nuisance may be found liable demonstrates as much. And the same is true even when the defendant causes the interference directly. “Nuisance is a term used to cover a wide variety of tortious acts or omissions and in many negligence in the narrow sense is not essential”: Overseas Tankship (UK) Ltd v The Miller Steamship Co Pty Ltd [1967] 1 AC 617 at 639. It follows that s 43A has no operation.

17. [177]

    Thirdly, and consistently with the above, in Gales Holdings Pty Limited v Tweed Shire Council (2013) 85 NSWLR 514; [2013] NSWCA 382 this Court rejected reliance on s 43A in answer to a claim in nuisance. The correctness of this decision was not challenged. Emmett JA observed at [196]-[197] that:

18. [178]

    Leeming JA and Sackville AJA agreed that the civil liability in nuisance on which Gales sued was not based on the exercise or failure to exercise a statutory power within the meaning of s 43A: at [283] and [284].

19. [179]

    If a special statutory power is exercised negligently, or if it is alleged that it was negligent not to exercise a special statutory power, then the action for negligence is one that is based on the exercise or failure to exercise a special statutory power. In such a case, the effect of s 43A is that rather than needing to satisfy ss 5B and 5C, it is necessary to show that the exercise or failure to exercise the statutory power is unreasonable in the sense stated by s 43A(3). But the claim for nuisance in this case is not one which is based on the exercise of a special statutory power.

20. [180]

    Fourthly, it is not the exercise of a special statutory power to plan the construction of a light rail system. Nor is it the exercise of a special statutory power to determine to start construction after there had been incomplete identification of utilities along the route. The gravamen of the plaintiffs’ case was that TfNSW had either failed sufficiently to incentivise the contractor to adhere to the staging contemplated, or else had failed to address the risk posed by unknown utilities along the route. Those were not failures to exercise special statutory powers; they were complaints that TfNSW had not insisted on the right terms in a contract, or had not caused different things to have taken place before construction began.

21. [181]

    Those reasons are sufficient to conclude that these grounds are not made out. In an appeal where the application of s 43A does not alter the result, it is neither necessary nor desirable to address the operation of s 43A comprehensively.

The cross-appeal – should damages include the funder’s 40% fee?

1. [182]

   For the reasons already given, the appeal must be allowed. Turning to the cross-appeal, whether or not the partial costs order made by the primary judge and noted at [4] above discloses reviewable error is entirely arid, because the order will in any event be set aside. Likewise, the question whether damages should include a litigation funder’s fee does not arise. However, unlike the exercise of the discretion as to the costs of this particular trial, the question concerning the recoverability of the funder’s fee was the subject of full submissions, is a question of principle, and one which will recur in other cases. It is addressed below.

2. [183]

   At trial and on appeal, the cross-appellants submitted that the award of damages should include compensation for the litigation funder’s commission. The litigation funding agreement with International Litigation Partners No 16 Pte Ltd provided, relevantly, that the funder would pay all legal costs and disbursements, meet any adverse costs orders, provide any security for costs, provide “litigation management services”, and upon receipt of a “Resolution Sum” would be entitled to reimbursement of the costs it had paid, plus a “Funder’s Commission”. The amount of the commission principally depended on when the resolution was reached, and ranged between 25% and 40%. In the circumstances which led to the judgments being obtained in favour of Hunt Leather and Ancio, the fee was 40%.

3. [184]

   Common question 10 asked whether the cross-appellants were entitled to claim as a head of damages their reasonable litigation funding costs without needing to show that it was TfNSW’s nuisance which rendered them impecunious, that they would have pursued the claim without litigation funding had they the means, or that they negotiated over the terms of the litigation funding agreement. Common question 11 asked, assuming common question 10 was answered in the affirmative, what a reasonable rate of recoverable litigation funding costs would be.

4. [185]

   The primary judge found against the cross-appellants on this point, holding that litigation funding costs were not recoverable as a head of damages: Hunt Leather Pty Ltd v Transport for NSW (No 4) [2024] NSWSC 140. His Honour found that reasonable foreseeability was not a sufficient criterion to extend legal responsibility for loss, and considered that the claimed loss was too remote: at [100]-[102]. His Honour considered that the loss was not caused by the cross-respondent’s nuisance but rather by the cross-appellants’ decision, made freely and willingly, to enter into a litigation funding agreement: at [103]-[111]. Further, the primary judge did not consider it appropriate to extend legal responsibility for the loss to the cross-respondent: at [112]-[122]. Common question 11 did not arise.

5. [186]

   On appeal, the cross-appellants submitted that the loss and damage suffered by the group members included the cost of obtaining the funding necessary to bring the present proceeding, and that the trial judge erred in concluding otherwise. Necessarily, in light of his Honour’s reasons, they contended that the primary judge had erred in holding that the loss was too remote, that there was a break in the chain of causation, and that it was not appropriate for the cross-respondent to be responsible for the obligation to pay the litigation funder.

6. [187]

   The cross-appellants submitted that the trial judge’s approach to remoteness was contrary to the principle that the defendant is liable for reasonably foreseeable damage. They relied on Landoro (Qld) Pty Ltd v Jensen International Pty Ltd [1999] QCA 318, where the Queensland Court of Appeal allowed amendments to a statement of claim by which the appellant sought to recover the cost of litigation finance it had incurred as a result of the respondents’ alleged breach of contract. They also cited Thomas (as liquidator of Anne Lewis Pty Ltd (in liq)) v Arthur Hughes Pty Ltd [2016] NSWSC 1861 at [19], which refers to Landoro.

7. [188]

   The cross-appellants contended that the trial judge erred in finding that entry into the litigation funding agreement was not a loan to finance litigation, and could not be characterised as litigation finance or funding. They contended that finance can take on many forms, and that the funder’s obligation to pay legal costs and disbursements is in effect a loan to be paid from the final sum, together with a commission in the form of a contingency fee, especially in the case of funded group members, who have assigned part of the “value” of their claim to the litigation funder. They submitted that the real benefit of the funding arrangement was the prosecution of these proceedings, rather than, as the trial judge concluded, the avoidance of fees and exposure to adverse costs orders. They alleged that this error infected the trial judge’s assessment of whether the loss was caused by the cross-respondent’s conduct or was too remote.

8. [189]

   Further, the cross-appellants submitted that the litigation funding loss was a reasonably foreseeable consequence of the cross-respondent’s nuisance, and contended that the test for remoteness is whether harm of the kind suffered was reasonably foreseeable. In particular, they submitted that it was reasonably foreseeable that the nuisance would impact a wide class of persons, that it would be uneconomic for these individuals to bring separate claims and that they would be unwilling to take on the adverse costs risk, and thus that it was likely a funded class action would be brought. As they put it, but for the nuisance, the funding agreement would not have been entered into, and the case itself could not have been brought without funding.

9. [190]

   Finally, the cross-appellants submitted that the trial judge erred in concluding it was not appropriate that the cross-respondent be held responsible for the litigation funding losses. They noted again that the agreement would not have been entered into but for the cross-respondent’s conduct, that a funded class action was reasonably foreseeable, and that in the absence of a damages order accommodating this type of loss, the applicants and group members would not be put in the same position had the tort not been committed. They submitted that the question was not about scope of risk. Instead, they submitted, the purpose and policy of the law of nuisance made it appropriate for liability to extend to these losses.

10. [191]

    In response, TfNSW observed that the primary judge’s view that the litigation funding was not litigation finance was not of decisive significance in his Honour’s reasoning, and that in any event, the funding was more than simply passive litigation finance, as it rewarded the funder for the risk it assumed and provided remuneration for management services.

11. [192]

    TfNSW also submitted that reasonable foreseeability is not the sole restriction upon the recoverability of damages, and that remoteness principles and normative considerations also apply. Further, they submitted that they could not have reasonably been expected to foresee the events which transpired – specifically, the litigation and this type of funding. They noted that the scope of their obligation in issue was to avoid substantial and unreasonable interference with the use of the land, and that litigation funding was not within the scope of risk created by the nuisance. They submitted that the cross-appellants entered into the funding agreement independently of TfNSW’s conduct, breaking the chain of causation, and that it cannot be the case that litigants can deliberately choose to increase their loss. They also submitted that allowing this cost to be recoverable would advantage funded plaintiffs over others, and that it was not appropriate for the scope of a defendant’s liability to extend to payments made for the purposes of enabling litigation against that defendant.

12. [193]

    The primary judge was correct to dismiss this aspect of the plaintiffs’ claim, for the following reasons, which substantially overlap with those given by his Honour.

13. [194]

    First, there can be no doubt that the entry into the funding agreements was a voluntary act by some group members. Moreover, the premise of common question 10 was that no assumption was to be made that the group member was impecunious or that any impecuniosity was brought about by the tortious conduct of the defendant, thereby serving to emphasise the voluntariness of the group member’s decision. Their decisions to do so, or to choose not to do so, might be attended by very different considerations. A supermarket which is a franchisee and a supermarket which is a branch of a major listed Australian company might both be affected in similar ways by the construction of the Sydney Light Rail, but each might have quite different approaches as to how they would approach the decision to enter into the funding agreement.

14. [195]

    Those and similar considerations suggest that there are difficulties in treating the cost of the funding agreement as a component of the damages for which TfNSW would be liable. It seems decidedly odd that TfNSW would be liable for 40% more to a franchisee who entered into a funding agreement, than to a public company which operated a similarly sized and similarly affected supermarket next door.

15. [196]

    The group members who entered into funding agreements have chosen to bargain away the risk of being exposed to an adverse costs order and a liability to pay security for costs, as explained by the primary judge:

16. [197]

    We respectfully agree, although rather than describing the plaintiffs’ decision as being to increase “their loss” by 40%, we would say that they have promised to pay 40% of the damages to which they would otherwise have been entitled in return for an indemnity as to the costs of the litigation (including the potential liability for adverse costs orders and orders for security for costs). But we cannot accept the cross-appellants’ submission that the “real benefit” of the funding agreement was the prosecution of the proceedings. The “real benefit” was an indemnity for the certainty of the incurring of legal costs, and the practical certainty of a liability to provide security for costs, and the potential liability for an adverse costs order, all of which were incidents of the lead plaintiffs’ decision to prosecute the proceedings, and which were the quid pro quo for the funder’s “commission”.

17. [198]

    In those circumstances, the litigation funder’s fee is not to be regarded as a foreseeable loss caused by the defendant’s nuisance, but instead as the voluntary act of the particular plaintiff.

18. [199]

    Secondly, if the funder’s commission of 40% is recoverable as damages, then there would be every incentive for every group member, large or small, to enter into such an agreement. There would be no downside for any group member. Nor would there be any incentive to bargain for a smaller fee. We fail to see how that would serve any of the ends to which the legal system is directed. Some of the submissions advanced on behalf of the cross-appellants had the tenor that any measure which improved the ability of plaintiffs to bring proceedings against a defendant enhanced access to justice and was, for that reason alone, within the policy of the law. But in truth the position is more nuanced, as was observed by Glazebrook J writing for the New Zealand Supreme Court in Waterhouse v Contractors Bonding Ltd [2014] 1 NZLR 91; [2013] NZSC 89 at [41]-[42]:

19. [200]

    It is to be borne steadily in mind that litigation funders are not altruistic organisations dedicated to enhancing access to justice; they are self-interested investors, seeking to profit by involving themselves in litigation. The fact that litigation may be funded of itself provides no good reason to alter the rules of compensable loss for tort.

20. [201]

    Nor can we accept the cross-appellants’ submission that there is something about the policy of the law of nuisance which renders fees paid to a litigation funder recoverable. There is nothing special about the law of nuisance in this respect, nor is there any relevant difference between the funding of this representative proceeding and any other form of representative proceeding that seeks pecuniary remedies.

21. [202]

    Thirdly, at least in large measure, the benefit of the funding agreement was in meeting the group members’ costs and any liability for an adverse costs order or an order for security for costs. This suggests that if the quid pro quo provided by the group member – namely, the commission – were recoverable at all, it would be recoverable as a cost or expense of litigation, rather than as a component of damages. But if that is the correct characterisation, it presents an obstacle to the claim for the same amount as damages, in accordance with the principle in Anderson v Bowles (1951) 84 CLR 310 at 323; [1951] HCA 61:

22. [203]

    A Full Court of the Federal Court said in Gray v Sirtex Medical Ltd (2011) 193 FCR 1; [2011] FCAFC 40 at [15]:

23. [204]

    To substantially the same effect is the rule stated by Devlin LJ in Berry v British Transport Commission [1962] 1 QB 306 at 328 that expenditure on litigation is not recoverable loss.

24. [205]

    Fourthly, the entitlement of the litigation funder to its commission only accrued upon the entry of a judgment (or alternatively settlement). That not only tends to confirm the force of the previous point but also presents a conceptual difficulty, if not a paradox, identified by the primary judge at [67]:

25. [206]

    The plaintiffs are thus seeking to recover “a loss” which does not crystalise until after the defendant has been ordered, or has agreed, to compensate the plaintiffs in respect of their economic loss. It is a loss which the plaintiffs are not even obliged to tell the defendant about until a time of their choosing.

26. [207]

    We also cannot accept the cross-appellants’ submission that it is sufficient that the loss be reasonably foreseeable. For the reasons in the previous paragraph, it may be doubted that it is reasonably foreseeable, but the more fundamental consideration is that reasonable foreseeability is a necessary but not a sufficient touchstone for damages in this area. Some losses are taken as being caused by the plaintiff’s own voluntary act, and some are outside the scope of liability.

27. [208]

    Finally, the foregoing is consistent with the absence of any authority supportive of the cross-appellants’ submissions. Landoro does not assist them. A decision which permitted an amendment to go to trial is not authority for the proposition that the claim is sound, and indeed, Davies JA explicitly said in Landoro at [12] that “[t]o allow these claims to proceed is not, by any means, to assert their correctness”. The principles governing the circumstances in which novel claims may be pleaded are summarised in Perera v Genworth Financial Mortgage Insurance Pty Ltd (2017) 94 NSWLR 83; [2017] NSWCA 19 at [32]-[38]. The reference in Thomas v Arthur Hughes Pty Ltd takes the matter no further.

28. [209]

    For those reasons, the funder’s commission is to be regarded as the incurring of a liability by the group members’ separate voluntary decisions, rather than a consequence of any actionable nuisance. If that be wrong, then at least in the premises of common question 10, which is that it is unnecessary to establish that the nuisance rendered a group member impecunious, we would not regard the scope of the defendant’s liability to extend to the funder’s commission.

29. [210]

    The cross-appeal must be dismissed.

Conclusion and orders

1. [211]

   For those reasons, the appeal should be allowed, the orders made by the Court at first instance in favour of Hunt Leather and Ancio should be set aside, and in lieu thereof, judgments should be entered in favour of TfNSW. TfNSW did not suggest that any of the answers to the common questions given by the primary judge needed to be set aside. The cross-appeal should be dismissed. There is no reason of which the Court is presently aware for costs not to follow the event, both at first instance and in this Court, but that will not prevent any party from making an application within the time stated in UCPR r 36.16.

2. [212]

   The Court makes the following orders: