Assumption of Responsibility and the Law of Negligence

It is first important to deal with a misunderstanding that can arise in this context. A useful way to do so is to examine an argument presented by Robertson and Wang in this volume.

Whether an implied undertaking has been given by conduct is a matter of interpretation, determined by convention as to when it can be concluded that one party is assuming responsibility for another. Like a surgeon, a solicitor owes a duty of care to his client even if he provides his services gratuitously.^[519]

They then maintain that:

The essential problem with this idea is that the courts do not require that the defendant manifest any consent to the obligation in question, and the cases do not support the notion that there is any such requirement. Allan Beever claims that the right protected by the tort of negligent misrepresentation ‘is based on the defendant’s consent' and ‘is a right given by the defendant to the claimant’.^[520] But Beever later acknowledges that ‘the issue is not whether the defendant agreed to place herself under a legal obligation’ or ‘agreed to be liable’ but ‘whether the defendant placed herself under a legal obligation by guaranteeing the accuracy of some information, by consenting to do something, etc.’ ^[521] But a person who consents to do something does not thereby place himself or herself under a legal obligation unless he or she makes a commitment and manifests an inten­tion to be legally bound by that commitment. If the defendant only consented ‘ to do something’ and did not manifest any consent to the legal obligation that arises from that conduct then there is nothing distinctive about this category of obligation. It is simply imposed by law on the basis of conduct that is potentially harmful.^[522]

The claim made here could be one of the following three.

The first is correct on its face. The problem is that is relies on an inappropriate understanding of the meaning of ‘ consent’. In this context,^[523] ‘consent' alludes to the work of Peter Birks in particular, where it is proffered as the basis of a legal event that gives rise to legal relationships in many areas of the law, particularly, of course, in the law of contract.^[524] Here, then, the claim that A has consented to do x does not imply that he has merely indicated willingness to do x or the like, it means that a legal event has occurred in which A comes under an obligation to do x by consenting to do it. In this sense of the term, it is not coherent to say that A consents but does not commit himself. The claim that in Hedley Byrne-type cases the defendant’s potential liability ‘ is based on the defendant’s consent’ is to be understood in this fashion.^[525]

The second claim is false. It is not true that one must consent to an obligation for that obligation to exist. If I commit myself to doing x, then I have an obligation to do x. I do not additionally have to consent to the obligation to do x. A promise to do something, for instance, is a commitment to do that thing that creates an obligation. It is not also a commitment to the commitment, a consent to the obli­gation. All that is redundant.

The third claim is true, but it does not advance the argument.

Let us take some examples. Imagine a conversation in which you tell me that you are considering investing in a company. In response, I agree to find out what I can about the company and to provide you with a report. In this report, I tell you that I can guarantee that I have accurate information regarding the company’s financial position and that in my view you ought to invest. Imagine that I am so confident that I say ‘If you invest, I guarantee that you will double your money in five years’. Finally, imagine that there is nothing in the context of this conversation to indicate that I mean anything other than what I say—eg my guarantee is not hyperbole or puffery but is meant literally. But say also that the company is in fact in poor shape and that you suffer significant loss as a result of your subsequent decision to invest.

If we assume, as seems reasonable, that you have a Hedley Byrne action against me here, what is the obligation that I am said to have violated? According to the contract model, it is the obligation that I undertook, guaranteeing that the invest­ment was a good one. Note that it is not the obligation to provide a report. I may not have had any such obligation and, if I had, I did not violate it. Of course, this is an extreme case, likely to occur only infrequently (though it is far from impos­sible). Let us then consider more probable scenarios.

Imagine the same case but where my report lacks the guarantee of profit. Put positively, the report guarantees that I have accurate financial information and recommends that you invest. In this case, the obligation that I violate is the obli­gation to provide accurate information regarding the financial position of the company, an obligation that I expressly undertook.

Imagine that we have the same case but that my report indicates that I have looked into the matter and have discovered that the company is in good shape and that, in the light of that information, I recommend that you invest. What is the obligation here? Normally, it will be that I observed due diligence in obtaining the information and in giving the advice. This is not because I expressly undertook that obligation, but because it is an implied undertaking given my behaviour.^[526]

Perhaps it is at this point that the objection is meant to bite. As this obligation was not expressly undertaken by me, it may be tempting to conclude that ‘It is sim­ply imposed by law on the basis of conduct that is potentially harmful’.^[527] But that position would involve a strange understanding of the nature of communication.

Think of this case in abstraction from the law. Apart from wanting compensa­tion from me, you may be angry to discover that I had not taken sufficient care in preparing my report. You come to me and complain about this. I respond by saying ‘I agreed to prepare the report but I did not promise to do it carefully’. Who would find this reply at all compelling? It is clear that I am being obtuse. In agree­ing to prepare the report, I implicitly agree to do so carefully—unless, of course, I indicate otherwise.^[528] And this is not because something is imposing on me any obligation. It is because that is what my explicit agreement entails. This is what it means to say that the obligation is implied—it is implicit in the express undertak­ing I made. It is not imposed.

Similarly, imagine that I agree to babysit your children while you have an even­ing out. Coming home, you are astonished to see your sons drunk having helped themselves to a large portion of your liquor cabinet and even more shocked to find me playing strip poker and smoking crack with your daughter.

With this in mind, we can turn to the notion that the idea that ‘ The essen­tial problem with [the contract model] is that the courts do not require that the defendant manifest any consent to the obligation in question, and the cases do not support the notion that there is any such requirement'.^[530]

The first thing to be said is that the claim is false. Given what has been said above, it is clear that many courts have required precisely this. But one would certainly be right to maintain that this requirement is far from universal. This is a result of the confused state of the case law. But the crucial point in this context is that, when courts reject the contract model, they replace it with approaches that are in fact best explained by the model that they reject. The problem, in other words, is that we have once again replaced what actually matters with a defective proxy.

I have examined the sometime requirements of skill, purpose, representation, proximity, reasonable foreseeability, reliance and negligence as well as the treat­ment of disclaimers and casual conversations elsewhere. In each case, I show that these are best explained as defective proxies for assumptions of responsibility.^[531] The reader can be directed to those arguments. Suffice it for now to examine one case, perhaps the most confused of all, the decision of the New Zealand Court of Appeal in AG v Carter.

The plaintiffs purchased a vessel, the Nivanga, from a third party. They did so in the light of two interim certificates of survey and one certificate of survey issued by the Marine Division of the Ministry of Transport (‘MOT') in accordance with section 206 of the Shipping and Seamen Act 1952 (NZ).

In Carter, the Court resoundingly rejected the contract model in favour of the negligence model. It also maintained that a duty of care could be owed by the defendant to the plaintiff only if the parties were in a relationship of proximity. That would be the case if and only if the plaintiff's reliance on the defendant's statement was reasonable and reasonably foreseeable.^[534]

The Court concluded that the plaintiffs were not owed a duty of care, as a relationship of proximity did not exist. First, as the certificates were issued for the purpose of ensuring the safety and seaworthiness of the vessel, it was not reasonable for the plaintiffs to rely on them for the different purpose of making investment decisions.^[535] Secondly, the Court maintained that the plaintiffs' reliance was not reasonable, as the relevant sections of the Shipping and Seamen Act 1952 indicate that certificates of survey are prepared for the benefit of passengers and crew of ocean-going vessels, other seafarers and the sea itself. The certificates are not prepared for the benefit of prospective purchasers of vessels.^[536] Hence, in ‘none of the capacities in which the plaintiffs claim to have suffered loss were they the person or within the class of persons who were entitled to rely on the certificates'.^[537]

Rejecting the contract model on the ground that ‘In tort obligations are imposed, not assumed, as they are in contract',^[538] the Court went on:

But, that said, the idea of one person assuming, in the sense of coming under, a respon­sibility to another in tort, does have value when understood in the sense that in certain circumstances the law requires responsibility to be assumed. Responsibility is then deemed in law to have been assumed. The expression ‘ deemed assumption of responsi­bility' conveniently expresses this process of thought and is thus conceptually consistent with the conventional difference between tort and contract.^[539]

This is at least close to double Dutch. The label ‘deemed assumption of responsi­bility' is obscurantist. The phrase ‘the idea of one person assuming, in the sense of coming under, a responsibility' is hard to decipher. A deemed assumption of responsibility is simply a deemed responsibility. It cannot but serve to confuse matters to refer to it as a deemed assumption of responsibility. So why so refer to it? As we will see, it is because a ‘deemed assumption of responsibility' is in fact just an assumption of responsibility. More fully, though the Court explicitly rejected the contract model in favour of the notion of deemed assumptions of responsibility, it propounded a test that entails that a defendant will be deemed to have assumed responsibility when and only when he did in fact assume responsibility.

Moreover, the Court's application of the negligence model is unstable. As indi­cated above, proximity between the parties is held to exist if the plaintiff's reliance on the defendant's statement was reasonable and foreseeable. But the Court adopted an odd view of reasonable reliance. The Court insisted that if a defend­ant makes a statement for one purpose, then any reliance on that statement for a different purpose is unreasonable. Moreover, the Court asserted that it is not rea­sonable for a person to rely on a statement unless the statement was made for the benefit of that person. While this position produces outcomes that fit the settled law, it is not remotely plausible.

It is often perfectly reasonable to rely on statements in the relevant circum­stances. If an auditor prepares a glowing report into a publicly listed company for purposes of the company's AGM, then it is entirely reasonable for me to rely on that report (perhaps not in isolation) when deciding whether to invest in that company. In fact, it would be quite irrational of me not to do so. This is so whether or not I was a shareholder of the company at the time of the AGM and hence whether or not the report was prepared for me.^[540] Every day, people rely on reports of this kind to make similar decisions. People routinely decide to buy cars, for example, on the at least partial basis that they have recently received a Warrant of Fitness, MOT, etc. They are not acting unreasonably when they do this. In Carter, the plaintiffs' reliance was also entirely reasonable.

Similarly, I am familiar with a New Zealand car auction company that prepares mechanical reports for its customers on the state of its vehicles. The report con­tains the phrase: ‘Note: This report is not a warranty. This report is a guide only'. This prevents Hedley Byrne-type liability arising in relation to these reports, but that cannot be because a customer's decision to rely on the report would be unrea­sonable. If that were so, what could the point of preparing the reports be?

In response to related concerns, La Forest J in London Drugs Ltd v Kuehne & Nagel International Ltd distinguished ‘mere reliance in fact and reasonable reli­ance on the employee's pocket-book'.^[541] Hence, for La Forest J, the question was not whether it was reasonable for the plaintiff to rely on the statement, but whether it was reasonable for the plaintiff to expect the defendant to compensate her for her, perhaps reasonable, detrimental reliance on the statement. Perhaps, then, it is pos­sible to concede that the plaintiffs' reliance in Carter and in the hypothetical cases discussed was reasonable in fact, but maintain that it would not be reasonable for the plaintiffs to rely on the defendant's pocket-book.

The problem is that there is a mere illusion of an argument here. The claim is that, while it may have been reasonable for the plaintiff to have relied on the defendant's statement, it is not reasonable for the plaintiff to expect to be com­pensated by the defendant for the consequences of that reliance. But we need to know why that expectation is unreasonable. Given that the plaintiff's reliance was reasonable in fact, the reasonable foreseeability of the plaintiff's injury and the commitment to the negligence model, it is obscure why the expectation of com­pensation is unreasonable. Nor is it any better to say that recovery would not be fair, just or reasonable. Why would it not be?

In fact, the plaintiff's expectation of compensation is unreasonable because the defendant did not assume responsibility for the plaintiff's reliance on the state­ment in the first place. Hence, it is unreasonable for a plaintiff to expect compen­sation for detrimental reliance on a statement for a purpose other than that for which the statement was made, because the defendant did not assume responsi­bility to the plaintiff for that reliance. Similarly, it is unreasonable for a person to expect compensation for detrimental reliance on a statement if that statement was not made for her benefit, because the defendant did not assume responsibility to the plaintiff at all. In Carter, though the plaintiffs' reliance on the MOT's reports was perfectly reasonable—it would have been irrational for them not to rely—it was unreasonable for them to expect compensation from the defendant because the MOT did not assume responsibility to the plaintiffs as to the quality of the vessel; in fact, the MOT did not assume responsibility to the plaintiffs at all.

It unnecessarily complicates the enquiry and disguises the nature of the action to base liability on a restrictively defined reasonable and foreseeable reliance or to distinguish ‘mere reliance in fact and reasonable reliance on the employee's pocket­book' when the question is simply whether the defendant assumed responsibility to the plaintiff for the plaintiff's reliance. Though the Court of Appeal insisted on the negligence model, it did not apply it in fact. Rather, the actual decision in the case, as well as the more general rules concerning limitations on the duty of care, are based on the contract model. Again, the defective proxy is hiding the basis of this action.^[542]

VIII.