‘Three Party’ Cases—Into the Wasteland

In the second type of case, the provision of false information to the third party impacts detrimentally upon the plaintiff's economic interests, without the plain­tiff himself ever actually using or relying on the information. It is the third party ^,s reliance on the information (or the reliance of other market actors) that provides the causal pathway between the defendant's carelessness and the plaintiff's loss. This type of case is typified by solicitor-beneficiary cases such as White,⁵⁵ Hill^[1573] and Ross v Caunters;^[1574] and by cases involving negative job references supplied by a defendant to a plaintiff's prospective employer, such as Spring v Guardian Assurance,^[1575] although there are other examples. White, Hill and Ross are more usually categorised as cases of negligent service provision, not ‘negligent misstate­ment', because in each of them the solicitor's error lay not in making any care­less statement, but in omitting to act so as to secure the wishes of his client to confer economic benefit on the plaintiff.^[1576] I have discussed such cases elsewhere.^[1577] It might be possible to dispose of them justly by expanding the boundaries of the law of contract, although changing privity rules in the way that is required would not be as straightforward as is often assumed.^[1578] It is the clear failure of the solicitor to perform his contract with his client in cases of this sort that attracts some com­mentators and judges to the idea that the solicitor's duty to the plaintiff is based on a relationship ‘akin to contract'.^[1579] The High Court has not directly considered a negligent reference case such as Spring and seems reluctant, in any event, to allow such claims to undermine defences in the law of defamation,^[1580] which leaves the job reference category empty and the applicable principles under Australian law yet to be determined.

There is too little space to consider the proper approach to the duty question in such cases here and I confine myself to the obvious observation that the concept of a special relationship of ‘ reasonable reliance' will most likely be a redundant conceptual tool in such instances, because the plaintiff himself or herself never relies on the advice given.^[1581] Below, I focus instead on the first, more familiar type of three-party case, in which a plaintiff himself makes use of information supplied by the defendant that has been obtained indirectly via a third party.

B. The Wasteland of Esanda

Esanda was a typical case of this sort. It involved a careless audit report relating to the financial affairs of a particular company that was detrimentally relied on by the company's financier. The confusion surrounding its contribution to the duty debate stems from several factors. First, the plaintiff's pleadings were struck out on a straightforward basis that did not require the court to address the finer points of the duty question. This renders virtually all of the court's reasoning on that issue obiter. Secondly, the five judgments in the case vary significantly in form and detail, although there are undoubtedly some commonalities, which I draw out below. Thirdly, the case was decided at a time when ‘proximity' reasoning was still good currency in Australian law, which means that one must now bypass that language in order to extract any intellectual sustenance from the case. Finally, the judgments were all given several years before the Tepko case, in which the High Court affirmed the Barwick approach. It is unclear to what extent the principles articulated in Esanda in 1997 have been superseded by the more recent statements of the law in the latter case (in 2001). Three members of the Court (Gummow, Gaudron and McHugh JJ) sat in both cases.

Despite Tepko, Esanda is still regularly cited in cases of the three-party configura­tion. This may be because Tepko is itself understood as a case of the more straight­forward two-party type—the plaintiff in that case had been in direct contact with the defendant and had previously requested the information supplied in a way that the plaintiff in Esanda had not.

How, if at all, do the principles in Esanda differ from those postulated by Barwick CJ in Evatt? If one looks beyond the language of proximity and asks what it was that the judges thought might make a relationship proximate, the approaches of Brennan CJ, Dawson J, and Toohey and Gaudron JJ are actually all pretty close to Barwick CJ's own. Dawson J thought that a proximate relationship would be trig­gered by a defendant's actual or constructive knowledge of a plaintiff's reasonable reliance on information provided. Indeed, he saw his reasonable reliance principle as simply a more modern way of expressing Barwick CJ's requirement of ‘ trust'. He rationalised the additional principles stated by the majority judgment in San Sebastian^[1583] above as simply assisting in the application of that approach in cases in which the plaintiff had not himself requested the information^[1584]—a circumstance that Barwick CJ himself had not addressed in any real detail.

The joint judgment of Toohey and Gaudron JJ was similar in general import. Stripping away the language of assumption of responsibility (about which their Honours were themselves sceptical), they concluded that the essential elements of a special relationship were knowledge and reasonable reliance.^[1585]

Brennan CJ drew on both Barwick CJ's approach and the speeches of members of the House of Lords in Caparo. His formulation is close to that of Lord Bridge in that case. He said:

In every case, it is necessary for the plaintiff to allege and prove that the defendant knew or ought reasonably to have known that the information or advice would be communi­cated to the plaintiff, either individually or as a member of an identified class, that the information or advice would be so communicated for a purpose that would be very likely to lead the plaintiff to enter into a transaction of the kind that the plaintiff does enter into and that it would be very likely that the plaintiff would enter into such a transaction in reliance on the information or advice and thereby risk the incurring of economic loss if the statement should be untrue or the advice should be unsound.^[1586]

Emphasis here is placed on a high degree of probability that the information will be used and will cause loss; on the purpose for which the information is provided; and on the defendant's knowledge that the advice will be used in a transaction of the same kind as the defendant contemplated.

McHugh J's judgment stood apart from the others and drew directly from the majority judgment in San Sebastian. It concluded:

[T]he position in Australia to date with respect to liability for pure economic loss caused by negligent misstatement is that, a bsent a statement to a particular person in response to a particular request for information or advice or an assumption of responsibility to the plaintiff for that statement, it will be difficult to establish the requisite duty of care unless there is an intention to induce the recipient of the information or advice, or a class to which the recipient belongs, to act or refrain from acting on it. Mere knowledge by a defendant that the information or advice will be communicated to the plaintiff is not enough.^{[1588] [1589]}

This section has given rise to considerable uncertainty regarding the importance of a defendant's intention in cases of the remoter three-party configuration. On the one hand, McHugh J clearly approved of the decision in R Lowe Lippmann Figdor and Franck v AGC (Advances) Ltdⁿ in which the Victorian Court of Appeal had interpreted S an Sebastian to emphasise the significance of the defendant's intention in cases of this type. At the same time, however, he suggested that the fact that a defendant does not intend to induce the plaintiff to rely is ‘not neces­sarily fatal' to a claim.^[1590]

Gummow J disposed of the case on the very limited basis that the pleadings on their face were insufficient to disclose a case for a duty of care.

C. Lower Court Confusion

Lower courts have been unsure how to interpret or apply the various judgments in Esanda. One can sympathise with their dilemma. The extent of the confusion is illustrated in a Federal Court decision of 2004, Charben Haulage Pty Ltd v Environ­mental & Earth Sciences Pty Ltd.^[1594] The plaintiff in that case was a redeveloper who purchased land in reliance on an environmental report that had been prepared by the defendant company for the land's vendor. The report was prepared under contract with the vendor, but as Wilcox J found, the defendant knew full well that the vendor was selling the site and that the reason the vendor wanted the report was in order to persuade potential purchasers that it was suitable for domestic redevelopment. He also found that the defendant knew, or ought to have known, that the report would be shown to potential purchasers in order to get them to make the purchase.^[1595] He nonetheless rejected the negligence claim. His reasoning was brief and, with respect, obscure. Importantly, he accepted that the claim might well succeed under Brennan CJ's approach, but nonetheless rejected it on the basis that it was not clear that ‘the other members of the court' in Esanda would have imposed a duty of care on the facts, because the plaintiff had not requested the report and the defendant had not produced it ‘with the intention of causing [the plaintiff] to rely upon it'.^[1596]

An obvious difficulty with this thinking appears to be that, on the analysis pre­sented above, none of the approaches of Dawson, Toohey and Gaudron JJ would have ruled out the existence of a duty of care on the facts—indeed, the approaches of these judges seem at least as generous as that of Brennan CJ and to have been broadly in line with the Barwick approach that prevails in the basic two-party case.

This is clearly very problematic and it is hard to understand why, on the facts of Charben, a duty was not found. The most likely explanation is that Wilcox J was simply reluctant to step out upon such uncertain ground and did not need to do so, because the plaintiff had a clear, alternative, statutory claim for misleading and deceptive conduct, which succeeded.

Other lower court decisions since Esanda do little to resolve the uncertainty. In Derring Lane Pty Ltd v Fitzgibbon in 2007, the Victorian Court of Appeal ulti­mately upheld the trial judge's view that a duty of care could be owed on either Brennan CJ or McHugh J's approach.^{[1598] [1599] By contrast, in ABN Amro v Bathurst Regional Council,32 the Full Federal Court thought it appropriate to apply the basic Barwick test as approved in Tepko, even in claims of the three-party, ‘indirect reliance' pattern.[1600] One of the defendants in this complex proceeding sought at one stage to try to deny liability by relying on McHugh J's more restrictive ‘inten­tion to induce reliance' test, but was apparently knocked back by the Court on the basis that the criteria of the two-fold test accepted in Tepko were sufficient.[1601]}

The upshot is that courts in three-party cases of the Esanda configuration are faced with three slightly different types of approach to apply, all of which have High Court approval. The first is the straightforward, two-stage Barwick approach based on knowledge and reasonable reliance, approved in Tepko. The second is Brennan CJ's approach in Esanda, which is very similar, but which focuses on the foreseeability of economic harm at a higher level or probability (‘very likely') and which clearly insists that the purpose for which an indirect recipient of informa­tion relies must be of the same type as that for which the defendant intended to supply the information. The third, which stems from San Sebastian and the judg­ment of McHugh J in Esanda, is an approach based on proof that the defendant intended the plaintiff to rely on the advice or information given. The differences between these approaches may appear slight and on many factual configurations will be immaterial, but cases like Charben show that they are real. That case also demonstrates the way in which courts are able to avoid grappling with the uncer­tainties of negligence law by deciding cases on an alternative, statutory basis.

D. Stabilising the Duty Test and the Subsidiary Role of Intention

The uncertainties about the duty issue are unhealthy and need to be resolved. Here, I suggest that the duty criteria in the type of three-party case we have just exam­ined should be no different to those that apply to the basic, two-party case. The appropriate approach is the two-stage version of the Barwick test, as applied by the High Court in Tepko, but accepting the slightly broader views of the minority in that case as regards the degree of knowledge a defendant must have before a duty can arise. Where information, opinion, or advice provided by a defendant is obtained by a plaintiff—whether directly from the defendant or indirectly via a third party—and relied upon, a duty of care should therefore arise in respect of the plaintiff's economic interests if, and only if:

1. the defendant knew or ought to have known that the plaintiff (whether individually or as a member of an ascertainable group) was likely to use the information in respect of some serious matter of business, and

2. the plaintiff reasonably relied upon that information in entering into the transaction in question.

Although the defendant's actual and apparent intentions and purposes do not themselves create the duty in such an instance, they are not irrelevant. In fact, they are germane to both limbs of the test in the following ways. First, the defendant's actual intentions as to how information will be used are clearly relevant to his actual knowledge that it may be relied on by the plaintiff. If he intends the infor­mation to be used by the plaintiff for investment purposes, for example, he simply cannot deny that he knew it was likely to be used in this way. On the other hand, he may still have actual or constructive knowledge that it would be used by the plaintiff for such a purpose (and therefore owe a duty) even if he does not actually intend that use. It is his knowledge, not his subjective intention, that is crucial. For the same reason, a defendant's disclaimer of responsibility is not invariably fatal to the existence of the duty of care.^[1602] Such a disclaimer may (in fact will it not always?) demonstrate that the defendant does not s ubjectively intend the plaintiff to rely, but this does not preclude his actual or constructive k nowledge that the plaintiff will in fact do so. The fact that it is knowledge and not intention per se that is important is demonstrated by those rare but important cases in which courts have been prepared to impose duties even when a disclaimer is present.^[1603]

Secondly, the defendant's apparent intentions (those that would appear to a rea­sonable observer) and the apparent purpose of the information that has been given are relevant both to the defendant's constructive knowledge of the use to which it might be put and to the question whether it was reasonable for the plaintiff to rely on it in the way that he did. The principles here are exactly the same as in product liability cases and there is no magic or mystery about them. A defendant manufac­turing a microwave oven owes no duty of care to a consumer who uses the oven to dry his chihuahua after a long walk in the rain, rather than to reheat his lunch. This is because the manufacturer cannot reasonably know that the oven will be put to this use, given the reasonably apparent purpose of such an oven. Furthermore, an owner using the oven for such a purpose is likely to be regarded as the real legal cause of the unfortunate dog's death, not the defendant manufacturer. It is a case where we reprimand him for his own stupidity, not one where we award him compensation against the manufacturer. As with products, so with words, which are simply products of a more ephemeral kind. One who ignores the apparent purposes of information (that is, the purposes which a reasonable person would think it had) and relies on it for some other, unreasonable design of his own is simply the cause of his own financial loss and must bear the risk of his decision to invest on the back of it. It may be hard, I accept, to accuse an investor of quite the same rank stupidity as the dog owner, where he uses information for a different financial purpose to that for which it was produced, but this is simply a matter of fact and degree. I would prefer to say that the claim fails for lack of causation, or for remoteness of the damage, but accept that the current strategy of courts is to say that no duty of care is owed in such a case.

This way of understanding the role of a defendant's actual and apparent inten­tions is potentially helpful, I suggest, in resolving some of the current tensions in the case law. It means that there is no separate or distinct, restrictive, intention­based ‘ test' that applies in controversial misstatement cases of a remoter pattern, but it still keeps faith with the regularly stated judicial view that the purposes for which advice is given are of crucial importance in working out whether a duty of care is owed in respect of it. The approach is also consistent with courts' now almost universal insistence that duties of care in misstatement cases of all con­figurations are imposed by courts, not genuinely ‘voluntary' in the sense of being a product of the defendant's own will or apparent intention as such. And it can help to explain why judges, including McHugh J himself, have expressly accepted that it is not necessarily fatal to a misstatement claim that the defendant did not intend the plaintiff to rely on his advice.

Understanding courts' references to intention in this way is, in my view, prefer­able to understanding modern negligence liability as the product of the parties' subjective or objective intentions, as voluntarists do; or using ‘intention to induce reliance' as a restrictive alternative ‘practical test’^[1604] for duty in categories of mis­statement case in which the risks of indeterminate liability are perceived to be higher than usual. True it is that when the law sets about protecting new, more controversial types of interest, it tends first to protect them against intentional interference; and then only later to lower the bar so as to admit negligence claims.^[1605] But the signs are that the law of negligence has moved beyond restrictive ‘inten­tion' tests in deciding when to protect economic interests, just as it has moved beyond restrictive contractual paradigms as a way of organising its thinking about the modern law of tort. Where courts currently insist on intention in negligence cases, it is not because intention is an element of the cause of action in negligence, as it is in deceit. It is because it is required for some other ‘practical' (ie pragmatic) reason.^[1606] Historically, this reason relates to the need to keep the bounds of liability under control for fear of creating indeterminate or mar ket-dist ortive liabilities. But the concept of intention is not needed, I suggest, as a ‘control test' artificially restricting the ambit of the duty of care to these instrumental ends.

E. Accommodating Residual Concerns

1 suspect at this point that some pragmatists may think that adopting the basic Barwick two-stage test in all cases in which a plaintiff has relied on information or advice may be going too far, and that it does too little to deal with concerns about indeterminate liability in three-party cases. Properly understood and applied, this is not the case. The knowledge requirements adverted to in Section IIIB are per­fectly adequate to perform this task. The requirement that a defendant's reliance on information be reasonable ought also to cut out a very large number of speculative cases in which plaintiffs stand in remote, indirect relationships with those upon whose advice they rely. It is also always open to a defendant providing information on a commercial basis to limit or exclude its liability for carelessness; and when this is done courts are then also likely to refuse to impose any duty of care towards remoter parties that would undermine the way in which the defendant's liability has been contractually allocated.^[1607] It is important to remember in this regard that the proposed two-stage test must always be met for a duty to be owed, but that such a duty can always still be negated if it would contradict a contractual alloca­tion of economic risk, subject an advisor to conflicting duties,^[1608] undermine the coherence of other rules of law,^[1609] or frustrate the purposes of a statutory scheme under which advice has been supplied. Remember also that, even if information is provided without a contract, a disclaimer of liability can normally provide a defendant with a good measure of protection for reasons already discussed.

If all of this is insufficient to allay residual concerns about extensive or indeter­minate liabilities in cases of the remoter three-party configuration such as Esanda and Charben, on the basis that clearer, fixed limits to liability are needed to protect the health of information markets and to enable defendants to set their pricing and provide for sufficient levels of liability insurance, I suggest that a viable way of meeting these concerns is to introduce a system of statutory liability-caps for advi­sors operating in particularly high-risk areas of the information economy, such as auditors and public bodies.^[1610] The benefit of caps is that they can be introduced after a thorough empirical study of the state of the relevant markets, which relieves courts of the need to limit liabilities through their own guesswork about the potential empirical effects of tort liabilities, when they really have no reliable facts and figures to go on. Furthermore, since insurance and information markets seem to go through cyclical phases of health just like the rest of us, caps can be adjusted periodically and relatively quickly, to adjust to prevailing social conditions.

The resulting pattern of liability provides a reasonable balance in terms of the onus of protection as between the suppliers and users of advice. Furthermore, if advisors are still concerned that they may end up paying the bill for a plaintiff's full economic losses when they are really only ‘ secondarily' responsible for them relative to some other fraudster or negligent party who is more at the heart of the matter, they are now accorded significant additional protections in Australian law as a result of the introduction of reforms that abolish the principle of joint and several liability in a wide range of cases. My own view is that this abolition was a precipitous step too far, taken on too little evidence, and that capping is prob­ably a more effective mechanism for the reassurance of markets.^[1611] In any event, however, the point is that advisors in Australia need have no anxiety about having to shoulder financial losses caused by other wrongdoers who are allegedly more centrally implicated in a financial disaster, but who are now insolvent.

F. Reconciling the Rules with the Multifactoral Approach

A final puzzle is how the rules I have outlined above for misstatement cases sit with the ‘multifactoral' approach to duty that is now taken by the High Court in cases falling outside existing precedents. According to this approach, in ‘new' cases, courts consider a number of different ‘factors' or ‘salient features' in determin­ing whether a duty is owed. The only fixed requirement, for a duty is that the loss be reasonably foreseeable. Otherwise, no factor is absolutely necessary and courts simply weigh them up in the round.^[1612] In one recent case, it has been suggested that there could be more than 18 different factors to consider,^[1613] depending on the sort of case one is dealing with; and that the list is an open one.

This is not the place to provide a full critique of this approach, but one can see how, if multiple, flexible factors were to be allowed to replace basic duty rules in cases of negligent misstatement, liability would become almost impossible to predict. The law would rapidly move to such a high state of entropy as to dis­solve into total chaos. Some misstatement cases of the three-party configuration have recently begun to bump up against the multifactoral approach and the Full Federal Court has interestingly gone so far as to suggest that ‘notwithstanding the guidance provided by the Evatt [Barwick] principles' a court must now undertake a ‘case-by-case identification of whether a duty arises' in a negligent misstatement case ‘c onsistently with the rather broad principles [ie factors] mentioned in Perre v Apand and other cases that deal with liability for pure economic loss'.^[1614] Herein lies the potential for a set of reasonably clear, discrete rules governing the duty question in misstatement cases to be replaced or unsettled by a broad and uncertain smorgasbord.

This is clearly not what the High Court intended and we must be careful that it does not happen. The rules set out above provide a reasonably predictable pattern and should not now be replaced by an open list of discretionary factors tout court. Nor, I suggest, would it be helpful to lump misstatement cases in with all other types of pure economic loss case in a grand m elange, since they evince at least some discrete concerns. Rather, the multifactoral approach should be used only in cases that fall entirely outside existing precedents (ie those that are truly ‘novel')^[1615] as a tool to assist in the further development of legal rules in existing categories of case.

Courts are aware of these issues and are doing their best to reconcile the tra­ditional rule-based approach with the more open-textured, multifactoral one. In Amro, for example, the Full Federal Court thought that there was no need to give the factors of ‘indeterminate liability' or ‘vulnerability' a distinct role in rea­soning about misstatement cases, precisely because the rules in such cases have already been designed in such a way as to accommodate them.^[1616] This is a sensible approach. Whilst it is therefore true to say, as the High Court has done on several occasions, that there is a clear link between the ‘ vulnerability' factor that is now widely used to assist in determining the existence of duty in new cases and the rule of factors identified in Perre needed to be satisfied before one could proceed to consider the duty issue further—see Perre at [133] and even more clearly Crimmins at [93]. Those small certainties now appear to have been discarded. This would be a matter of dismay to McHugh J, who emphasised the need to keep the list of factors small and directly relevant to the relevant category of law, failing which decision making would lapse into discretionary remedialism: Crimmins at [77].

in misstatement cases that a defendant's reliance on advice must be ‘reasonable',^[1617] the latter rule should be understood as a concrete manifestation of the broader principle, rather than the broader principle now itself being used to determine every misstatement case. General principles should be used to inform the develop­ment of particular rules pedetemptin in discrete types of case, not as a substitute for rules. The ‘ factors' which appear most obviously relevant to the development of rules regarding misstatement duties are the foreseeability of harm, the defend­ant's knowledge, indeterminate liability, respect for the defendant's autonomy (the need not to undermine legitimate competition or free social interaction), conflict with contractual allocations of risk or statutory purposes, and the need to develop the law of tort in a way that coheres with other fields of law dealing with misstatements and economic loss, such as the law of defamation.

V.