CONTRACT breakers do not have to compensate their counterparties for all losses caused by their breach of contract. Some losses are too remote. In Hadley v Baxendale (1854) 9 Ex. 341, 355, Alderson B. held that the only types of loss that were recoverable were “those arising naturally or ordinarily from the breach” and those in the reasonable contemplation of the parties at the time of contracting. What degree of probability of the kind of loss must be in reasonable contemplation at the time of contracting? That was the question tackled by Lord Hodge in Attorney General of the Virgin Islands v Global Water Associates Ltd. [2020] UKPC 18, [2020] 3 W.L.R. 584 (hereafter “Global Waters”).

The dispute arose from two contracts between the Government of the British Virgin Islands (BVI) and a private contractor. The first required the contractor to design and build a water plant in the BVI; the second required the contractor to manage, operate and maintain that plant for 12 years. The Government was obliged to provide the contractor with a site for the proposed plant which, in breach of the first contract, it failed to do. As a result, the contractor lost the profit it would have earned under the second contract. Astonishingly, the arbitrators held that this loss was too remote and the BVI Court of Appeal agreed. The Privy Council quite rightly allowed the appeal.

The likelihood of loss required under the remoteness rules for contract damages is a matter of some debate. Despite contrary dicta in Parsons (Livestock) Ltd. v Uttley Ingham & Co. Ltd. [1978] Q.B. 791 (C.A.), the starting point is that the test for reasonable contemplation in contract is more robust that the equivalent test in tort. Tort law takes its lead from The Wagon Mound (No 2) [1967] 1 A.C. 617 (P.C.), although that was really a decision about fault. At 642–3, Lord Reid is taken to have set down a rule that a foreseeability of “real risk” of the type of loss sufficed for damages in negligence. But a year later, in The Heron II [1969] 1 A.C. 350 (H.L.), Lord Reid said that for contract damages the type of loss must be “not unlikely” (p. 383) or “quite likely to happen” (p. 390). In so holding, Lord Reid criticised the test articulated in Victoria Laundry (Windsor) Ltd. v Newman Industries Ltd. [1949] 2 K.B. 528 (C.A.). There, Asquith L.J. concluded that the test was whether it was reasonably in contemplation that there was a “serious possibility” or a “real danger” of the type of loss arising from the breach (pp. 539–40). In The Heron II, Lord Reid was alone in criticising this aspect of Victoria Laundry. Lord Morris regarded Asquith L.J.'s judgment as “valuable by way of exposition” (p. 399); Lord Hodson admitted that it was “not … possible to improve” (p. 411); Lord Pearce considered “the expressions used … were right” (p. 415); and Lord Upjohn was “content to adopt” that language as the test (p. 425). Nonetheless, after The Heron II, a consensus has gathered around Lord Reid's “not unlikely” formulation of the test (see e.g. Chitty on Contracts, 33rd ed. (London, 2018), para. 26.121; McGregor on Damages, 20th ed. (London, 2017), paras. 8.183–86).

In Global Waters, Lord Hodge appeared to reinstate Victoria Laundry's “serious possibility” test to express the requisite degree of probability (at [32]). Presumably this was because he considered it best encapsulated the examples given by the Law Lords in The Heron II that a 10:1 chance of the ceiling of the court falling down was good enough odds to allow damages but a 51:1 chance of drawing a particular card in a pack of cards or winning on the premium bonds were both too unlikely for the purposes of contractual remoteness (cf. The Wagon Mound where oil catching fire on water was practically impossible but still a sufficient prospect to satisfy the remoteness rules in negligence).

Less helpfully, Lord Hodge treated the difference between Asquith L.J. and Lord Reid as being merely a disagreement about the “choice of language” (at [28]). He considered “not unlikely” was synonymous with “serious possibility”. That is unconvincing. In The Heron II, Lord Reid was clear that he disagreed with the substance of a serious possibility test; his “not unlikely” test was intended to be more restrictive. Perhaps Lord Reid's objection in The Heron II was prompted by Asquith L.J.'s decision in Victoria Laundry to treat “serious possibility” and “real danger” and (worse still) “on the cards” as expressions of the same idea. The imprecision of “on the cards” was rightly called out by Lords Reid, Morris, Pearce and Upjohn in The Heron II. As for “real danger”, its problem is that it is difficult to see how it differs from a “real risk” test in tort law, despite the fact that the remoteness thresholds in contract and tort are meant to be different, and certainly operate differently in practice. In Global Waters, Lord Hodge was careful not to endorse Asquith L.J.'s “real danger” test. Given his conflation of the “serious possibility” and “not unlikely” tests, it seems that “serious possibility” in Global Waters does not quite mean the same thing as “serious possibility” in Victoria Laundry. To avoid semantic confusion, it would be preferable for the judiciary to retain Lord Reid's language of “not unlikely” where possible.

Lord Hodge also held that “the criterion for deciding what the defendant must be taken to have had in his or her contemplation as the result of a breach of their contract is a factual one” (at [35]). This is puzzling. The normal meaning of “criterion” is a test or standard by which something is to be judged; that must, by definition, be a question of law not fact. It is also regrettable that Lord Hodge expressed the reasonable contemplation test in terms of what the party in breach should reasonably have contemplated at the time of contracting (see also [33]–[34]). It makes no sense, at the remedies stage, to limit damages by reference to the position of one party only. The essence of a contract is its synallagmatic nature. The better view, therefore, is that the court should consider what would reasonably have been in the contemplation of both parties when contracting.

The most fundamental criticism to be made of Global Waters is its confused treatment of the assumption of responsibility test articulated by Lords Hoffmann, Hope and Walker in The Achilleas [2008] UKHL 48, [2009] 1 A.C. 61. Lord Hodge, at [26], associated this test with only the first two of these Law Lords and characterised it as an additional restriction on the recoverability of damages that arose only in unusual cases. What constitutes an unusual case is entirely opaque. Then, at [28], Lord Hodge emphasised that the serious possibility test “was not simply a question of probability”, citing Lord Walker's very exposition of the assumption of responsibility test in The Achilleas. While not free from controversy, the better view is that determining the parties’ assumptions of responsibility is as an exercise in construction: see A. Kramer, The Law of Contract Damages, 2nd ed. (Oxford, 2017), paras. 14.06–08. Thus, the reasonable foreseeability test is not an external test imposed by law. Rather, it provides a default answer which might be restricted or expanded once the relevant text and context are considered. The parties are presumed to have agreed that reasonably contemplated kinds of loss caused by a future breach of contract would be compensated by the contract breaker, unless some other allocation of risk is apparent on the contract's express or implied terms.

The disposal of the appeal in Global Waters itself demonstrates the centrality of contract terms to issues of remoteness. On the facts, the two contracts were entered into at the same time between the same parties and concerned different stages in the life cycle of the same water plant. It was clearly not unlikely (to put it at its lowest) that the failure to complete the first contract would prevent performance of the second. Lord Hodge held that the existence of two contracts, rather than one, could not be taken implicitly to rebut that presumption (at [39]). Nor could the loss be treated as too remote because of the possibility that the Government could have engaged another contractor to build a water plant which could then have been handed over to the claimant to manage it in performance of the second contract. That was not what had been envisaged under the terms of the second contract when properly construed: the parties had agreed that the claimant would manage the water plant it had constructed under the first contract, not another plant nominated by the Government (at [42]).

The apparent side-lining of The Achilleas in Global Waters is regrettable. But the basis on which the appeal was determined will nonetheless give some encouragement to proponents of the assumption of responsibility approach to remoteness in contract damages.