This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

Spandeck Engineering (S) Pte Ltd v Defence Science & Technology Agency
[2007] SGCA 37

Case Number : CA 3/2007
Decision Date : 08 August 2007
Tribunal/Court : Court of Appeal
Coram : Chan Sek Keong CJ; Andrew Phang Boon Leong JA; V K Rajah JA
Counsel Name(s) : Mohan R Pillay (MPillay) and Gopinath Pillai (Tan Peng Chin LLC) for the appellant; Tai Chean Ming, Chong Kuan Keong and Tan Joo Seng (Chong Chia & Lim LLC) for the respondent
Parties : Spandeck Engineering (S) Pte Ltd — Defence Science & Technology Agency
Tort  – Negligence  – Damages  – Pure economic loss  – Whether pure economic loss recoverable in Singapore  – Whether different test applying in relation to pure economic loss to determine duty of care  – Single test to apply notwithstanding damage claimed as pure economic loss

Tort  – Negligence  – Duty of care  – Applicable test to determine existence of duty of care  – Relationship between two-stage test and incremental approach  – Application of two-stage test comprising first proximity and second policy considerations with threshold consideration of factual foreseeability  – Incremental approach as methodological aid in applying specific criterion of two-stage test

Tort  – Negligence  – Duty of care  – Applicable test to determine existence of duty of care  – Whether type of damage claimed should result in different test  – Application of single (two-stage) test irrespective of type of damage claimed

Tort  – Negligence  – Duty of care  – Whether there was proximity between contractor and certifier given that contractor could submit disputes for arbitration  – Whether there was proximity between contractor and certifier given no direct contractual relationship between contractor and certifier  – Whether policy considerations negating finding of duty of care

8 August 2007

 

Chan Sek Keong CJ (delivering the grounds of decision of the court):

Introduction

1          This was an appeal against the decision of the trial judge, who dismissed the claim of Spandeck Engineering (S) Pte Ltd (“the appellant”) in the tort of negligence against the Defence Science & Technology Agency (“the respondent”). Her judgment is reported in Spandeck Engineering (S) Pte Ltd v Defence Science & Technology Agency [2007] 1 SLR 720 (“the Judgment”). At the end of the hearing, we unanimously dismissed the appeal. We now give our reasons.

2          Apart from the substantive rights and liabilities of the parties, which must surely form the primary focus of any judicial decision, this case was also about an intensely practical problem in the law of negligence: How does one ascertain whether a duty of care exists in law between one person and another where there is no pre-existing contractual or legal relationship between them? The common law is rudderless and surprisingly fraught with difficulties and uncertainties in relation to this deceptively simple and yet important question. The courts in Singapore are no less spared from this problem which we will now address. We turn first to the facts of the present case.

Background facts

Award of the contract for the project

3          The appellant’s claim stemmed from a dispute relating to the building project known as “The Proposed Redevelopment of Medical Camp in Nee Soon Camp Lot 212 PT MK 13 at Transit Road – Project LE(D) 1068 C08 MINDEF/DSTA” (“the Project”). As its name suggests, the Project was to redevelop a medical facility at Nee Soon Army Camp for the Ministry of Defence (“Mindef”). The Government of Singapore (“the Employer”) put the Project out for tender to which the appellant submitted a “base tender” and an “alternative tender”. In relation to this “alternative tender”, which was submitted on 15 March 1999,[note: 1] the appellant offered an alternative pre-cast/structural design. This alternative design consisted of a change in the original cast-in-situ design to the pre-cast structure design which would result in cost savings of about $200,000 and a shortening of the construction time by two months.[note: 2] The Employer accepted the alternative design and awarded the appellant a contract dated 24 June 1999 (“the Contract”) at a lump sum price of $31.78m to execute and complete the Project. The agreed commencement date of the Contract was 15 June 1999 and the completion date was 14 January 2001.

Salient terms of the Contract

4          The Contract incorporated the terms of the Public Sector Standard Conditions of Contract (“PSSCOC”), by which an appointed superintending officer (“SO”) was responsible for the administration and supervision of the Project, including certifying interim payments in respect of the appellant’s work for the Project. In particular, the Contract provided in cl 2.8(1) that:

The Superintending Officer, the Superintending Officer’s Representative or any assistant appointed … shall at no time be under any obligation or duty to the Contractor [ie, the appellant] either on behalf of the Employer or his own account to exercise or not to exercise any of his powers under the Contract, nor shall any failure to do so on his own part in any way prejudice the rights of the Employer against the Contractor or render the Employer liable to the Contractor.

5          In relation to the settlement of disputes relating to the Project, cl 34 was the governing clause. Its provisions, in so far as they are material, are set out below:

34.1      Reference to the Superintending Officer

(1)        If a dispute or difference of whatsoever kind shall arise between the Employer or the Superintending Officer or the Superintending Officer’s Representative and the Contractor in connection with or arising out of the Contract or the execution of the Works … including any dispute or difference as to any … certificate or valuation of the Superintending Officer or the Superintending Officer’s Representative, it shall in the first place be referred by either party in writing to the Superintending Officer for his decision. …

34.2      Reference to Arbitration

(1)        If either the Employer or Contractor is dissatisfied with the decision of the Superintending Officer made pursuant to Clause 34.1 hereof, … then the Employer or the Contractor may … give notice to the other party with a copy for information to the Superintending Officer of his intention to refer the decision or the dispute or difference that had not been decided to an arbitrator. …

However, the amount of damages for which the Employer might be liable to the appellant was qualified by cl 32.8 of the Contract, which provided as follows:

32.8      Delay in Certification

Under no circumstances shall the Employer be liable to pay to the Contractor [ie, the appellant] any damages, whether by way of interest or otherwise, for any failure or delay by the Superintending Officer in certifying any payment due or payable to the Contractor.

[emphasis added]

Appointment of the SO and consultants

6          On 7 July 1999, Mindef advised the appellant by letter that Quek Kheng Song (“QKS”) and Huang Siong Hui were appointed the SO and SO’s representative respectively, with Peh Chew Seng as the alternate SO’s representative in the absence of the SO’s representative. At that time, QKS was employed by the Employer as an officer in the Land & Estate Organisation in Mindef. In addition, various individuals and entities were also appointed as consultants to provide professional services for the Project (“the Consultants”). Among the Consultants appointed was KPK Quantity Surveyors (1995) Pte Ltd (“KPK”), which was appointed the quantity surveyor for the Project. In addition, RDC Architects Pte Ltd (“RDC”) was appointed the architect for the Project.

7          Subsequent to that, the respondent was constituted under the Defence Science and Technology Agency Act (Cap 75A, 2001 Rev Ed) in March 2000 and it was around that time that the respondent assumed the role of the SO. QKS was no longer the SO for the Project by this time but continued to represent the respondent at the Project after his employment was transferred from the Employer to the respondent.

Summary of tender and cost breakdown

8          The tender documents for the Project required the appellant to submit a summary of tender (“SOT”) as well as a cost breakdown (“CBD”). The SOT and CBD were documents prepared by KPK listing out all the items of work required for the Project being tendered. It required the contractors who were tendering for the Project to price this itemised list as a cost breakdown of the lump sum contract tender price, in the form of the CBD. The SOT was a summary of the detailed cost itemisation set out in the CBD.

Original SOT and CBD

9          As alluded to earlier (at [3] above), as part of its bid, the appellant had offered an alternative tender in addition to the base tender. The SOT and CBD which the appellant had submitted with its tenders were based on the base tender and did not reflect the breakdown under the alternative tender. This meant that the original SOT and CBD did not accurately reflect the scope and actual value of the works to be carried out by the appellant under the alternative tender.

SOT and CBD incorporated into the Contract

10        As such, on the basis of the Employer accepting the appellant’s alternative tender, KPK requested the appellant to submit a revised SOT and CBD to reflect the change of the design. Pursuant to this request, the appellant submitted a revised SOT and CBD dated 24 May 1999,[note: 3] which it alleged was incorporated into the Contract without any further revision. On the other hand, the respondent contended that there were further revisions to the SOT and CBD dated 24 May 1999 and that the final SOT and CBD were finalised by KPK with the appellant and incorporated into the Contract in or about October 1999.

11        In the Judgment at [9], the trial judge resolved this issue in favour of the respondent, where she said:

After the Employer had accepted the plaintiff’s alternative proposal, KPK requested the plaintiff to submit a revised SOT to reflect the change in design. The plaintiff submitted the first revision to KPK on 24 May 1999. After further revisions by the plaintiff, the revised SOT and cost breakdown (“CBD”) for the contract price were finalised by KPK with the plaintiff and incorporated into the Contract documents in October 1999. [emphasis added]

We agreed with the trial judge’s finding. It was clear to us that the SOT bound into the Contract documents[note: 4] was not the same as the appellant’s revised SOT dated 24 May 1999. That the SOT bound into the Contract was not a version prior to the SOT dated 24 May 1999 was supported by the fact that the total cost and contract period were based on the alternative tender. Given that the revised SOT dated 24 May 1999 was the first SOT based on the alternative tender, it must follow that any other SOT based on the alternative tender must have come after the SOT dated 24 May 1999. Accordingly, contrary to the appellant’s contention, the SOT dated 24 May 1999 was indeed revised before it was incorporated into the Contract. This will henceforth be referred to as “the Contract SOT” while the CBD incorporated into the Contract will be known as “the Contract CBD”.

Exercise to revise the Contract SOT and the Contract CBD

12        The appellant contended that, in the course of submitting its progress claims, it discovered that numerous items related to the alternative tender were omitted from the Contract SOT and the Contract CBD and this in turn allegedly led to the under-certification of the progress payments to the appellant. This had occurred even though the Contract SOT and the Contract CBD were bound into the Contract documents after revisions by the appellant itself. These alleged deficiencies in the Contract SOT and the Contract CBD were highlighted by the appellant by a letter dated 29 September 2000, in which it also submitted a revised SOT and CBD to KPK and requested KPK to review and approve them for use in the future assessment of the appellant’s progress claims.[note: 5]

13        The request was initially rejected by the Consultants by their letter of 11 October 2000,[note: 6] who regarded it as a departure from the terms of the Contract. However, by this time, the respondent and the Consultants had received many unsolicited letters from sub-contractors and suppliers of the appellant complaining that they had been underpaid by the appellant. They also requested the Employer to arrange for direct payment to be made to them for work done.[note: 7] Eventually, in view of these developments, a meeting was held on 16 November 2000 between the representatives of the appellant and respondent. At this meeting, the appellant’s managing director, Dr Chi Chao-Ton Tony (“Dr Chi”), repeated the request for the Contract SOT and the Contract CBD to be revised so as to alleviate the appellant’s cash-flow problems.[note: 8]

14        The respondent, on behalf of the Employer, expressed its willingness to consider the appellant’s request to adjust the Contract SOT and the Contract CBD subject to the following conditions:

(a)        KPK would evaluate the new SOT and CBD proposed by the appellant to try to ensure that adequate money was allocated for the uncompleted works; and

(b)        the appellant was to submit a performance bond, equivalent to the amount of money proposed to be reallocated from the uncompleted works, to safeguard the Employer’s interest.

At the meeting, the appellant agreed to the conditions stated above and KPK was tasked to evaluate the appellant’s proposed revision of the Contract SOT and the Contract CBD and to subsequently submit its recommendations to the respondent.

15        KPK completed its assessment on or about 22 January 2001 and a copy of the proposed revised SOT and CBD (“KPK-revised SOT and CBD”) was sent by KPK to the appellant by a letter dated 1 February 2001. By way of a letter dated 20 February 2001,[note: 9] the appellant agreed to adopt the KPK-revised SOT and CBD. Subsequently, by way of a letter dated 6 April 2001,[note: 10] RDC reminded the appellant of the discussion that had taken place at the meeting in November 2000 and requested it to propose the amount of bond money for the Employer’s consideration.

Abandonment of exercise to revise the Contract SOT and the Contract CBD

16        Later, at a meeting held on 10 April 2001 between the representatives of the appellant and the respondent, RDC said that if the KPK-revised SOT and CBD were implemented, an extra amount of $200,000 would have to be paid upfront to the appellant in the forthcoming progress payment. This represented a risk to the Employer and that was why a performance bond was necessary before the KPK-revised SOT and CBD could be implemented. As recorded in the minutes of the meeting,[note: 11] Dr Chi then informed the respondent and its representatives that the appellant had decided not to proceed with the implementation of the KPK-revised SOT and CBD because there was “insufficient incentive” to do so. As such, and at the appellant’s request, the exercise to revise the Contract SOT and the Contract CBD was abandoned. 

Novation of the Contract

17        During a meeting on 24 May 2001 and by way of a letter dated 25 May 2001, the appellant informed the respondent that it was considering the possibility of novating the Contract to another contractor. The Contract was eventually novated to Neo Corporation Pte Ltd (“NeoCorp”). The novation agreement between the appellant, the Employer and NeoCorp was signed on 16 July 2001.

18        At the time of the novation, the appellant had only completed about half of the Contract work. In the ordinary course of events, the appellant would have been paid the full lump sum at the end of the Project and would also have been entitled to submit to the Consultants for certification any claim for works not certified. If disputes arose, the appellant was entitled to submit these disputes for resolution in accordance with the arbitration clause (ie, cl 34 of the Contract). However, the appellant lost this right when it novated the Contract to NeoCorp in July 2001. This was the reason why the appellant had decided to pursue a claim against the respondent in tort to recover its alleged losses.

Appellant’s claim and trial judge’s findings

19        The appellant claimed against the respondent for negligence on the basis that the respondent owed it a duty of care to apply professional skill and judgment in certifying, in a fair and unbiased manner, payment for work carried out by the appellant to avoid causing it any loss due to undervaluation and under-certification of works. The appellant claimed that the respondent breached this duty by negligently undervaluing and under-certifying the appellant’s works, leading to the following losses:

(a)        $2,234,704.24, being the difference between the amount claimed by the appellant for works done on site, and the certified amount in progress payment No 23;

(b)        $290,102.25, being underpayment of variation orders;

(c)        $629,994.59, being loss of profit from the balance of works;

(d)        $124,827.02, being loss of management fees from mechanical and electrical variation orders; and

(e)        $530,324.08, being loss of interest.

20        In the event, the trial judge dismissed the appellant’s claim. The trial judge found that the respondent did not owe a duty of care to the appellant to avoid causing the latter to suffer pure economic loss through its (the respondent’s) negligent conduct: the Judgment at [92]. Notwithstanding this finding, the trial judge went on to consider the other “main issues that arose from the evidence adduced at the trial” and held that even if a duty of care had been owed, the evidence was “against the [appellant]” and “clearly disproved the breach of any such duty by the [respondent] to the [appellant]”: the Judgment at [92], [93] and [190].

Issues on appeal

21        It is trite law that, in order to succeed in a claim under the tort of negligence, a claimant has to establish that (a) the defendant owes the claimant a duty of care; (b) the defendant has breached that duty of care by acting (or omitting to act) below the standard of care required of it; (c) the defendant’s breach has caused the claimant damage; (d) the claimant’s losses arising from the defendant’s breach are not too remote; and (e) such losses can be adequately proved and quantified: see, for example, Clerk & Lindsell on Torts (Sweet & Maxwell, 19th Ed, 2006) at para 8-04.

22        Accordingly, the first issue which we had to consider was whether there was a duty of care owed by the respondent to the appellant in the manner we have outlined earlier. As we found against the appellant on this threshold issue, we did not have to consider the other issues of breach, causation and remoteness.

Trial judge’s decision in relation to the duty of care

23        In considering the duty of care issue, the trial judge applied both the two-stage test in Anns v Merton London Borough Council [1978] AC 728 (“Anns”) and the three-part test in Caparo Industries Plc v Dickman [1990] 2 AC 605 (“Caparo”) and reached the same result that there was no duty of care, an implicit judicial recognition of the state of confusion which belabours the local law in respect to the applicable test to be used to determine the existence of a duty of care for cases involving claims for pure economic loss (“cases of pure economic loss”) or, indeed, cases involving claims of personal injuries and physical damage (collectively “cases of physical damage”).

24        In coming to her decision, the trial judge placed considerable emphasis on the following factors: (a) there was no assumption of direct responsibility by the respondent to the appellant for the economic loss claimed; (b) there was no direct contractual relationship between the appellant as the main contractor and the respondent as the SO; (c) the appellant had recourse under the Contract to claim additional payments for variation works from the Employer; and (d) the property in question was an army camp, not a residential development: see the Judgment at [54].

25        Before we deal with the substantive arguments in relation to the duty of care, it is first necessary for us to determine the applicable test for ascertaining the existence of a duty of care.

Applicable test for duty of care

The parties’ arguments

26        Before us, the appellant submitted that the proper test to ascertain the existence of a duty of care for cases of pure economic loss should be a combination of the common elements from both the two-stage test in Anns and the three-part test in Caparo, citing the views contained in Andrew Phang, Saw Cheng Lim & Gary Chan, “Of Precedent, Theory and Practice – The Case for a Return to Anns” [2006] Sing JLS 1. Adopting this approach, the appellant submitted that the following issues arose for our consideration in relation to the existence of a duty of care in this case: (a) whether, in the circumstances of this case, there existed a relationship of sufficient proximity so as to give rise to a duty of care on the part of the respondent in respect of a pure economic loss; and (b) if so, whether there was any policy consideration that should nonetheless limit or negate the duty established under the first stage. However, the appellant also submitted that before these two issues could even be considered, the court should determine a preliminary issue first, viz, whether the damage suffered by the appellant was foreseeable by the respondent. We shall come back to this submission after we have addressed the current tests in the developed common law jurisdictions on the duty of care.

27        The current state of the law in Singapore on liability in negligence for pure economic loss, and indeed for physical damage, is not entirely satisfactory. Liability for pure economic loss (which is the subject matter of the appeal before us) is an issue of great complexity and one that imposes on the courts a great responsibility as this is an area of law where the determination of liability has been and will continue to be, unless the Legislature intervenes, the exclusive domain of the Judiciary. Judicial adjudication of claims for pure economic loss calls for caution and judiciousness in not extending the limits of liability for negligent acts to avoid any unforeseen economic cost to the public in the court’s desire to do justice to the claimant in the particular case. It is par excellence a policy decision for the court to make. To begin with, although the English courts have thought fit not to allow recovery of pure economic loss, the Singapore courts have taken a different path and have allowed recovery of such losses in a few decisions. Whether these decisions have laid down a general principle, that pure economic loss can be recovered in all cases of negligence, is a matter we wish to re-examine because of its implications for the redistribution of wealth in the economy. Furthermore, it is important that we revisit this issue as the Singapore courts have also applied a different test to determine the existence of a duty of care when the type of damages claimed is pure economic loss, and when it is for loss associated with physical damage, even though the foundation of the two different tests is derived from a common source, viz, the well-known English decisions in Anns and Caparo.

Preliminary observations

28        A few preliminary observations are apposite at this juncture. The common law is built on interconnected layers of principles, universal and particular, each dependent on and interacting with the other, held together by the overarching goal of fairness and justice. To understand this is to comprehend the common law’s eternal search for rational and universal principles in the law. Thus, in the area of negligence, the principles relating to the imposition of a duty of care must be rational and coherent and be reliably and consistently determined. With the universal in the form of general principles, the law will be a coherent body of rules, and not consist of singular instances of judicial decisions unrelated to each other. However, to have the universal without the particular is to lose sight of the need for practical solutions for the myriad of disputes that come before the courts. These general principles provide the focus to which the particular (viz, the facts of decided cases and the present case) must then be analysed so that the courts may, within this aggregation of the universal and the particular, reach a result that is fair and just. As Cooke P put it in South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd [1992] 2 NZLR 282 at 294:

A broad two-stage approach or any other approach is only a framework, a more or less methodical way of tackling a problem. However it is formulated should not matter in the end. Ultimately the exercise can only be a balancing one and the important object is that all relevant factors be weighed. There is no escape from the truth that, whatever formula be used, the outcome in a grey area case has to be determined by judicial judgment. Formulae can help to organise thinking but they cannot provide answers. [emphasis added]

Ultimately, the resolution of a claim in negligence should be based on general principles that have universal application to every factual situation which gives rise to a legal dispute either on liability or damages.

29        In the tort of negligence, careless conduct cannot, by itself, be used as a basis for tortious liability. It is only carelessness within the context of breach of a duty to take care vis-à-vis another person that there is liability in negligence, and hence the necessity to determine the circumstances in which the duty of care arises. In other words, even if duty sets the boundaries of liability (as Prof Tan Keng Feng put it in “The Three-Part Test: Yet Another Test of Duty in Negligence” (1989) 31 Mal L Rev 223 at 225), it also sets the limits on duty where liability is seen to be indeterminate. The two concepts of duty and liability are of course related but at the same time each affects the operation of the other. The difficulty is in balancing the goal of a fair and just result for the parties and the imposition of indeterminate liability on an indeterminate class of tortfeasors. That is the primary responsibility of the courts in developing the tort of negligence. How to do so without compromising on the usefulness of the tort of negligence as a tool for the redistribution of economic wealth fairly between tortfeasors and their victims is essentially the crucial issue for the courts. Unlike contract, where the rights and liabilities of the parties are primarily agreed or fixed between themselves, tortious liability occurs independently of the parties’ intentions and is hence liable to manifest itself in, potentially, an infinite number of situations, given that carelessness is a human frailty. This was what Cardozo CJ was afraid of in Ultramares Corp v Touche, Niven & Co 255 NY 170 (1931) at 179, ie, the imposition of “liability in an indeterminate amount for an indeterminate time to an indeterminate class”. Lord Buckmaster also alluded to this concern when he said in Donoghue v Stevenson [1932] AC 562 at 577, “If one step, why not fifty?”

30        The answer to this lies in legal control mechanisms for which the courts are primarily responsible. Lord Atkin, who may be regarded as the founder of the modern law of negligence, referred to this heavy responsibility in Donoghue v Stevenson (at 583–584):

[I]n the branch of the law which deals with civil wrongs, dependent in England at any rate entirely upon the application by judges of general principles also formulated by judges, it is of particular importance to guard against the danger of stating propositions of law in wider terms than is necessary, lest essential factors be omitted in the wider survey and the inherent adaptability of English law be unduly restricted. [emphasis added]

Indeed, the various tests which have been applied by the courts following Donoghue v Stevenson are but a manifestation of one or more of the legal control mechanisms formulated by the courts within the general principles. It is essential, at the outset, to note that although there are now many such mechanisms, they all serve the singular purpose of limiting liability, albeit in different ways. To put it another way, these legal control mechanisms ultimately seek to limit liability by resorting to the use of different conceptual tests, such as reasonable foreseeability, proximity or public policy.

The different tests in England

The neighbour principle and its development

31        Although there were some cases prior to 1932 which purported to articulate some kind of general principle to determine the existence of a duty of care (see, for example, Heaven v Pender (1883) 11 QBD 503 at 509, per Brett MR), the search for a universal basis for duty is generally regarded to have begun with the House of Lords decision of Donoghue v Stevenson. In that case, Lord Atkin famously said (at 580):

The rule that you are to love your neighbour becomes in law, you must not injure your neighbour; and the lawyer’s question, Who is my neighbour? receives a restricted reply. You must take reasonable care to avoid acts or omissions which you can reasonably foresee would be likely to injure your neighbour. Who, then, in law, is my neighbour? The answer seems to be – persons who are so closely and directly affected by my act that I ought reasonably to have them in contemplation as being so affected when I am directing my mind to the acts or omissions which are called in question.

This statement was then generally understood to mean that liability was founded on the test of pure foreseeability alone: see Hay or Bourhill v Young [1943] AC 92 at 107, where Lord Wright accepted Lord Atkin’s statement as establishing the “general concept of reasonable foresight as the criterion of negligence”.

32        However, it soon became apparent that a test based on foreseeability alone could be too wide and too pliable as a basis of liability in some situations. In order to limit or exclude what was deemed as unwarranted recovery, the courts frequently resorted to deciding artificially that certain claimants were “unforeseeable”: see Prof Tan ([29] supra) at 225. This led Lord Denning to consider foreseeability and the existence of a duty as simply a question of public policy. In Spartan Steel & Alloys Ltd v Martin & Co (Contractors) Ltd [1973] QB 27, Lord Denning MR said (at 36):

Whenever the courts draw a line to make out the bounds of duty, they do it as matter of policy so as to limit the responsibility of the defendant. [emphasis added].

Similarly, in Dorset Yacht Co Ltd v Home Office [1969] 2 QB 412, Lord Denning MR said (at 426):

[Duty] is, I think, at bottom a matter of public policy which we, as judges, must resolve. This talk of “duty” or “no duty” is simply a way of limiting the range of liability for negligence.

The two-stage test in Anns

33        The pendulum which swung between foreseeability and policy as the determinant of duty showed signs of resting in Arenson v Casson Beckman Rutley & Co [1977] AC 405, in which Lord Simon of Glaisdale appeared to endorse the idea that where Lord Atkin’s neighbour principle was seen as fixing the defendant with responsibility for the claimant’s damage, liability should follow, save where there was some good reason why it should not, and identified “public policy” as the qualifying factor (at 419). That pendulum finally became stationary in favour of an approach which meshed foreseeability with public policy in Anns ([23] supra), where Lord Wilberforce influentially said (at 751–752):

[I]n order to establish that a duty of care arises in a particular situation, it is not necessary to bring the facts of that situation within those of previous situations in which a duty of care has been held to exist. Rather the question has to be approached in two stages. First one has to ask whether, as between the alleged wrongdoer and the person who has suffered damage there is a sufficient relationship of proximity or neighbourhood such that, in the reasonable contemplation of the former, carelessness on his part may be likely to cause damage to the latter – in which case a prima facie duty of care arises. Secondly, if the first question is answered affirmatively, it is necessary to consider whether there are any considerations which ought to negative, or to reduce or limit the scope of the duty or the class of persons to whom it is owed or the damages to which breach of it may give rise … [emphasis added]

34        It is well known that there is some controversy regarding the meaning of the first stage of the test: Did it relate only to the factual issue of reasonable foreseeability or did it introduce the legal conception of proximity (see further Phang, Saw & Chan ([26] supra) at 40–41)? We will revisit this problem later (see [74] below). For the moment, it suffices to note that this pronouncement in Anns is not only significant for the elucidation of a two-stage test to determine duty, it also introduced the concept of a truly “universal” test in that it could be applied without reference to previous decisions where the duty might or might not have been seen to have existed. This attracted considerable criticisms from many other judges who thought that the much sought-after universal basis for duty is unattainable and even undesirable, and that an incremental approach based on previous decisions on a case by case basis was less subjective and therefore preferable.

35        In our view, these criticisms have arisen because of the perceived divorce of the particular from the universal (see [28] above). Whether or not Lord Wilberforce intended for his test to apply in a vacuum, distanced from all considerations of prior decisions, is not exactly clear, since the essence of the common law (which Lord Wilberforce must have known) is by its very nature incremental. A fair reading of the sentence (see [33] above) beginning with “in order to establish … it is not necessary to bring the facts within those of previous situations” does not preclude an incremental approach as the expression “the facts” does necessarily connote all the facts, and is capable of implying that it is only necessary to bring only some of the facts. However, as the formula was perceived as such, it led to the concern of a potentially uncontrollable expansion of liability.

The two-stage test in Anns doubted and overruled

36        This concern led to a check of the advance of the two-stage test in Anns as doctrine. In Governors of the Peabody Donation Fund v Sir Lindsay Parkinson & Co Ltd [1985] AC 210, after quoting Lord Wilberforce’s test, Lord Keith of Kinkel made the criticism (at 240) that:

There has been a tendency in some recent cases to treat these passages as being themselves of a definitive character. This is a temptation which should be resisted.

Similarly, the Australian High Court gave reasons in Sutherland Shire Council v Heyman (1985) 60 ALR 1 (“Sutherland”) for declining to follow the direction in which Anns appeared to be taking the law in England. Brennan J stated (at 43–44) that it was preferable that:

[T]he law should develop novel categories of negligence incrementally and by analogy with established categories, rather than by a massive extension of a prima facie duty of care restrained only by indefinable “considerations which ought to negative, or to reduce or limit the scope of the duty or the class of person to whom it is owed”.

37        Following this, Lord Brandon of Oakbrook remarked in Leigh and Sillavan Ltd v Aliakmon Shipping Co Ltd [1986] AC 785 at 815 that the propositions laid down by Lord Wilberforce in Anns “[do] not provide, and cannot … have been intended by Lord Wilberforce to provide, a universally applicable test of the existence and scope of a duty of care in the law of negligence”. The Privy Council in Yuen Kun Yeu v Attorney-General of Hong Kong [1988] AC 175 (“Yuen Kun Yeu”) further expressed the opinion that whether or not a duty of care in negligence existed would depend primarily upon foreseeability of damage, together with the existence of a close and direct relationship or proximity between the parties, and that, occasionally, it would be necessary to go on to consider whether public policy required that liability should not attach. Lord Keith of Kinkel observed in Yuen Kun Yeu (at 191) that the two-stage test in Anns “has been elevated to a degree of importance greater than it merits, and greater perhaps than its author intended”.

38        The two-stage test in Anns was finally overruled by the House of Lords in Murphy v Brentwood District Council [1991] 1 AC 398 (“Murphy”). In unanimously holding that the defendant council owed no duty of care to the claimant (a subsequent home owner) in respect of pure economic loss, Lord Keith of Kinkel observed thus (at 471):

In my opinion it is clear that Anns did not proceed upon any basis of established principle, but introduced a new species of liability governed by a principle indeterminate in character but having the potentiality of covering a wide range of situations, involving chattels as well as real property, in which it had never hitherto been thought that the law of negligence had any proper place.

He continued (at 471–472):

In my opinion there can be no doubt that Anns has for long been widely regarded as an unsatisfactory decision. In relation to the scope of the duty owed by a local authority it proceeded upon what must, with due respect to its source, be regarded as a somewhat superficial examination of principle and there has been extreme difficulty, highlighted most recently by the speeches in [D & F Estates Ltd v Church Commissioners for England [1989] AC 177], in ascertaining upon exactly what basis of principle it did proceed. I think it must now be recognised that it did not proceed on any basis of principle at all, but constituted a remarkable example of judicial legislation. It has engendered a vast spate of litigation, and each of the cases in the field which have reached this House has been distinguished. Others have been distinguished in the Court of Appeal. The result has been to keep the effect of the decision within reasonable bounds, but that has been achieved only by applying strictly the words of Lord Wilberforce and by refusing to accept the logical implications of the decision itself. These logical implications show that the case properly considered has potentiality for collision with long-established principles regarding liability in the tort of negligence for economic loss. There can be no doubt that to depart from the decision would re-establish a degree of certainty in this field of law which it has done a remarkable amount to upset.

The three-part test in Caparo

39        With the fall of the two-stage test in Anns, another approach to determine the existence of a duty of care emerged. In Smith v Eric S Bush [1990] 1 AC 831, Lord Griffiths, in dealing with the question of when the law deems a person who has given advice to have assumed responsibility to the recipient who has acted on it and suffered damage, replied (at 865):

I would answer – only if it is foreseeable that if the advice is negligent the recipient is likely to suffer damage, that there is a sufficiently proximate relationship between the parties and that it is just and reasonable to impose the liability.

This trend towards a three-part test was summarised by Lord Bridge of Harwich in Caparo ([23] supra) thus (at 617–618):

What emerges is that, in addition to the foreseeability of damage, necessary ingredients in any situation giving rise to a duty of care are that there should exist between the party owing the duty and the party to whom it is owed a relationship characterised by the law as one of “proximity” or “neighbourhood” and that the situation should be one in which the court considers it fair, just and reasonable that the law should impose a duty of a given scope upon the one party for the benefit of the other.

40        The three-part test in Caparo is, by no means, unambiguous in its application; indeed, the very judges who had formulated it expressed their belief that it was unlikely to be a workable test. Lord Bridge himself cautioned (at 618) that:

[T]he concepts of proximity and fairness embodied in these additional ingredients are not susceptible of any such precise definition as would be necessary to give them utility as practical tests, but amount in effect to little more than convenient labels to attach to the features of different specific situations which, on a detailed examination of all the circumstances, the law recognises pragmatically as giving rise to a duty of care of a given scope.

Similarly, Lord Roskill gave a warning (at 628) that:

Phrases such as “foreseeability,” “proximity,” “neighbourhood,” ”just and reasonable,” “fairness,” … will be found used from time to time in the different cases. But … such phrases are not precise definitions. At best they are but labels or phrases descriptive of the very different factual situations which can exist in particular cases and which must be carefully examined in each case before it can be pragmatically determined whether a duty of care exists and, if so, what is the scope and extent of that duty.

In other words, some overlapping or merging of the three criteria will occur according to the particular circumstances: see also Steyn LJ’s comments in Elguzouli-Daf v Commission of Police of the Metropolis [1995] QB 335 at 349. In a similar vein, in Marc Rich & Co AG v Bishop Rock Marine Co Ltd [1994] 1 WLR 1071, Balcombe LJ said (at 1088) that he doubted whether the words “fair, just and reasonable” imposed a test additional to that of “proximity”. He went on to state that “these are criteria to be adopted in considering whether the necessary degree of proximity exists”.

The incremental approach

41        Additionally, it ought to be noted that the House of Lords in Caparo also endorsed the incremental approach advocated by Brennan J in the Australian High Court decision of Sutherland ([36] supra). Lord Bridge observed (at 618) that:

Whilst recognising, of course, the importance of the underlying general principles common to the whole field of negligence, I think the law has now moved in the direction of attaching greater significance to the more traditional categorisation of distinct and recognisable situations as guides to the existence, the scope and the limits of the varied duties of care which the law imposes.

42        Although the English courts have accepted incrementalism as a useful guide, Prof Keith Stanton has noted in “Incremental Approaches to the Duty of Care” (in Torts in the Nineties (Nicholas J Mullany ed) (LBC Information Services, 1997) at p 44), that no consistent pattern emerges to indicate what the courts understand by this term. Prof Stanton is of the view (at p 39) that, at a general level, “incrementalism may be regarded as restoring a traditional method of adjudication to the tort of negligence” in that it remains intertwined with a formula-based approach, and that the general principle should henceforth be applied in a more cautious manner. On the other hand, at a more detailed level, incrementalism can be taken to mean the development of “pockets of duty of care”, modified and developed by the slow incremental accretion of case law, in lieu of a general principle which provides the basis of analysis. Indeed, the three-part test in Caparo stands in apparent distinction from the incremental approach which was, somewhat curiously, advocated in the same case.

43        If incrementalism is intended to return the development of liability in negligence to the basis that discrete sets of rules should govern different factual scenarios, the central problem, as Prof Stanton noted at p 48, is that it requires a return to an approach which has been systematically destroyed since Donoghue v Stevenson, as generations of lawyers have since been trained to believe that the tort of negligence is based on a common general principle. In this respect, we agree with Lord Bingham of Cornhill in Customs and Excise Commissioners v Barclays Bank plc [2007] 1 AC 181 (“Customs and Excise Commissioners”), where he observed (at [7]) that the “incremental test is of little value as a test in itself, and is only helpful when used in combination with a test or principle which identifies the legally significant features of a situation”. In our view, incrementalism is not a distinct and alternative method which results in a plethora of discrete sets of rules (as opposed to a set of general (and universal) principles). Instead, it is an analogical process to apply specific facts to the general principle. According to this approach, a court should, when applying the specific step of the general principle, utilise the incremental methodology to decide first whether that specific step (or criterion) can be satisfied with reference to decided cases. This will provide an essential check to any unwarranted expansion of liability for negligent acts. However, as we will further observe (at [72] below), where there are no decided cases, recourse to general principles is not only valid but also desirable. Such an approach, an interaction of the universal with the particular (see [28] above), would, in our view, aid in the development of a rational and purposive concept of negligence which can achieve fairness and justice in each case.

Other tests

44        The cases referred to above may be regarded as the main current in the search for a test for the existence of a duty of care in novel situations. But there have been other side-currents, particularly in relation to liability for psychiatric harm and negligent misstatements causing pure economic loss. A line of authority from Hedley Byrne & Co Ltd v Heller & Partners Ltd [1964] AC 465 (“Hedley Byrne”) has grounded liability for the latter on an assumption of responsibility towards the recipient of the statement in question and reliance by him on its accuracy.

45        As for acts or omissions causing pure economic loss, English law currently prevents recovery for such claims. In Murphy ([38] supra), the House of Lords adopted what is known as the “exclusionary rule” in relation to cases involving pure economic loss flowing from careless acts or omissions. Lord Oliver of Aylmerton said (at 487) that if the infliction of loss by the defendant on the claimant was to be categorised as wrongful, it was necessary to find some factor beyond the mere occurrence of the loss and the fact that its occurrence could be foreseen. What was necessary was evidently a relationship of proximity so close that it was “akin to contract” (per Lord Bridge of Harwich in Murphy at 481) to introduce the element of reliance such that the scope of the duty of care owed by the defendant to the claimant was wide enough to embrace pure economic loss. Lord Oliver further noted that the categorisation of damage as economic served at least the useful purpose of indicating that something more was required. However, it has long been recognised that the effect of Murphy is simply that English law will deny the existence of a duty of care to any individual suffering pure economic loss caused by careless acts or omissions where there is neither personal injury nor property damage: see Charlesworth & Percy on Negligence (Sweet & Maxwell, 11th Ed, 2006) at para 2-134. The reason for this appears to be a fear of indeterminate liability owing to the fluid nature of pure economic losses which are not strictly restricted by the physical interaction of the parties involved (see further [66] below).

Current position in England

46        The search for clarity and certainty in a single formula has proved to be elusive even some 60 years after the first articulation of the “neighbour” principle in Donoghue v Stevenson. Lord Oliver of Aylmerton lamented in Caparo ([23] supra at 633) thus:

… I think that it has to be recognised that to search for any single formula which will serve as a general test of liability is to pursue a will-o’-the wisp. The fact is that once one discards, as it is now clear that one must, the concept of foreseeability of harm as the single exclusive test – even a prima facie test – of the existence of the duty of care, the attempt to state some general principle which will determine liability in an infinite variety of circumstances serves not to clarify the law but merely to bedevil its development in a way which corresponds with practicality and common sense.

47        In Customs and Excise Commissioners v Barclays Bank plc [2005] 1 WLR 2082, Longmore LJ identified (at [23]) three separate and independent tests in England to find a duty of care:

The modern law of negligence derives primarily from four decisions of the House of Lords: Caparo Industries plc v Dickman [1990] 2 AC 605; Henderson v Merrett Syndicates Ltd [1995] 2 AC 145; White v Jones [1995] 2 AC 207 andPhelps v Hillingdon London Borough Council [2001] 2 AC 619. Caparo laid down the threefold test of foreseeability, proximity and fairness and emphasised the desirability of incremental development: see in particular Lord Bridge of Harwich, at pp 617–618. Henderson and White both emphasised the concept of assumption of responsibility in cases where the negligent performance of services had caused economic loss. In the light of these authorities this court has held (see Bank of Credit and Commerce International (Overseas) Ltd v Price Waterhouse [1998] BCC 617) that an appropriate course for ascertaining whether there is a duty of care, at least in an economic loss case, is to look at any new set of facts by using each of the three approaches in turn, viz, the threefold test, the voluntary assumption of responsibility test and the incremental test – “if the facts are properly analysed and the policy considerations are correctly evaluated the several approaches will yield the same result”: per Sir Brian Neill, at p 634.

48        The judicial response of Longmore LJ that has led to the establishment of no less than three separate tests for duty, in our view, puts paid to the search for a general principle applicable to all situations where tortious liability may arise. Damage arising from different situations (viz, pure economic loss or physical damage) is seen to require the application of different tests, and even within the rubric of physical damage, there is some confusion as regards which of the three tests is applicable to determine a duty of care.

49        On appeal in Customs and Excise Commissioners ([43] supra), the House of Lords did not express a definitive view on whether or not the three “tests” mentioned by Longmore LJ were substantive tests in their own right and were all to be applied in any given case. Indeed, the general position adopted by the House itself is, with respect, none too clear (see also Steven Gee, “The Remedies Carried by a Freezing Injunction” (2006) 122 LQR 535, especially at 536). Thus, as Andrew Phang Boon Leong J (as he then was) observed in Sunny Metal & Engineering Pte Ltd v Ng Khim Ming Eric [2007] 1 SLR 853 (“Sunny Metal”) at [79], the position in England, even at the present, continues to be in a “state of flux and confusion”. This situation does not make it easier for the Singapore courts, but it does present them with an opportunity to revisit the subject and to try to fashion a general test that is principled and logical and capable of practical application to all cases where damage is caused by a negligent act. In this connection, it should be borne in mind that all such problems that come before the courts are evaluated and assessed on a retrospective and not a prospective basis; the damage has arisen and the only question is whether there is liability for it. This condition should, in most situations, give the courts an advantage in being able to see whether, on the facts of the case, the application of the test is likely to lead to an unmanageable expansion of liability for negligent acts or omissions that may be detrimental to the national welfare.

The different tests in Singapore

50        We turn now to examine the current position in Singapore. The decisions have, as with the case in England, applied different tests to different types of damages claimed. For convenience, we will deal first with cases of pure economic loss and then cases of physical damage.

“Two-stage process” for cases of pure economic loss

51        As we have stated earlier, the courts in Singapore have in certain cases declined to apply the “exclusionary rule” adopted by the House of Lords in Murphy ([38] supra) and have held that pure economic loss is recoverable in tort: see RSP Architects Planners & Engineers v Ocean Front Pte Ltd [1996] 1 SLR 113 (“Ocean Front”) which concerned the faulty construction of common property in a condominium. The rationale for this departure from the English position was explained in the subsequent case of RSP Architects Planners & Engineers v MCST Plan No 1075 [1999] 2 SLR 449 (“Eastern Lagoon”), where this court stated (at [43]) that there were cogent policy reasons to allow for claims of pure economic loss in relation to immovable property since land was scarce in Singapore and the financial outlay in such properties represented a significant investment in a person’s lifetime.

52        In Ocean Front, the management corporation (“MC”) of Bayshore Park Condominium took out a writ against the developer of that condominium for damages arising out of the faulty construction of common property. This court applied the two-stage test in Anns ([23] supra) and relied on Junior Books Ltd v Veitchi Co Ltd [1983] 1 AC 520 (“Junior Books”) and found the developer liable in negligence for economic loss caused to the MC with whom the developer had no contractual relationship. The approach which this court took was that it first considered whether there was (on the facts of the case) sufficient proximity between the parties, which would give rise to a duty of care. L P Thean JA, who delivered the judgment of the court, observed (at 139, [69]):

But the approach of the court has been to examine a particular circumstance to determine whether there exists that degree of proximity between the plaintiff and the defendant as would give rise to a duty of care by the latter to the former with respect to the damage sustained by the former. Such proximity is the ‘determinant’ of the duty of care and also the scope of such duty. [emphasis added]

53        This court found that there was a duty of care from the combination of several factors evidencing proximity:

(a)        the developer developed the condominium (which included the common property as part of it) and undertook that the condominium would be constructed in a good and workmanlike manner;

(b)        after completion of the condominium, the developer remained the legal owner of the common property and was responsible for its maintenance and upkeep;

(c)        the developer incorporated the MC to hold and to maintain the common property; and

(d)        the developer was obviously aware that if the common property was defective as a result of defective workmanship, any damage to it would have to be made good by the MC.

The relationship between the developer and the MC was as close as could be short of actual privity of contract.

54        Having found that there was proximity on these bases, this court then proceeded “to consider whether there is any policy consideration in negativing such [a] duty of care” (at 142, [75]). It found no policy reason negativing the duty of care primarily because the ambit of liability was determinate. Indeed, in our view, for the very reasons given by the court in deciding that there was proximity in that case, the policy considerations would be in favour of imposing liability on the developer.

55        Ocean Front was followed by Eastern Lagoon ([51] supra), in which this court revisited its holdings in Ocean Front. This court expressly rejected the “two-stage test” in Anns but adopted what was substantively a similar test. L P Thean JA, who delivered the judgment of the court, observed thus (at [18]):

It seems to us that what is objectionable in that passage [containing Lord Wilberforce’s test in Anns] is firstly his Lordship’s sweeping proposition of a single general rule or principle which can be applied in every situation to determine whether a duty of care arises and secondly the fact that the test propounded by his Lordship in the first stage was based on foreseeability of damage alone.

And later, at [29]–[30]:

It is abundantly clear that in Ocean Front this court did not follow the broad proposition laid down by Lord Wilberforce in Anns. True, the court reached its conclusion by a two-stage process. In principle, there is no objection to such approach. It depends on what is involved and considered in each stage. The court certainly did not apply the first test in Anns. The court’s finding that there was sufficient degree of proximity giving rise to a duty on the part of the developers to avoid the loss sustained by the management corporation was not premised on foreseeability of damage alone, but on the consideration of other relevant facts. Nor did the court accept Lord Wilberforce’s proposition that in any given situation a single general rule or principle can be applied to determine whether a duty of care arises.

It does not follow from the mere fact that the court in the course of their determination examined the facts by the two-stage process that the court in effect followed Anns.

[emphasis added]

56        As to what this “two-stage process” entailed, Thean JA continued at [31]:

Stripped of the verbiage, the crux of such approach [ie, the “two-stage process”] is no more than this: the court first examines and considers the facts and factors to determine whether there is sufficient degree of proximity in the relationship between the party who has sustained the loss and the party who is said to have caused the loss which would give rise to a duty of care on the part of the latter to avoid the kind of loss sustained by the former. … Next, having found such degree of proximity, the court next considers whether there is any material factor or policy which precludes such duty from arising.

57        That the “two-stage process” was only applicable to cases of pure economic loss was confirmed in The Sunrise Crane [2004] 4 SLR 715. In that case, the majority (at [31]–[37]) referred to Man B&W Diesel S E Asia Pte Ltd v PT Bumi International Tankers [2004] 2 SLR 300 (“Man B&W Diesel”) and Ocean Front and held that the approach used in those cases (ie, the “two-stage process”) was limited to pure economic loss situations and thus was not applicable to The Sunrise Crane, which concerned physical damage. The majority therefore applied the three-part test in Caparo ([23] supra) to the case at hand. However, the dissenting judge in The Sunrise Crane took the view that a uniform approach should be applied to all cases of negligence, with the salient features of each case being determinative of the existence and scope of duty.

58        The result of the majority’s pronouncement in The Sunrise Crane, as Dr Kumaralingam Amirthalingam notes in “The Sunrise Crane – Shedding New Light or Casting Old Shadows on Duty of Care?” [2004] Sing JLS 551 at 555, is that in Singapore, the “two-stage process” in Ocean Front is to be applied in cases of pure economic loss, while the three-part test in Caparo is to be applied in all other situations (ie, cases of physical damage).

59        Subsequently, in United Project Consultants Pte Ltd v Leong Kwok Onn [2005] 4 SLR 214 (“United Project Consultants”) at [27], this court held that, in determining whether to impose a duty of care in cases of pure economic loss (which was the case in United Project Consultants), the courts have “consistently adopted a more restrictive approach” [emphasis added] and applied the “two-stage process” in Ocean Front. After endorsing Thean JA’s statement of the “two-stage process” in Eastern Lagoon (see [56] above), this court went on to state (at [35]) that:

In essence, before liability may be imposed upon a defendant for pure economic loss, a court must be satisfied that all the circumstances of the case give rise to a relationship whereby the defendant owes a duty to the plaintiff to avoid the particular loss suffered by the plaintiff. In doing so, the court must likewise be satisfied that there are no policy reasons why such a duty ought not to be imposed.

However, it then expressly went on to clarify (at [36]) that the “restatement of the principle in [Eastern Lagoon] should not be construed as reverting to the two-stage test in Anns” [emphasis added].

60        At first glance, the “two-stage process” is substantively similar to the two-stage test in Anns; indeed, it even involves the same elements as stated by Lord Wilberforce in Anns: see also Man B&W Diesel at [29]. On another view, however, the “two-stage process” is a “hybrid or composite test” in that it is not wholly “universal” in nature, since the court in Ocean Front pointed out that the facts of that case were similar to Junior Books ([52] supra) and Bryan v Maloney (1995) 69 ALJR 375, ostensibly using both these cases as precedents from which to advance the imposition of a duty of care in that case, a process not dissimilar to the incremental approach alluded to in Caparo: see Debbie Ong Siew Ling, “The Test of Duty for Defects in Property Causing Pure Economic Loss” [1999] Sing JLS 667 at 670–671.

61        In our view, if one were to construe Lord Wilberforce’s words in Anns literally (see [33] above), such that there was no longer a need to draw analogies with decided cases, it would be true that the “two-step process” does constitute a “hybrid or composite” test. However, as we stated earlier (see [43] above), the common law methodology is essentially incremental in nature, drawing its conclusions from precedents, and this necessarily presupposes that even within the confines of a general principle, there remains the scope for the application of an analogising or incremental approach, and the two must not be viewed as being mutually exclusive. As such, it may be better to view this court’s characterisation in that case of the “two-stage process” as being wholly unrelated to Lord Wilberforce’s two-stage test in Anns and as a reflection of the reluctance of the local courts at that time to follow an English authority which had been discredited. It would therefore have been better had this court in Eastern Lagoon simply acknowledged that it was, in substance and effect, applying the two-stage test in Anns.

Three-part test for cases of physical damage

62        We turn now to the decisions applying the three-part test in Caparo. As we mentioned earlier, for cases not involving pure economic loss, it is the three-part test in Caparo which has also been endorsed in the Singapore context: see, for example, Ikumene Singapore Pte Ltd v Leong Chee Leng [1993] 3 SLR 24 (“Ikumene”); Standard Chartered Bank v Coopers & Lybrand [1993] 3 SLR 712; Pang Koi Fa v Lim Djoe Phing [1993] 3 SLR 317; Mohd bin Sapri v Soil-Build (Pte) Ltd [1996] 2 SLR 505; D v Kong Sim Guan [2003] 3 SLR 146; TV Media Pte Ltd v De Cruz Andrea Heidi [2004] 3 SLR 543 (“TV Media Pte Ltd”); and The Sunrise Crane ([57] supra).

63        It is clear that the courts take the view that the three-part test in Caparo involves an incremental approach. In Ikumene, this court reiterated (at 29, [18]) what Brennan J said in the Australian High Court in Sutherland ([36] supra) at 43–44 (see [36] above). However, for reasons we have mentioned earlier, our courts have not made clear the relationship between the three-part test in Caparo and the incremental approach. We repeat our view that the application of a general principle does not exclude an incremental approach; the latter is a method by which the general principle is to be applied. The universal and the particular interact to produce a result that is fair and just.

Overview of the problems

64        The search for the applicable test is not restricted to preferring one test over the other; a related question is whether there should be different tests to determine the existence of a duty of care for cases of pure economic loss and cases of physical damage. Why should there be two different tests, and what difference does it make, since the final result the courts wish to achieve is the same, viz, that which is fair and just between the parties without imposing an unacceptable economic cost to the public. We now consider whether a different or modified approach should apply to determine the existence of a duty of care in cases of pure economic loss.

A different approach to determine duty of care in cases of pure economic loss?

Rationale for different approach

65        As stated earlier (see [45] above), pure economic loss arising from acts or omissions is not recoverable in England, although losses arising from negligent misstatements are recoverable, provided that there exists a special relationship between the parties concerned. In Singapore, while pure economic loss is recoverable, the courts have advocated a more restricted approach in its recovery, to the extent of purporting to apply a different test (ie, the “two-stage process” in Ocean Front) to supposedly reflect this different treatment of such losses.

66        A possible explanation for the courts’ adoption of a more restrictive approach could be the express reference made by this court in The Sunrise Crane to a passage from Clerk and Lindsell on Torts (Sweet & Maxwell, 18th Ed, 2000) which sets out a rationale for the difference between pure economic loss and losses associated with physical damage. This passage is as follows (at para 7-85):

Three features distinguish economic loss claims from most physical damage claims. First, whilst the links between negligence and physical damage depend largely on the laws of nature and necessarily limit the type of relationship giving rise to a claim, those between negligence and pure economic loss are primarily human in creation and can form a complex web through which economic losses can ripple out from the one negligent act. Secondly, because the economic relationships are frequently created rather than imposed, the participants in the web have a greater opportunity to use contracts to determine the level of risk to be taken and the degree of protection from loss required. … Thirdly, as Hobhouse L.J. has observed, “in a competitive economic society the conduct of one person is always liable to have economic consequences for another and, in principle, economic activity does not have to have regard to the interests of others and is justifiable by the actor having regard to his own interests alone”. Furthermore, as McHugh J. has noted, “pure economic losses frequently result in mere transfers of wealth. The plaintiff’s loss is the defendant’s or a third party’s gain”, whereas “harm to a person or property ordinarily involves a new loss to social wealth.” These features have led the courts to take a more restrictive approach to the imposition of a duty of care in relation to economic loss than in relation to physical damage.

Is a different approach justified?

67        In relation to the reasons traditionally given as justification for a different approach, Robby Bernstein points out in Economic Loss (Sweet & Maxwell, 2nd Ed, 1998) (at p 21) that it is important to note that it is not anything inherent in the type of the damage itself that necessitates a different approach:

Rather, it is the circumstances in which the economic loss has arisen which sound an alarm that something more is required to warrant the imposition of a duty of care on the defendant for the plaintiff’s loss than the mere fact that it was reasonably foreseeable that the defendant’s negligent act or omission would cause that loss to the plaintiff. [emphasis added]

In the case of Leigh and Sillivan Ltd v Aliakmon Shipping Co Ltd [1985] QB 350, Oliver LJ (as he then was) recognised this point when he said (at 375):

The premise appears to be that the reason why, in the older authorities, the plaintiff failed to recover damages was that the damage which he suffered was “economic” … Speaking for myself, I have not felt able to accept that premise, for as I read those cases it was not the “economic” quality of the damage which prevented recovery but the reason why that damage had occurred. [emphasis added]

Similarly, he, as Lord Oliver in Murphy ([38] supra), reiterated this point (at 485):

[I]t is far from clear from these decisions [ie, the various economic loss cases which he had just referred to] that the reason for the plaintiff’s failure was simply that the only loss sustained was “economic.” Rather they seem to have been based either upon the remoteness of the damage as a matter of direct causation or, more probably, upon the “floodgates” argument of the impossibility of containing liability within any acceptable bounds if the law were to permit such claims to succeed.

68        Unfortunately, as Bernstein notes in Economic Loss (at p 21), the House of Lords in Murphy failed to realise that if the concerns articulated above do not in fact exist, then there is no need for the law to adopt an unduly restrictive approach towards the imposition of duties of care for all cases involving pure economic loss. To do so based only on the type of damages sustained (ie, economic loss as opposed to physical damage) would be to misunderstand the foundational principle upon which the restrictive principle towards cases of pure economic loss was conceptualised in the first place. There will be cases of pure economic loss in which the concern of indeterminate liability does not arise. Each case must be looked at on its own facts. Indeed, in the Australian High Court decision of Sutherland ([36] supra), Mason J (as he then was) said (at 32):

The proposition that, in general, damages are not recoverable for economic loss unless it is consequential upon injury to the plaintiff’s personal property is by no means absolute or inflexible; it is a reflection of the law’s concern about endless indeterminate liability. In the absence of any such concern in a particular class of case, there is no necessity to give effect to the proposition. [emphasis added]

As Mason CJ, he reiterated this point in Bryan v Maloney ([60] supra). Mason CJ, when delivering the majority judgment (at 380), said:

[T]he policy considerations underlying the reluctance of the courts to recognise a relationship of proximity and a consequent duty of care in cases of mere economic loss are inapplicable to a relationship of the kind which existed between [the parties in this case] … [T]here is no basis for thinking that recognition of a relationship of proximity between builder and first owner with respect to that particular kind of economic loss would give rise to the type of liability “in an indeterminate amount for an indeterminate time to an indeterminate class” which the courts are reluctant to recognise.

69        We respectfully agree that there is no justification for a general exclusionary rule against recovery of all economic losses and indeed, this is already the position the Singapore courts have taken, following Ocean Front. In this connection, we would also note that the Singapore cases which have allowed claims for pure economic losses have all been related to the economic value of the land. Indeed, for the condominium cases, as we have considered above at [51], the common property was intended to be for the enjoyment of the condominium owners and, by purchasing the condominium units, each owner was also effectively purchasing a share of the common property, as contemplated by the statutory regime on strata titles. For this reason, the MC may be regarded as the alter ego of the owners of the condominium units. Without the artificial but necessary interposition of the MC, there would have been a direct contractual relationship between the developers and the condominium owners. This very close relationship between the developers and the MC alleviates any fear of indeterminate liability owing to the nature of pure economic losses. Although the Singapore decisions on pure economic loss have been largely restricted to such situations, we are of the view that there is no reason not to extend liability for pure economic loss to other situations, provided that the issues of indeterminate liability and policy can be adequately dealt with. In this connection, the incremental approach would provide a necessary safeguard against the unintended consequence of indeterminate liability as well as discourage arbitrariness in determining liability.

70        However, the question then is whether we should continue to adopt a different test for cases of pure economic loss, as distinguished from cases of physical damage. In our opinion, quite apart from the doctrinal untidiness of having two applicable tests, depending on the nature of the damage (ie, the “two-stage process” in Ocean Front for cases of pure economic loss and the three-stage test in Caparo for cases of physical damage), this approach does not, in fact, address the concerns which prompted the more restrictive approach for some, but not all, cases of pure economic loss.

71        As such, in our view, a single test is preferable in order to determine the imposition of a duty of care in all claims arising out of negligence, irrespective of the type of the damages claimed, and this should include claims for pure economic loss, whether they arise from negligent misstatements or acts/omissions. In cases of physical damage, there is usually no difficulty with regard to a control mechanism to prevent indeterminate liability. However, the adoption of a single test would serve to constrain liability even in those extremely rare cases where physical damage might possibly result in indeterminate liability. It may well be that there are policy considerations in restricting recovery for pure economic loss in certain situations, but this in itself does not make it necessary for a wholly different approach in the form of a separate test altogether.

72        Ultimately, a single test to determine the existence of a duty of care for all claims of negligence would do well to eliminate the perception that there are, at once, two or more tests which are equally applicable. While it may be that all these tests could yield the same result, their serial applicability diminishes the desirability of having a general principle that can provide a coherent, consistent and reliable way of determining or recognising a duty of care. We now turn our attention to considering what the single applicable test may be.

The applicable test in Singapore

73        In our view, a coherent and workable test can be fashioned out of the basic two-stage test premised on proximity and policy considerations, if its application is preceded by a preliminary requirement of factual foreseeability. We would add that this test is to be applied incrementally, in the sense that when applying the test in each stage, it would be desirable to refer to decided cases in analogous situations to see how the courts have reached their conclusions in terms of proximity and/or policy. As is obvious, the existence of analogous precedents, which determines the current limits of liability, would make it easier for the later court to determine whether or not to extend its limits. However, the absence of a factual precedent, which implies the presence of a novel situation, should not preclude the court from extending liability where it is just and fair to do so, taking into account the relevant policy consideration against indeterminate liability against a tortfeasor. We would admit at this juncture that this is basically a restatement of the two-stage test in Anns, tempered by the preliminary requirement of factual foreseeability. Indeed, we should point out that this is the test applied in substance by many jurisdictions in the Commonwealth: see, for example, the Canadian case of Cooper v Hobart (2001) 206 DLR (4th) 193; the New Zealand case of Rolls-Royce New Zealand Ltd v Carter Holt Harvey Ltd [2005] 1 NZLR 324; and the general summary provided in Phang, Saw & Chan ([26] supra) at 9–28.

74        We now set out in detail the various elements in the approach outlined above. We will also illustrate how this approach is to be applied practically when we apply it to the facts of the present case.

Factual foreseeability

75        As we mentioned earlier (see [34] above), there is some controversy in relation to the interpretation of the first stage of Lord Wilberforce’s test in Anns, ie, whether it refers to mere factual foreseeability or legal foreseeability or proximity. In our view, factual foreseeability is too wide a criterion to be effective as a legal control mechanism if all that it means is that the defendant ought to have known that the claimant would suffer damage from his (the defendant’s) carelessness. If this is the approach to be adopted, it would be fulfilled in almost all cases, because the two parties are likely to be in some degree of physical relationship in the relevant case. As Phang J said in Sunny Metal ([49] supra) at [55]:

The first part of the “three-part test” (relating to “foreseeability”) is, in effect, a reference to the factual conception of reasonable foreseeability or proximity. It is, in other words, what I have earlier termed … as “reasonable foreseeability”. Although this consideration has been incorporated as an element within the “three-part test” itself, its incorporation is, with respect, unnecessary. As I have already pointed out above … the requirement of reasonable foreseeability from a factual perspective will almost always be satisfied, simply because of its very nature and the very wide nature of the “net” it necessarily casts. There is therefore no practical need to include such a factual element within a legal formulation such as exists with regard to the existence (or otherwise) of a duty of care. … This is not to state that this last-mentioned (factual) element is to be discarded. As I have already mentioned, it will almost always be satisfied. [emphasis in original]

76        However, that is not to say that it is not a necessary element in any claim in negligence, just that it is a threshold question which the court must be satisfied is fulfilled, failing which the claim does not even take off. Indeed, as the Privy Council said in Yuen Kun Yeu ([37] supra) at 191–192:

Further, the expression of the first stage of the test [in Anns] carries with it a risk of misinterpretation. As Gibbs C.J. pointed out in Council of the Shire of Sutherland v Heyman, 59  A.L.J.R. 564, 570, there are two possible views of what Lord Wilberforce meant. The first view, favoured in a number of cases mentioned by Gibbs C.J., is that he meant to test the sufficiency of proximity simply by the reasonable contemplation of likely harm. The second view, favoured by Gibbs C.J. himself, is that Lord Wilberforce meant the expression “proximity or neighbourhood” to be a composite one, importing the whole concept of necessary relationship between plaintiff and defendant described by Lord Atkin in Donoghue v. Stevenson [1932] A.C. 562, 580. In their Lordships’ opinion the second view is the correct one. As Lord Wilberforce himself observed in McLoughlin v. O’Brian [1983] 1 A.C. 410, 420, it is clear that foreseeability does not of itself, and automatically, lead to a duty of care. There are many other statements to the same effect. The truth is that the trilogy of cases referred to by Lord Wilberforce in Anns v. Merton London Borough Council [1978] A.C. 728, 751, each demonstrate particular sets of circumstances, differing in character, which were adjudged to have the effect of bringing into being a relationship apt to give rise to a duty of care. Foreseeability of harm is a necessary ingredient of such a relationship, but it is not the only one. Otherwise there would be liability in negligence on the part of one who sees another about to walk over a cliff with his head in the air, and forbears to shout a warning.

Proximity

77        The first stage of the test to be applied to determine the existence of a duty of care is that of proximity, ie, that there must be sufficient legal proximity between the claimant and defendant for a duty of care to arise. The focus here is necessarily on the closeness of the relationship between the parties themselves, as alluded to by Bingham LJ in the Court of Appeal stage of Caparo Industries Plc v Dickman [1989] QB 653, where he said (at 679) that while “[t]he content of the requirement of proximity, whatever language is used, is not … capable of precise definition” and “[t]he approach will vary according to the particular facts of the case, … the focus of the inquiry is on the closeness and directness of the relationship between the parties” [emphasis added]. Indeed, in Hedley Bryne ([44] supra) itself, the House of Lords used language which pointed to the relationship between the parties as being determinative of duty.

78        Although Lord Keith in Yuen Kun Yeu referred (at 194) to the concept of “close and direct relations”, he did not clarify what the concept means. However, in the Australian High Court decision of Sutherland ([36] supra), Deane J has observed thus (at 55–56):

The requirement of proximity is directed to the relationship between the parties in so far as it is relevant to the allegedly negligent act or omission of the defendant and the loss or injury sustained by the plaintiff. It involves the notion of nearness or closeness and embraces physical proximity (in the sense of space and time) between the person or property of the plaintiff and the person or property of the defendant, circumstantial proximity such as an overriding relationship of employer and employee or of a professional man and his client and what may (perhaps loosely) be referred to as causal proximity in the sense of the closeness or directness of the causal connection or relationship between the particular act or course of conduct and the loss or injury sustained. It may reflect an assumption by one party of a responsibility to take care to avoid or prevent injury, loss or damage to the person or property of another or reliance by one party upon such care being taken by the other in circumstances where the other party knew or ought to have known of that reliance. Both the identity and the relative importance of the factors which are determinative of an issue of proximity are likely to vary in different categories of case. That does not mean that there is scope for decision by reference to idiosyncratic notions of justice or morality or that it is a proper approach to treat the requirement of proximity as a question of fact to be resolved merely by reference to the relationship between the plaintiff and the defendant in the particular circumstances. The requirement of a relationship of proximity serves as a touchstone and control of the categories of case in which the common law will adjudge that a duty of care is owed. Given the general circumstances of a case in a new or developing area of the law of negligence, the question what (if any) combination or combinations of factors will satisfy the requirement of proximity is a question of law to be resolved by the processes of legal reasoning, induction and deduction. On the other hand, the identification of the content of that requirement in such an area should not be either ostensibly or actually divorced from notions of what is “fair and reasonable” … or from the considerations of public policy which underlie and enlighten the existence and content of the requirement. [emphasis added]

79        We respectfully agree with this analysis which merely unpacks “proximity or neighbourhood” as a composite idea, importing the whole concept of the necessary relationship between the claimant and the defendant described by Lord Atkin in Donoghue v Stevenson (at 580). However, in this regard, we also acknowledge that there are numerous judicial pronouncements on the difficulty (and indeed, redundancy) in defining “proximity”. Some of these examples have already been cited above and it suffices only to reproduce Lord Oliver of Aylmerton’s speech in Alcock v Chief Constable of South Yorkshire Police [1992] 1 AC 310, where he said (at 411):

[I]n the end, it has to be accepted that the concept of “proximity” is an artificial one which depends more upon the court’s perception of what is the reasonable area for the imposition of liability than upon any logical process of analogical deduction.

80        Notwithstanding these judicial views, we agree with Phang, Saw & Chan ([26] supra at 42) that these observations are “puzzling, to say the least”. If indeed the “proximity” concept is merely a label or an artificial exercise in judicial creativity, then one must ask why the concept is still resorted to or utilised in the various tests. Its very presence suggests that it has some substantive content that is capable of being expressed in terms of legal principles. Rather than denouncing it as a mere “label”, the courts should strive to infuse some meaning into it, if only so that lawyers who advise litigants and even law teachers can make some sense of the judicial formulations.

81        In our view, Deane J’s analysis in Sutherland, that proximity includes physical, circumstantial as well as causal proximity, does provide substance to the concept since it includes the twin criteria of voluntary assumption of responsibility and reliance, where the facts support them, as essential factors in meeting the test of proximity. Where A voluntarily assumes responsibility for his acts or omissions towards B, and B relies on it, it is only fair and just that the law should hold A liable for negligence in causing economic loss or physical damage to B: see Phang, Saw & Chan at 47, where the authors wrote:

[B]oth perspectives are, at bottom, two different (yet inextricably connected) sides of the same coin and ought therefore to be viewed in an integrated and holistic fashion. [emphasis in original]

82        We only need to add, further, that in determining proximity as expounded by Deane J in Sutherland, the court should apply these concepts first by analogising the facts of the case for decision with those of decided cases, if such exist, but should not be constrained from limiting liability in a deserving case only because it involves a novel fact situation.

Policy considerations

83        Assuming a positive answer to the preliminary question of factual foreseeability and the first stage of the legal proximity test, a prima facie duty of care arises. Policy considerations should then be applied to the factual matrix to determine whether or not to negate this duty. Among the relevant policy considerations would be, for example, the presence of a contractual matrix which has clearly defined the rights and liabilities of the parties and the relative bargaining positions of the parties.

84        We also recognise that the obvious objection to utilising policy as the overarching determinant of liability is its potential to result in arbitrary decisions. Although it is generally recognised that public policy is an unruly horse (per Burrough J in Richardson v Mellish (1824) 2 Bing 229 at 252; 130 ER 294 at 303), it cannot be completely ignored. The danger is not with judges deciding cases based on policy considerations but rather with judges deciding cases based solely on them. We agree with Prof Tan ([29] supra at 228) that “[t]he truth lies somewhere in between pure principle-based decisions and policy-based decisions” and that “[i]t is obviously impossible to decide cases in vacuo, exclusive of the interests and the context of the community for which the decisions are made”. In our view, it is inescapable that some measure of public policy must be considered but it must not be the sole determinant.

85        We would also caution that when applying policy considerations to negate the imposition of a duty of care, the courts must be careful to differentiate such considerations from the requirement of proximity in the first stage of the test we have articulated. The courts must, as far as possible, avoid giving the impression that there remain “unexpressed motives” behind their finding for or against a duty. The courts must also not have litigants believe, as J A Smillie put it in “The Foundation of the Duty of Care in Negligence” (1989) 15 Monash U L Rev 302 at 302, that “none of [the tests articulated by the courts] is truly explanatory of judges’ reasoning or provides a helpful framework for analysis of the duty question”. If there is truly a pertinent and relevant policy consideration involving value judgments which reflect differential weighing and balancing of competing moral claims and broad social welfare goals, we feel that it would be better if the courts were to articulate these concerns under the requirement of policy considerations, rather than subsume these concerns within the proximity requirement, which may then lead to an overall distortion of the legal test to determine the existence of a duty of care. In this respect, we agree with Nicholas J Mullany when he notes in “Proximity, Policy and Procrastination” (1992) 9 Aust Bar Rev 80 at 83 that:

Judges should openly express the true premises of their decisions and, if recovery is considered to be undeserved, in the light of greater moral, social, economic, administrative or philosophical public perceptions, then these reasons and not others inherently uncertain in nature, should be expressed as the true foundation for denial of recovery.

86        With these general propositions in mind, we turn now to the facts of the present case.

Duty of care in the present case

Foreseeability of damage

The parties’ arguments

87        Adopting the test which we have set out, the threshold issue was the foreseeability of damage to the appellant. In this respect, it contended that the respondent’s responsibility under the PSSCOC, as the SO, included certification responsibilities under cl 32. Given that payment to the appellant under the PSSCOC structure was entirely dependent on the issuance of payment certificates certified by the respondent, the appellant submitted that it must have been foreseeable to the respondent that any negligence in its certification would deprive the appellant of moneys he would otherwise have been entitled to.

88        In response, the respondent contended that the appellant had not even crossed the low threshold of reasonable foreseeability as it was not even able to show that the damage allegedly suffered by it could have been foreseeable. In particular, it was argued that the appellant was in effect saying that the alleged under-certification led to its financial difficulties and these difficulties led to the need for it to novate the Contract. If it had not novated the Contract, it might still have been able to recover the amounts under-certified. In other words, the appellant had relied on the correctness of the interim certificates issued by the respondent in deciding to enter into the novation agreement. However, the respondent contended that it could not have reasonably foreseen that the appellant would rely on the said certificates for the novation given that the appellant had since 29 September 2000 been challenging the correctness of the respondent’s certifications.

Our decision

89        We rejected the respondent’s submission that this factual requirement was not satisfied because the respondent could not have foreseen (essentially) the kind of losses suffered by the appellant. This is because the test envisaged by the respondent (ie, foreseeability of the kind of losses) is properly the test for the remoteness of damages in tort, and is not the test for factual foreseeability in relation to the duty of care: see Ho Soo Fong v Standard Chartered Bank [2007] 2 SLR 181. As such, the factual requirement of foreseeability was satisfied in this case because, as the appellant rightly pointed out, it must have been foreseeable to the respondent that any negligence in its certification would directly deprive the contractor of moneys he would otherwise have been entitled to, and that if it had been paid the correct amounts, it might not have got into financial difficulties.

Proximity

The parties’ arguments

90        The issue of proximity obviously formed the main focus of contention between the parties, dominant as its role is in the determination of the existence of a duty of care (see [77] above). We therefore consider it useful to provide a closer analysis of the parties’ arguments.

(1)        The appellant’s arguments

91        The appellant submitted that it, as the contractor in the Project, had to rely on the professional judgment of the respondent, as the certifier, in agreeing to enter into a PSSCOC form of contract, under which it was held out to the contractor that payment for work done in the correct amount would depend entirely on the respondent, as an independent certifier, discharging its duties professionally, ie, fairly and impartially. Specifically, the appellant pointed out that the respondent must have known that any certificate it issued for payment would determine the amount payable to the appellant under cl 32 of the Contract. It was therefore unrealistic for the respondent to deny knowledge of the obvious reliance that would be placed on its certificate since it was not open for the appellant under the PSSCOC framework to obtain payment for work done by it in any other way.

92        Further, the appellant referred to cl 2.8 of the Contract as providing some specific indications of the respondent’s assumption of responsibility towards the contractor. As we mentioned earlier (at [4] above), this clause states that the SO or any of its representatives “shall at no time be under any obligation or duty to the Contractor either on behalf of the Employer or his own account to exercise or not to exercise any of his powers under the Contract”. The appellant contended that, properly construed, this clause meant that it would only apply if the respondent did nothing, but not if he did something, and that once he decided to exercise his power under the Contract it must follow that he had assumed responsibility to the appellant to do it properly and professionally.

93        Finally, the appellant pointed out that cl 32.8 of the Contract made it clear that the appellant had no remedy vis-à-vis the Employer for “any damages, whether by way of interest or otherwise, for any failure or delay by the [SO] in certifying any payment due or payable to the Contractor”. Since cl 32.8 had severely limited any recourse by the appellant to the Employer for damages arising from any failure or delay by the respondent in certifying any payment due or payable to the contractor, it was argued that the decision of the English Court of Appeal in Pacific Associates Inc v Baxter [1990] 1 QB 993 (“Pacific Associates”), which denied a duty of care in similar circumstances, was distinguishable as in that case there was no such restriction of the contractor’s right of recourse to the employer.

(2)        The respondent’s arguments

94        The respondent argued that the case was on all fours with Pacific Associates if the court took into account the existence of the appellant’s right to arbitration under cl 34 (see [5] above). It submitted that cl 34 gave extensive rights to the appellant to submit disputes for arbitration including submitting a claim against the Employer for any under-certification by the respondent. As such, the respondent contended that, on the facts of the case, it could not be said that the SO had assumed any responsibility for the appellant for the consequences of any under-certification or that the appellant had relied solely on the respondent’s certification to recover what was due to it for work done.

95        In relation to the appellant’s interpretation of cl 2.8, the respondent submitted that it was wholly misconceived and that it actually meant the opposite. It was argued that cl 2.8 referred not only to the absence of a duty to exercise the SO’s powers, but also of a duty not to exercise them, which meant that the SO could exercise them, if he wished, which he had done. Finally, as regards cl 32.8, the respondent submitted that while cl 32.8 precluded the appellant from claiming against the Employer damages for the failure or delay of the respondent in certification, the appellant had the right to claim the amounts under-certified and interest thereon by commencing arbitration proceedings against the Employer. This was clear from a perusal of cll 32 and 34.

96        Ultimately, the respondent submitted that, given the carefully drafted and detailed provisions governing the rights and liabilities of the Employer and the appellant which took into account the role of the respondent, it would be wrong for the court to impose a duty of care on it vis-à-vis the appellant. If the appellant had wanted the respondent to assume some liability to it in discharging the respondent’s duties to the Employer, it could have proposed a direct contractual arrangement with the respondent or introduce the respondent as a further party to the Contract but it chose not to do so. Alternatively, the appellant could have decided not to tender for the work since it knew or must have known about the contractual structure before it even submitted its bid.

Our decision

97        Adopting an incremental approach with respect to the requirement of proximity, we found the salient facts in Pacific Associates to be materially the same as those in the present case. The claimant in Pacific Associates was a contractor engaged in dredging work under the supervision of the defendant engineer who was retained by the employer. The claimant’s contract with the employer contained clauses providing that the engineer would not be personally liable for acts under the contract and providing for the arbitration of disputes between the contractor and the employer. The contractor claimed that the geological information in the tender document issued by the engineer had under-estimated the amount of hard materials to be dredged and that the engineer had acted negligently in rejecting the contractor’s claims for extra payment for removal of unforeseen hard materials. The contractor recovered some of its alleged losses from the employer following an arbitration settlement and then sought to recover the balance through a negligence action against the employer.

98        The Court of Appeal held (at 1032) that it would not be reasonable to impose a Hedley Byrne ([44] supra) duty on the engineer because it “would cut across and be inconsistent with the structure of relationships created by the contracts, into which the parties had entered”.

(1)        Reasonable reliance and assumption of responsibility

99        The court in Pacific Associates also held that the engineer did not owe a duty of care to a contractor who had suffered economic loss as a result of his decisions as the contractor was safeguarded by the terms of its contract with the employer. Purchas LJ considered that the courts should be slow to superimpose an added duty of care in excess of the rights the contractor was content to acquire against the owner in those cases where the parties came under a contractual framework which provided for compensation in the event of the failure of one of the parties involved. On the question of reliance, Purchas LJ held that if the contractor required extra-contractual protection for the defaults of the engineer, it was open to the contractor to stipulate for it when contracting. By accepting the invitation to tender the contractor must be taken to accept the role to be played by the engineer as defined in the contract. As such, the court could not accept that the engineer had held himself out as accepting a duty of care outside the contractual framework, or that the contractor had relied on such an assumption of responsibility. Purchas LJ, however, noted that the position might have been different if the arbitration clause had not been included in that contract.

100      Ralph Gibson LJ gave a similar analysis. In his judgment, he noted that there was no express or implied assumption of responsibility, emphasising that there was no approach or request by the contractor inviting the provision of a service or advice or information directly to the engineer. The engineer’s relationship with the contractor arose from the engineer’s contract of employment with the owner. It was to the owner that the engineer owed his contractual duty to use proper care. Since the contract provided for correction by arbitration, the engineer was not employed to exercise due care in the interest of the contractor, so he could not be regarded as assuming responsibility to the contractor. The duty to act fairly and impartially was owed only to the employer. Essentially, Ralph Gibson LJ denied the general duty on the grounds of both foreseeability (which he evidently regarded as a legal, as opposed to factual, requirement) and proximity, in view of the absence of voluntary assumption of responsibility.

101      On the other hand, Russell LJ, while reaching the same result, regarded the policy test as pivotal in the appeal. He thought that the engineer, in preparing tenders and contract documents, held himself out as an expert on whom the contractor was entitled to rely. However, on the policy test, he held that the contractor had freely entered into the contract and knew that a claim against the engineer must be ex-contractual. The contractor had its rights adequately protected by the contract with the owner and, had it thought the protection inadequate, it was at liberty to insist on a tri-partite contract. The parties having sought to regulate their relationships by contract, the law should be very cautious before grafting onto the contractual relationship what might be terms of a parasitic duty unnecessary for the parties’ protection. In Russell LJ’s view, while the requirement of foreseeability and proximity might be satisfied, it was not just and reasonable that there should be imposed on the engineer a duty which the contractor chose not to make a contractual one because, inter alia, the arbitration clause provided an adequate remedy in the event of under-certification.

102      The core of the judgments, as Ian Duncan Wallace noted in “Charter for the Construction Professional?” (1990) 6 Const LJ 207 at para 13, lies in the view that, merely by virtue of the professional entering into a contract of employment under which he is to perform the various duties required by his client’s contract with a contractor, and by so acting, it cannot be inferred as a fact that the professional (an engineer in Pacific Associates) has voluntarily assumed responsibility to the contractor for a careful discharge of his duties. As Ralph Gibson LJ’s judgment shows, only in circumstances in which the contractor makes some positive approach to the engineer for some service, advice, etc, followed by some positive statement or action on the part of the engineer on which it would be reasonable for the contractor to rely without further checking, may a voluntary assumption of responsibility be inferable.

(2)        Other jurisdictions following Pacific Associates

103      Pacific Associates has been followed in the Australian case of John Holland Construction and Engineering Pty Ltd v Majorca Projects Pty Ltd (2000) 16 Const LJ 114 (“John Holland”). Byrne J thought that the lowest common denominator in Pacific Associates, which prevented the court from finding a duty, was the arbitration clause. Regarding the duties of a certifier generally, Byrne J had the following to say (at 125–126):

To my mind it is reasonably foreseeable that the decisions of a certifying architect might cause loss to a contractor in a conventional building project if made negligently. Whatever that loss might be will depend upon the nature of the decision in question; it is in any event pure economic loss. This foreseeability is not removed by a right to review the decision by arbitration. Insofar as an entitlement to payment is concerned, a contractor who is successful upon arbitration must necessarily incur the cost of that arbitration and may suffer a deferment of payment, perhaps without a right to interest pending payment ... Nor am I troubled by either of the examples of policy consideration referred to in Bryan v. Moloney [sic] (1995) 182 C.L.R. at 618–9 ... The Builder in this case is clearly an identifiable member of the limited class of two persons who must have been in the contemplation of the Architect as being directly affected by its decisions. It is not an affront to the standards expected of a professional person that he or she should exercise the onerous responsibilities of certifier with due care and without partiality or unfairness. I suspect that the community, and particularly those concerned in the building industry, would expect no less.

The solution to the present problem must be found in the answer to the questions whether, in the contractual framework in which the parties to this project operated, it is established that the Builder relied on or depended upon the careful and impartial performance by the Architect of its certifying functions as here alleged and whether the Architect, for its part, assumed legal responsibility to the Contractor so to perform them.

104      Byrne J went on to find (at 126), on the basis of the contractual background, that it was “not appropriate ... to seek to engraft upon the contractual background a tortious obligation”. However, Byrne J could see no policy reasons for not imposing a duty in appropriate circumstances and accordingly accepted that in principle a certifier had a duty to make “decisions which are professional, careful and even-handed, not in the interests of any one party” (at 126). Further, Byrne J expressed the view that the right to review an architect’s decision by arbitration did not remove the foreseeability of economic loss being caused by a certifier’s negligence. Seen in this light, the reasoning in John Holland appears to be tilted towards the policy aspect of the reasoning in Pacific Associates, as opposed to subsuming the ability of the contractor to seek recourse in arbitration under the rubric of “proximity”.

105      Pacific Associates has also been followed in the Hong Kong decision of Leon Engineering & Construction Co Ltd v Ka Duk Investment Co Ltd (1989) 47 BLR 139, where the court refused to impose a duty on the professional on the ground that there was an adequate machinery under the contract for the contractor to pursue grievances, including those relating to under-certification, against the employer. At the close of the judgment, Bokhary J said (at 151):

The principle which I extract from … Pacific Associates ... is one which I would state in these terms: Where, first, there is adequate machinery under the contract between the employer and a contractor to enforce the contractor’s rights thereunder and, secondly, there is no good reason at tender stage to suppose that such rights and machinery would not together provide the contractor with an adequate remedy, then, in general, a certifying architect or engineer does not owe to the contractor a duty in tort coterminous with the obligation in contract owed to the contractor by the employer.

106      Elsewhere, the effect of Pacific Associates has been distinguished, primarily on the basis that what the professional says or does before a contract is entered into is entirely different from what he does when he makes an interim certification of payment. However, it has never been expressly overruled and the legal principles which it stands for (as discussed above) remain good law, as far as we are aware, in all Commonwealth jurisdictions.

107      The Canadian Supreme Court in Edgeworth Constructions Ltd v ND Lea & Associates Ltd (1993) 107 DLR (4th) 169 has held that there might be situations where a negligent statement by the professional in preparing the tender documents would lead to liability under Hedley Byrne ([44] supra) but, in the absence of such a statement to the contractor, it seemed unlikely that a court would find a duty of care based on the mere conduct of the professional. Similarly, in J Jarvis and Sons Ltd v Castle Wharf Developments Ltd [2001] EWCA Civ 19, the English Court of Appeal distinguished Pacific Associates and held that the professional agent of an employer could be liable to a contractor for negligent misstatements made by that agent to the contractor with the aim of inducing the contractor to tender, provided that the contractor placed reliance on those statements. The question of whether a duty of care had arisen was dependent on the given facts and, in particular, on what had been said to the contractor.

(3)        Application to the present case

108      Applying the findings in Pacific Associates in relation to proximity to the present case, we were of the opinion that the requirement of proximity was not satisfied. The trial judge at [79] of the Judgment stated that there was no assumption of direct responsibility by the respondent to the appellant, but did not explain why this was so. In view of the presence of the arbitration clause in the Contract, ie, cl 34 (see [5] above) in the present case, it was not possible for this court to depart from the reasoning in Pacific Associates and hold that the respondent had a duty of care to the appellant to certify the payments for work done correctly. For this reason, we held that there was no legal proximity for a duty of care to arise.

109      With respect to the arguments based on cl 32.8, we were of the view that although the appellant had no remedy vis-à-vis the Employer for “any damages, whether by way of interest or otherwise, for any failure or delay by the [SO] in certifying any payment due or payable to the Contractor”, thereby precluding the appellant from claiming damages against the Employer for failure or for delay of the respondent in certification, the appellant was always free to claim the amounts under-certified and interest thereon by arbitration proceedings against the Employer. This was clear from a perusal of cll 32 and 34. In addition, cl 22 specifically allowed the contractor to claim “Loss and Expense” against the Employer under many stated grounds. Hence, Pacific Associates was not distinguishable because of cl 32.8.

110      In relation to the effect of cl 2.8, we accepted the respondent’s argument as set out in [95] above.

Policy considerations

111      In reaching our decision, we had also considered the parties’ arguments on whether if there had been proximity, there would have been any policy considerations that would have persuaded us to deny any remedy to the appellant.

The parties’ arguments

112      In this regard, the appellant first submitted that given the recognised and accepted position of the certifier having to be independent and impartial to both the employer and the contractor, it was logical that a duty of care be imposed on the certifier to ensure that he would discharge his duties professionally to all parties who might be affected by his certification. This would also help to improve the overall standards in the industry. To this submission, the respondent argued that it was merely speculative.

113      Secondly, the appellant stated that the trial judge’s concern about opening the floodgates must be considered in its appropriate context. In particular, while there might be additional claims made (such as from sub-contractors), it was within the power of the courts to control the risk by recognising the appellant’s claim and denying the other claims. In this case, the respondent’s duties, certification and otherwise, were aimed at the party in direct contract with the Employer, ie, the main contractor, and there was no risk of indeterminate liability. The respondent’s argument, in response, was that it would not be just or reasonable to impose a duty of care on the SO in view of the contractual framework to which the appellant had agreed.

Our decision

114      On the facts of the present case, the same reasons above articulated by Russell LJ in Pacific Associates can also be characterised as policy considerations under the second stage of the test in Anns. For the reasons considered above, viz, a duty of care should not be superimposed on a contractual framework, we considered that there were cogent policy reasons why, assuming that proximity was found, a duty of care should nonetheless not be imposed on the respondent.

Conclusion

115      To recapitulate: A single test to determine the existence of a duty of care should be applied regardless of the nature of the damage caused (ie, pure economic loss or physical damage). It could be that a more restricted approach is preferable for cases of pure economic loss but this is to be done within the confines of a single test. This test is a two-stage test, comprising of, first, proximity and, second, policy considerations. These two stages are to be approached with reference to the facts of decided cases although the absence of such cases is not an absolute bar against a finding of duty. There is, of course, the threshold issue of factual foreseeability but since this is likely to be fulfilled in most cases, we do not see the need to include this as part of the legal test for a duty of care.

116      In the circumstances, applying the two-stage test, we found that there was no duty of care as the appellant had contended for. As a result, we did not have to consider the additional issues of breach, causation and remoteness which had been canvassed before us. The appeal was dismissed with costs and with the usual consequential orders.



[note: 1] Record of Appeal, Vol III Part B at 658.

[note: 2] Record of Appeal, Vol III Part H at 2166.

[note: 3] Record of Appeal, Vol III Part B at 660.

[note: 4] Record of Appeal, Vol III Part A at 358.

[note: 5] Record of Appeal, Vol V Part N at 8338.

[note: 6] Record of Appeal, Vol V Part N at 8342.

[note: 7] Record of Appeal, Vol III Part I at 2328 to 2335.

[note: 8] Record of Appeal, Vol V Part N at 8418 to 8421.

[note: 9]Record of Appeal, Vol V Part Q at 9307.

[note: 10]Record of Appeal, Vol V Part R at 9478.

[note: 11]Record of Appeal, Vol V Part R at 9494 to 9496.

Copyright © Government of Singapore.

Back to Top

This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

Version No 0: 08 Aug 2007 (00:00 hrs)