LEGAL EDITING by Ava Chisling

LQR Cover 1-08:v

22/11/07

13:51

Page 1

January 2008

NOTES Character, credibility and truthfulness Richard III in New Zealand Excluding fiduciary duties: the problem of investment banks Unanticipated fiduciary liability

1 6 15 21

Equitable third party liability Final guidelines on compensation for commercial agents When is a risk of injury foreseeable? Refining the duty of care in Singapore

31 37 42

46 72 103 132

REVIEWS AND NOTICES Loughlin and Walker: The Paradox of Constitutionalism Goodwin-Gill and McAdam: The Refugee in International Law Gordley: Foundations of Private Law

160 163 166

Blair and Walker: Financial Markets and Exchanges Law Leverick: Killing in Self-Defence Smith: The Law of Assignment

170 172 175

ISSN 0023-933X Computerset by Laserwords Private Ltd, Chennai, India Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall No natural forests were destroyed to make this product; only farmed timber was used and re-planted

Vol.124 pp.1–180 January 2008

ARTICLES Legislation that would “Preserve” the Common Law: the Case of the Declaration of Intention Explaining Resulting Trusts Causation and Risk in the Highest Courts of Canada, England and France Deception, Mistake and Vitiation of the Victim’s Consent

THE LAW QUARTERLY REVIEW

Volume 124

Copies of articles from The Law Quarterly Review, together with other Sweet & Maxwell publications, can be obtained from Sweet & Maxwell (Yorkshire). Rates are £8.45 per case or article for orders by post. There is an additional charge for fax delivery of £1.15 per page. Please address any order to: Sweet & Maxwell (Yorkshire), The Hatchery, Hall Bank Lane, Mytholmroyd, Hebden Bridge, West Yorkshire HX7 5HQ. Tel: 01422 888000. Fax: 01422 888001. DX 715966 HEBDEN BRIDGE 2. Email: docdel@sweetandmaxwell.co.uk

100 Avenue Road, Swiss Cottage, London NW3 3PF AUSTRALIA: LBC INFORMATION SERVICES, SYDNEY CANADA AND U.S.A.: CARSWELL, TORONTO

Thomson, Sweet & Maxwell

*209743*

Volume 124

January 2008

Legislation that would “Preserve” the Common Law: the Case of the Declaration of Intention RODERICK MUNDAY Explaining Resulting Trusts WILLIAM SWADLING Causation and Risk in the Highest Courts of Canada, England and France PROFESSOR LARA KHOURY Deception, Mistake and Vitiation of the Victim’s Consent REBECCA WILLIAMS

LQR Cover 1-08:v

22/11/07

13:51

Page 2

THE LAW QUARTERLY REVIEW The Law Quarterly Review is regarded as the leading scholarly journal in its field in the English language. All contributors, whether Notes, Articles or Reviews, are selected as being of the highest quality, providing authoritative and innovative critical analysis of current legal issues in litigation, statute law, academic research, and commerce and industry.

Editor

F. M. B. REYNOLDS, Q.C. (HON.), D.C.L., F.B.A. Professor of Law Emeritus in the University of Oxford; Emeritus Fellow of Worcester College, Oxford

Assistant Editor D. D. PRENTICE, M.A., LL.B., J.D. Allen & Overy Professor of Corporate Law in the University of Oxford; Fellow of Pembroke College, Oxford

Book Review Editor W. E. PEEL, B.C.L., M.A. Fellow of Keble College, Oxford

Editorial Advisory Committee H.H. P. V. BAKER, Q.C. THE HON. SIR JACK BEATSON, D.C.L., F.B.A. THE RT. HON. SIR LAWRENCE COLLINS, LL.D., F.B.A.

S. M. CRETNEY, Q.C. (HON.), D.C.L., F.B.A.

THE RT. HON. LORD BINGHAM OF CORNHILL, D.C.L. (HON.), F.B.A. (HON.) PROFESSOR SIR NEIL MACCORMICK, Q.C., LL.D., F.B.A., F.R.S.E., M.E.P.

PROFESSOR C. F. H. TAPPER, B.C.L., M.A.

Articles, notes and correspondence should be addressed to Professor F. M. B. Reynolds, 61 Charlbury Rd, Oxford OX2 6UX, England : fax (+44) 1865–511894. Books for review should be addressed to W. E. Peel, Keble College, Oxford OX1 3PG, England : fax (+44) 1865–272742. Orders to Customer Service Department, Sweet & Maxwell Ltd, 100 Avenue Road, London NW3 3PF. http://www.sweetandmaxwell.co.uk/lqr

Each article, casenote and review in this issue has been allocated keywords from the Legal Taxonomy utilised by Sweet & Maxwell to provide a standardised way of describing legal concepts. These keywords are identical to those used in Westlaw UK and have been used for many years in other publications such as Legal Journals Index. The keywords provide a means of identifying similar concepts in other Sweet & Maxwell publications to which keywords from the Legal Taxonomy have been applied. The keywords have also been used to determine the main entries in the index in order to provide a much higher level of consistency between indexes in Sweet & Maxwell publications. Keywords will follow the taxonomy logo as a footnote at the end of each article, casenote or review. Suggestions to sweetandmaxwell.taxonomy@thomson.com

© Sweet & Maxwell Limited and contributors 2008 The Thomson trademark and Star Design are trademarks of Thomson Financial S.A. used herein under licence All rights reserved. Crown Copyright legislation is reproduced under the terms of Crown Copyright Policy Guidance issued by HMSO. No part of this publication may be reproduced or transmitted in any form or by any means, or stored in any retrieval system of any nature without prior written permission, except for permitted fair dealing under the Copyright, Designs and Patents Act 1988, or in accordance with the terms of a licence issued by the Copyright Licensing Agency in respect of photocopying and/or reprographic reproduction. Application for permission for other use of copyright material including permission to reproduce extracts in other published works shall be made to the publishers. Full acknowledgment of author, publisher and source must be given.

THE LAW QUARTERLY REVIEW Volume 124

January 2008 NOTES CHARACTER, CREDIBILITY AND TRUTHFULNESS

THE law of evidence classically divides material into that which is admissible by virtue of its relevance to some issue in the case, and that admissible because of its relevance to the credibility of some witness or statement. Though doubts have sometimes been expressed about the utility or good sense of the distinction, both by judges (see, e.g. R. v Nagrecha [1997] 2 Cr. App. R. 401) and by academic lawyers (see, e.g. Zuckerman, Principles of Criminal Evidence (1989), at pp.95–97), it seems to be so firmly embedded in the law as to be immoveable. An abiding problem with the old law relating to cross-examination of the accused about his bad character, under the second part of what finally became s.1(3)(ii) of the Criminal Evidence Act 1898, was that such a broad view was taken of relevance to credibility that the required direction to the jury, to treat it as going only thereto, and not at all to the issue, might be thought to amount to “gibberish” (see Cross (1969) 6 Syd. L. Rev. 173 at 182). The jury would solemnly be told that the accused, charged with an offence of violence, was not to be treated as more likely to have committed the offence by reason of his string of convictions for similar offences, which offences, the judge would go on to say, went solely to the credibility of his claim that it had been the complainant who had attacked him. There were two possible solutions to this problem. One would be to get rid of the direction, perhaps even encouraging the jury to make use of the evidence, as going directly to the issue. However, since, in the premisses, the common law similar fact rules must have denied the prosecution the use of such evidence, on the issue, as part of its own case, the result would then be that evidence failing to get through the front door would enter freely through the back. The second solution would be to take a more coherent view of what items of bad character really do go to (in)credibility of the accused’s evidence. That would rule out offences other than those 1

Law Quarterly Review

[Vol. 124

of dishonesty, but might rule out even such offences where not involving any element of deception (see Munday [1985] C.L.J. 62 and [1986] Crim. L.R. 511). Or, one might adopt both solutions. Under the new provisions of the Criminal Justice Act 2003 that deal with the accused’s bad character, the first solution had been quickly embraced. In R. v Highton [2005] EWCA Crim 1985; [2005] 1 W.L.R. 3472, Lord Woolf C.J. put it as follows (at [14]): “. . . a distinction must be drawn between the admissibility of evidence of bad character, which depends upon it getting through one of the gateways, and the use to which it may be put once it is admitted. The use to which it may be put depends upon the matters to which the evidence is relevant, rather than upon the gateway through which it was admitted.” (emphasis in original) Highton itself was concerned with evidence that was admissible via s.101(1)(g) of the Act, because the accused had “made an attack on another person’s character”. Therefore, it was directly on the point at hand. It no longer mattered that the evidence in question would not have got through the issue gateway (s.101(1)(d)) under its own steam. As long as (barely) legally relevant to an issue between prosecution and accused, it was perfectly permissible for the jury to use it in that way. Furthermore, cases directly concerned with s.101(1)(g) relevance rather suggested that the very wide view of credibility taken under the 1898 Act was equally to be taken under the new Act. The best example is R. v Nelson [2006] EWCA Crim 3412, where the Court of Appeal saw no reason to interfere with the judge’s decision to admit evidence of the accused’s convictions for drugs offences, on charges of affray and assault. In R. v Lawson [2006] EWCA Crim 2572; [2007] 1 W.L.R. 1191, though the case itself was concerned with s.101(1)(e), where a co-accused whose defence has been undermined by the accused seeks to call evidence of the latter’s bad character, in order to attack his credibility, Hughes L.J. took a very wide view of that concept. The co-accused was entitled to adduce evidence that the accused was “unscrupulous and/or otherwise unreliable”, which qualities were, “capable of being shown by widely differing conduct, ranging from large scale drug or people-trafficking via housebreaking to criminal violence” (see [34]). Though his Lordship expressly confined his remarks to the accused/co-accused case, it would have surprised nobody to read similar words in relation to s.101(1)(g). A very different view of the accused’s credibility has emerged in a separate strand of authority, concerned with s.103(1)(b) of the Act. Though the story begins with certain influential dicta of Rose L.J. in R. v Hanson [2005] EWCA Crim 824; [2005] 1 W.L.R. 3169 (at [13]), it has been taken a great deal further in the very important case of R. v Campbell [2007] EWCA Crim 1472. (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

Section 103(1)(a) and (b) render certain matters per se in issue between the prosecution and the accused, for purposes of s.101(1)(d). In general, the question whether the accused has a propensity to commit offences of the kind with which he is charged is put in issue by s.103(1)(a), whilst the question whether the accused has a “propensity to be untruthful” is put in issue by s.103(1)(b). Thus, s.103(1)(b) seems to envisage the prosecution calling evidence to attack the credibility of the accused, as part of its own case, and before any evidence at all has been called by the defence. It is not only in that subsection that the Act rejects the evidential logic embraced by R. v Butterwasser [1948] 1 K.B. 4, but it is there that it is rejected in the most stark of terms. At root, Butterwasser stands for the proposition that we cannot conceive of evidence being adduced to attack a witness’s credibility unless and until there is some evidence from that witness for him or her to be (in)credible about. That s.103(1)(b) does not declare “open season” on the accused with convictions is one effect of Rose L.J.’s dicta in Hanson. There, his Lordship focused on the word “untruthful”. Previous convictions, whether for offences of dishonesty or otherwise, are likely to show a propensity to be untruthful only where (see [13]): “. . . in the present case, truthfulness is an issue and, in the earlier case, either there was a plea of not guilty and the defendant gave an account, on arrest, in interview, or in evidence, which the jury must have disbelieved, or the way in which the offence was committed shows a propensity for untruthfulness, for example, by the making of false representations.” In several later cases, the logic of his position was applied in rulings that convictions for violence or robbery, though relevant to the issue, could not be relevant to show untruthfulness or lack of credibility (see R. v Meyer [2006] All E.R. (D) 472 (Mar); R. v Awaritefe [2007] EWCA Crim 706; R. v McDonald [2007] EWCA Crim 1194). Campbell takes things a long step further, in a way that is capable of having an effect no less on s.101(1)(g) than on s.101(1)(d). The facts of the case were straightforward enough. Campbell was accused of falsely imprisoning a woman with whom he was having a sexual relationship. At trial, the prosecution relied on s.101(1)(d), as allowing it to adduce evidence of two convictions, each involving violence against a woman who had been, at the time, his girlfriend. No complaint was made, on appeal, about the admission of those convictions, which was clearly based upon their issue relevance. Later in the trial, the accused himself had given evidence of the incorrectness of the complainant’s claims, saying that he had never done anything to restrain her freedom of movement, nor had there been any violence. That evidence would seem to have opened him up to the possibility of cross-examination under s.101(1)(g), since (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

it constituted an attack on her character, yet the prosecution did not, it seems, seek to invoke that subsection. When the judge came to direct the jury, in accordance with the terms of the relevant Judicial Studies Board specimen direction, he told them that they might take account of the convictions, not only as showing his propensity to be violent towards women, but also as tending to suggest that his evidence might be untruthful. It was to this aspect of the case that objection was taken on appeal. The argument put to the Court of Appeal was that, the evidence having been admitted through s.101(1)(d), as being relevant to the issue, it was wrong, in principle, for the jury to be told that they might use it as relevant to credibility. The court gave short shrift to that argument, as falling foul of the holding in Highton (see above). So, just as evidence admissible through s.101(1)(g), as relevant to credibility, may be used for any relevance that it has to the issue, so evidence getting through s.101(1)(d), as relevant to the issue, may be used for any credibility relevance it has. However, that was, by no means, the end of the case. Lord Phillips of Worth Matravers C.J. pointed out that s.103(1)(b) merely tells us that the accused’s propensity to be untruthful is (generally) in issue between the parties. Yet, s.101(1)(d) requires relevance to an important matter in issue between them, neither of which requirements, he suggested, would, in the ordinary case, be met. In his words (see [30]): “The question of whether a Defendant has a propensity for being untruthful will not normally be capable of being described as an important matter in issue . . . A propensity for untruthfulness will not, of itself, go very far to establishing the commission of a criminal offence. To suggest that a propensity for untruthfulness makes it more likely that a Defendant has lied to the jury is not likely to help them. If they apply common sense they will conclude that a Defendant who has committed a criminal offence may well be prepared to lie about it, even if he has not shown a propensity for lying whereas a Defendant who has not committed the offence charged will be likely to tell the truth, even if he has shown a propensity for telling lies. In short, whether or not a Defendant is telling the truth to the jury is likely to depend simply on whether or not he committed the offence charged. The jury should focus on the latter question rather than on whether or not he has a propensity for telling lies.” So, it now seems that even previous convictions, such as those for perjury, that demonstrate a very specific propensity to lie at trial, fail to pass muster. Paying attention to such a past record serves only to divert the jury from the key question of guilt or innocence in the case at hand. One might conclude that the effect of his Lordship’s reasoning is to render (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

s.103(1)(b) a dead letter. Indeed, in the very next paragraph, he seems clearly to admit as much (see [31]): “For these reasons, the only circumstance in which there is likely to be an important issue as to whether a Defendant has a propensity to tell lies is where telling lies is an element of the offence charged. Even then, the propensity to tell lies is only likely to be significant if the lying is in the context of committing criminal offences, in which case the evidence is likely to be admissible under s. 103(1)(a).” Even though it followed that the trial judge had been wrong to direct the jury that they might take into account Campbell’s previous convictions as going to his (un)truthfulness, the court ruled that his present convictions were not, on the particular facts, unsafe. Lord Phillips C.J.’s conclusion may be considered surprising, in terms of statutory interpretation, even by those who think s.103(1)(b) the most unsatisfactory provision in the whole of this part of the 2003 Act. However, its potential for broader application may be accounted of even greater importance. If learning about the accused’s convictions for offences involving elements of deception is unhelpful to the jury when s.101(1)(d) has been employed, why should it be any more helpful when s.101(1)(g) has been invoked? As we have seen, Campbell was itself a case where the prosecution might have sought to rely upon s.101(1)(g), yet all that was involved was a clash of evidence between the complainant and the accused. Towards the end of the judgment, Lord Phillips C.J. offered a direction that he thought might have been more helpful to the jury than the one that they were actually given (see at [44]). They might properly have been told that the accused was saying that the complainant’s story was “all a pack of lies”, and that they could convict only if sure that she had been telling the truth. In deciding that question, they might be assisted by their knowledge of the accused’s tendency to use violence on his girlfriends. In other words, it was because his record demonstrated a particularised propensity to violence that the jury might find the complainant’s story more credible than the accused’s. One gets to the credibility of the witnesses via evidence that is relevant to the issue, not the other way around. From that perspective, it would seem that it would have made no difference had Campbell indeed had convictions for perjury or for obtaining by deception. Such an interpretation of the proper use of s.101(1)(g) would not render it a dead letter. Contrast a case where the accused had mounted a direct attack on the (un)truthfulness of the complainant, via extraneous evidence. What if she had convictions for perjury or deception offences, and the defence sought leave, under s.100(1)(b), to ask her about them? Then, the accused would be backing up his claim that she was lying now, by reference to her demonstrated propensity to lie, so the evidential basis (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

for his application would be broadly equivalent to that embraced by s.103(1)(a). If leave were granted, on that basis, it would be strange indeed if the jury were to be denied knowledge of the accused’s own convictions for offences showing resort to untruthfulness. He would deliberately have set a distinct credibility issue in motion, so could hardly complain about the “tit for tat” nature of the response. Furthermore, the law would have firmly adopted the second solution to the problem adverted to at the beginning of this note, as well as the first. PETER MIRFIELD.* RICHARD III

NEW ZEALAND

OF the five judges of the New Zealand Supreme Court who decided Wholesale Distributors Ltd v Gibbons Holdings Ltd [2007] NZSC 37 four took the position that, when the meaning of a written contract is in dispute, the court can have regard to evidence of how the parties acted after the contract was made. The Supreme Court refused to follow the English rule that such evidence inadmissible: see James Miller & Partners Ltd v Whitworth Street Estates (Manchester) Ltd [1970] A.C. 583 and F.L. Schuler AG v Wickman Machine Tool Sales Ltd [1973] UKHL 2; [1974] A.C. 235. The Supreme Court’s position contrasts with recent Australian cases which have followed the English rule: Commercial Capital v Durman [2007] NSWSC 869 at [40]; Ku v Song [2007] FCA 1189 at [53]; Council of the City of Sydney v Goldspar Australia Pty Ltd [2006] FCA 472 at [164]. The exclusionary rule is subject to a number of qualifications: evidence of post-contract conduct may be admitted (i) where the contract is oral, or partly written and partly oral: Carmichael v National Power Plc [1999] UKHL 47; [1999] 1 W.L.R. 2042; Maggs v Marsh [2006] EWCA Civ 1058; [2006] B.L.R. 395 at [25]–[26]; Glendale Managed Services v Graham [2003] EWCA Civ 773; [2003] I.R.L.R. 465 at [19]; (ii) where there is a boundary dispute and the parcels clause in the conveyance, or the plan attached thereto, is unclear: Ali v Lane [2006] EWCA Civ 1532; [2007] 1 P. & C.R. 26 at [21]–[36]; Haycocks v Neville [2007] EWCA Civ 78 at [31]; (iii) where the issue is not one of interpretation, but whether a document or a particular clause in a document is a sham: see two important judgments of Arden L.J., Properties Ltd v Dunsford [2001] EWCA Civ 528; [2001] 1 W.L.R. 1369 at [44] and Hitch v Stone (Inspector of Taxes) [2001] EWCA Civ 63; [2001] S.T.C. 214 at [65]; (iv) where the issue is whether a term was incorporated into a contract: Great North Eastern Railway Ltd v Avon Insurance Plc [2001] EWCA Civ 780; * Jesus College, Oxford. Admissibility; Character; Civil evidence; Credibility

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

[2001] 2 Lloyd’s Rep. 649 at [29]; or whether a contract was formed: Tomko v Palasty [2007] NSWCA 258 at [67]; (v) where rectification is claimed; or (vi) to support an argument that a contract’s terms have been varied, enlarged or waived, or an argument that the parties made a new contract, or to found an estoppel by convention. In his article, “My Kingdom for a Horse: the Meaning of Words” (2005) 121 L.Q.R. 577, Lord Nicholls of Birkenhead argued that English law should abandon the rule which excludes evidence of the parties’ postcontract conduct, as well as that which generally excludes evidence of their pre-contract negotiations where the issue is one of interpretation (as opposed to a claim for rectification). Lord Nicholls’ article was cited by two members of the Supreme Court in Wholesale Distributors v Gibbons; Thomas J., who delivered the leading judgment, described the article as “seminal”. Thomas J. also took the opportunity to reassert his position in Yoshimoto v Canterbury Golf International Ltd [2000] NZCA 50; [2001] 1 N.Z.L.R. 523 that evidence of pre-contractual negotiations should be admissible. Thomas J. may have intended to counteract the effective rejection of that position by the New Zealand Court of Appeal in Potter v Potter [2003] N.Z.L.R. 145. The Court of Appeal’s judgment was upheld by the Privy Council [2004] UKPC 41; [2005] 2 N.Z.L.R. 1, although without express reference to the rule excluding evidence of pre-contractual negotiations. The issue in Wholesale Distributors v Gibbons was the meaning of a covenant in an assignment of a sub-lease of a warehouse and land, the covenant being made by the assignee in favour of Gibbons Holdings Ltd (the “plaintiff”). The head lease had been granted to the plaintiff in 1982. The initial term of the head lease expired on October 31, 2002, but the plaintiff had a right of renewal for a term to end on October 31, 2023. In 1991, the plaintiff granted the sub-lease to GUS Properties Ltd (the “sub-lessee”). The parties would have liked the sub-lease to be for a term ending in October 2010, but when the sub-lease was granted, the head lease had not been renewed. If therefore the sub-lease had been granted until October 2010, that would have been longer than the then existing term of the head lease, and the sub-lease would have taken effect as an assignment of the head lease. So the sub-lease was structured as follows: there was (i) an actual grant of a term to end on October 30, 2002 (one day before the end of the initial term of the head lease), and (ii) an agreement to enter into a new sub-lease for a term commencing on November 2, 2002 and ending on October 31, 2010. In 1997 Wholesale Distributors Ltd (the “defendant”) agreed to purchase the business and assets of the sub-lessee. As part of that transaction, the sub-lease was assigned by the sub-lessee to the defendant. The parties to the deed of assignment, dated April 3, 1997, were the sublessee, the defendant and also the plaintiff which, as sub-lessor, gave (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

its consent to the assignment. The deed of assignment was based on a standard precedent of the Auckland District Law Society evidently intended for the assignment of a head lease rather than a sub-lease. The clause in issue was as follows: “The Assignee [i.e. the defendant] covenants with the Landlord [i.e. the plaintiff] that from the Date of Assignment and during the remainder of the term of the lease the Assignee will pay the rent provided for in the Lease and keep and perform all the covenants in the Lease.” The deed of assignment identified “the Lease” as “the Deed of Lease dated 10 July 1991”, i.e. the sub-lease. Against the printed heading “Expiry Date of Lease” contained in the First Schedule were inserted the words “31 October 2002 with a New Lease being granted for a term expiring 31 October 2010”. In 2000, the defendant assigned its interest in the sub-lease to a company called Infogate. On October 18, 2002, just a fortnight before the renewal of the sub-lease was due to take effect, Infogate went into receivership, and a few months later it was in liquidation. The plaintiff sought to enforce the defendant’s covenant in the assignment, quoted above, to pay the rent provided for in the lease and to perform all the covenants in the lease. Those covenants, the plaintiff said, included the covenant to enter into a new sub-lease for a term from November 2, 2002 until October 31, 2010. The defendant’s reply was simple. Its covenant was expressed to be “during the remainder of the term of the lease”; the lease was defined as “the Deed of Lease dated 10 July 1991”; and the term of that lease had ended on October 31, 2002. This defence failed. The Supreme Court construed “during the remainder of the term of the lease” as referring to the expiry date of the new lease provided for by the 1991 sub-lease. The court was able to arrive at this conclusion by classic methods of interpretation of the kind employed by Lord Wilberforce in Prenn v Simmonds [1971] 1 W.L.R. 1381. But Thomas J. went further. He stated that his interpretation of the covenant in the 1997 deed of assignment was confirmed by the terms of the assignment to Inofogate which the defendant had executed in 2000 and also by the terms of a sub-lease of part of the premises which the defendant had granted in 1999. Although Thomas J. was the only member of the court who considered it necessary or relevant to have regard to those documents, Elias C.J. and Tipping and Anderson JJ. agreed with the general proposition that evidence of post-contract conduct is admissible; Blanchard J. reserved his position. In this part of his judgment, Thomas J. drew support from the writings of Professor David McLauchlan and also from Lord Nicholls’ “My Kingdom for a Horse” article. The latter article, Thomas J. said, had given (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

“an exponential boost” to the case for the admission of the evidence of the actual intention of the parties. That may be an understatement. Lord Nicholls stopped short, by a nanometre, of directly challenging the notion that the court’s function is to determine the parties’ presumed intent. But his view that pre-contractual negotiations and post-contract conduct should be taken into account makes little sense unless the court is trying to ascertain the parties’ actual intentions. Lord Hoffmann has said that there is “no window into the mind” of the parties: Kirin-Amgen Inc v Hoechst Marion Roussel Ltd [2004] UKHL 46; [2005] R.P.C. 9 at [32]. However, Lord Nicholls evidently considers that there is a rear window (prior negotiations) and a front window (post-contract conduct) and that the courts should look through both of them. According to Thomas J., it is not the business of a court to ascertain the “presumed intent” of the parties unless the effective issue is which party should bear a loss caused by an event which the parties never thought about. Outside cases of that kind, “the proper task of the court” Thomas J., said, “is to construe the contract objectively with the aim of discerning the actual intention of the parties. When the courts approach a fork in the road, and one road leads to a presumed intent based on the meaning of words and the other road leads to the meaning actually intended, the courts must take that road which leads to the parties’ actual intention.” In the present case, there was no room for the doctrine of presumed intent, “and the habit of thought or attitude of mind which goes with it.” A similar position was taken by Tipping J., with the qualification that he did not consider the conduct of one party admissible: “the parties’ shared conduct will be helpful in identifying what they themselves intended the words to mean. That, after all, must be the ultimate determinant.” Having quoted Lord Reid’s statement in James Miller v Whitworth Street Estates that admitting post-contract evidence “might have the result that a contract meant one thing the day it was signed, but by reason of subsequent events meant something different a month or a year later”, Tipping J. commented: “The subsequent course of English law was shackled by this rare case of even Homer nodding.” Tipping J. was evidently influenced by the fact that Art.2-208 of the Uniform Commercial Code and para.202(4) of the Restatement, Second, Contracts in the United States allow consideration to be given to specified subsequent conduct of both parties. However, both these provisions are of very narrow application. The contract has to involve “repeated occasions for performance by either party”, and only “a course of conduct accepted or acquiesced in without objection” is relevant to interpretation. In such circumstances, an English court would probably find that there had been a variation by conduct or a waiver. In the course of his judgment, Thomas J. referred to the present writer’s note ((2006) 122 L.Q.R. 354, in response to Lord Nicholls’ article. Thomas J. thought that note was arguing that a contract should (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

be construed without regard to its factual background. That has never been the writer’s view. In National Westminster Bank Plc v Spectrum Plus Ltd [2005] UKHL 41; [2005] 2 A.C. 680 at [145], Lord Walker of Gestingthorpe expressed agreement with the criticism of Slade J.’s interpretation of a debenture in Siebe Gorman & Co Ltd v Barclays Bank Ltd [1979] 2 Lloyd’s Rep. 142 which the present writer had made in an article some two years before the ICS case. That criticism, as summarised by Lord Walker, was that Slade J.’s construction was “based too much on linguistic considerations, in isolation from the matrix of facts in which the security was created”. Thomas J. observed that the “primary attack” of the present writer’s 2006 note was directed at the first of the five principles enunciated by Lord Hoffmann in Investors Compensation Scheme Ltd v West Bromwich Building Society [1997] UKHL 28; [1998] 1 W.L.R. 896 at 912, and that Lord Nicholls’ article “merely served to provoke his [the present writer’s] attack on that basic principle”. This is not entirely correct. There were serious concerns about Lord Nicholls’ article, to some extent prompted by the experience in the years following Pepper v Hart [1992] UKHL 3; [1993] A.C. 594 until the retreat in the face of Lord Steyn’s attack: (2002) 21 O.J.L.S. 59. During that period, it was thought that a solicitor’s duty of care was such that, if there was an important issue about the meaning of a section, he was bound to go through the Hansards reporting the debates on the clause to see if there was anything relevant. This was usually done, for example, if there was a worry that a transaction might be caught by a particular section of the companies or tax legislation. Little reliance was placed on Lord Browne-Wilkinson’s statement that Hansard should not be used unless the section was ambiguous or obscure, since those conditions were treated by the courts in a relatively relaxed manner: Vogenauer (2005) O.J.L.S. 629 at 635 and 639–641. Within a few years, what Lord Browne-Wilkinson had intended as a sensible adjustment to an exclusionary rule dating back to the mid-18th century was generating a minor industry: elaborate courses were (and are still) organised on “How to do Pepper v Hart research”. The concern about Lord Nicholls’ article was that it would lead to a similar situation: it would become unsafe, or even negligent, to advise on the meaning of a contract without checking what had been written and said during the negotiations. It was notable that, in the cases which Lord Nicholls cited, the pre-contractual negotiations had been much simpler than is now common. ICS 1 asserts: “Interpretation is the ascertainment of the meaning which the document would convey to a reasonable person having all the background knowledge which would reasonably have been available (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

to the parties in the situation in which they were at the time of the contract.” But ICS 3, following Prenn v Simmonds, excludes evidence of precontractual negotiations from the admissible background, although with the qualification that the boundaries of the exclusion are in some respects unclear (as to which see, for example, KMPG LLP v Network Rail Infrastructure Ltd [2007] EWCA Civ 363; [2007] Bus. L.R. 1336 at [38]–[43] and Chartbrook Ltd v Persimmon Homes Ltd [2007] EWHC 409 (Ch); [2007] 1 All E.R. (Comm) 1083 at [22]–[45]). The ICS 3 exclusion has been repeated by Lord Bingham of Cornhill in Alexiou v Campbell [2007] UKPC 11 at [15]; note also Berkeley Community Villages Ltd v Pullen [2007] EWHC 1330 (Ch) at [44]–[56]. In Chartbrook Ltd v Persimmon Homes Ltd Briggs J. noted that Lord Hoffmann did not spell out the policy reasons for ICS 3; nor had the researches of counsel identified any authoritative identification of those reasons. Briggs J. therefore resorted to the summary of the reasons for the exclusionary rule which Lord Nicholls gave in his article, before rejecting them. Of those reasons, Briggs J. considered that the adverse effect on third party rights was “compelling”. Briggs J. said this: “Most modern commercial contracts are to a greater or lesser extent assignable, such that the benefit or burden of the obligations therein contained can be, and often is, transmitted to third parties who took no part in the negotiation of the contract, and who may therefore be assumed to be wholly ignorant of what took place. . . . Furthermore, even in the absence of assignment, the rights and obligations created by a commercial contract may form an important part of the assets and liabilities of one or more of its parties, such that the reporting and auditing of its financial health may be dependent upon a proper understanding of its terms, again by persons with no participation in, or knowledge of, its negotiation. If the parties’ negotiations were, to the extent ‘helpful’, to be routinely admissible as an aid to contractual construction, then no such third parties reading, dealing with or having transferred to them rights or obligations under the contract could make any safe assumptions about its meaning without themselves carrying out an inquiry as to those negotiations, so as to put themselves in the same state of knowledge as the parties to the contract. Furthermore, since ambiguity is no longer (after ICS ) a prerequisite for recourse to the admissible background, a third party’s appreciation of the apparently unambiguous meaning of a word, phrase or term could be subverted by reference to the original parties’ negotiations, without which no secondary meaning was even capable of being guessed at.” (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

Briggs J. then had to explain why evidence of negotiations is admissible if rectification is claimed. (In the New Zealand Supreme Court, Tipping J. pointed out (at [51]) the difficulty of adopting different evidential rules for rectification and interpretation.) Briggs J.’s explanation was that rectification is an equitable remedy and “as an equity, it does not invade the legal rights of bona fide purchasers without notice”. With great respect, this seems correct only where the third party has acquired a proprietary interest in the subject-matter of the contract. For example, a tenant cannot claim rectification of the tenancy agreement as against a bona fide purchaser of the reversion: Smith v Jones [1954] 1 W.L.R. 1089; and it seems that the landlord could not claim rectification as against a person who had purchased the leasehold interest. However, take a simple case of a contract, creating no proprietary interest, in which A promises B to pay him £20,000; B assigns his rights to C. If A had, as against B, an equity to have the contract rectified to state the sum payable as £15,000, A can assert that equity as against C; s.136(1) of the Law of Property Act 1925 states that an assignee takes subject to equities. Section 136 (and its predecessor, s.25(6) of the Judicature Act 1873) was intended only to give an assignee the right to sue in his own name. He would acquire rights of action in courts of law similar to the rights of suit which he already possessed in the courts of equity. So, for example, an assignee of a debt takes subject to any equitable set-off which the debtor could have asserted against the assignor, provided that the right to set off arose before the debtor received notice of the assignment. It makes no difference that the assignment is a legal assignment complying with s.136 and that the assignee had no notice of the debtor’s equitable set-off: Derham, Set-Off, 3rd edn (2003), at paras 17–04 and 17–18. Given that an assignee has the risk of the contract being rectified adversely to him on the basis of pre-contractual negotiations of which he was unaware, what can be the objection to his being exposed to an adverse interpretation of the contract based on those same negotiations? There seem to be two answers. Cases of rectification are quite rare; it would be another matter if the law were such that pre-contractual negotiations normally had to be taken into account when interpreting a contract. It is also arguable that the law should be changed to protect a bona fide assignee for value against rectification (except where the contract assigned is an adhesion contract with a consumer). The main reason is the importance which assignments of choses in action now have in the financial markets. Briggs J.’s judgment makes it appear that the problem is to find a justification for the exclusionary rule. But the real problem is that ICS 1 casts the admissible background so wide that a specific, and seemingly anomalous, exclusion for negotiations then becomes necessary. Much of the problem is caused by the repeated use of the expression “the parties”. The immediate reaction is to imagine two individuals sitting on opposite (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

sides of a table. But take even a reasonably straightforward commercial agreement, such as a joint venture agreement or an agreement by which a company sells off one its subsidiaries. Each party is likely to have its own team of executives—a commercial person, a finance expert, an inhouse lawyer, a tax specialist and so on. These will be supplemented by teams from the law firms, and perhaps other specialists, such as investment bankers and accountants. When the contract has to be interpreted, some of these individuals may no longer be on the scene; they may have moved to different companies or retired. The relevant party may have been taken over, and the people who now want to understand what the contract means may be new management put in by the acquirer. In Pemsel’s case [1891] A.C. 531, the House of Lords was faced with the problem of how the expression “trust for charitable purposes” as used in the Income Tax Acts was to be interpreted. The Acts applied on both sides of the border, and the legal meaning of “charity” in Scotland was different from that in England. The solution devised by the Court of Session was that the term had to be given its popular and ordinary meaning. In the House of Lords Lord Macnaghten disagreed. The general principle was that, in construing Acts of Parliament, words must be taken in their legal sense. He continued as follows: “Then I ask, to whom is the Act speaking? In one sense, no doubt, it is speaking to all concerned. But it is addressed, I think, specially to that body under whose ‘cognizance and supervision’, to use the words of the Act, these particular allowances and exemptions are placed. All these applications for allowances and exemptions are to be made . . . to the Special Commissioners, and ‘at the head office for stamps and taxes in England.’ . . . How are the authorities at Somerset House to determine what constitutes a charitable purpose? . . . The question is determined for them by the law of the country in which they sit to exercise their jurisdiction.” On these broad grounds Lord Macnaghten came to the conclusion that the expression “trust for charitable purposes” in the Income Tax Acts was to be construed according to its “technical meaning according to English law”. Now, to revert to the example of a joint venture agreement: to whom is the contract speaking? The answer given by Lord Hoffmann is that a written contract is addressed to the parties: The Starsin [2003] UKHL 12; [2004] 1 A.C. 715 at [73]. This may be true of a contract between two individuals, for example, where a couple who are separating enter into an agreement for dividing up their jointly-owned property: compare Shaw v Hutton-Shaw [2006] EWCA Civ 1235; [2007] 1 F.L.R. 1839. Here “the parties” are also the individuals who will be interested in the meaning of the contract at any future date. But, in the case of a commercial (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

agreement between two companies, to say that the contract is addressed to “the parties” is true only in the same formal, and empty, sense in which it is said that the Income Tax Acts are addressed to all taxpayers. If, using Lord Macnaghten’s approach, one asks, “Who will actually have the job of ensuring that the contract is carried out, or deciding what to do if there is a breach?”, these may well be people who played no part in the making of the contract and who therefore have only a small part of the background knowledge possessed by those who did. It is to these people who come on to the scene at a later stage (not by reason of assignment, but through normal changes in the management of a company) and to their legal advisers that the contract is speaking. Those who negotiated and drafted the contract would have realised that people who would come on to the scene later on would have only part of the background knowledge that had been possessed by those involved in the preparation of the contract. It follows that, when interpreting a commercial contract, the admissible background should include only those matters likely to be readily available to the people who will have to interpret the contract in the future, including their legal advisers. This is merely to prefer a common sense to a literal reading of Lord Wilberforce’s famous statement in Reardon Smith Line Ltd v Yngvar Hansen-Tangen (The Diana Prosperity) [1976] 1 W.L.R. 989 at 997, that the court has to “place itself in thought in the same factual matrix as that in which the parties were”. If the background is restricted in the way suggested, what was said in the pre-contractual negotiations should normally be outside the limit. And the courts are beginning to realise that, without sensible restrictions, ICS 1 can increase the costs of litigation and allow scope for delaying tactics: ICI Chemicals & Polymers Ltd v TTE Training Ltd [2007] EWCA Civ 725 at [13]–[14]. The Chief Justice of New South Wales has recently given a noteworthy review of contemporary contractual in interpretation: “From Text to Context”, an address delivered in March 2007. The entire paper merits careful reading, but one of the most significant points is the mismatch between the simplicity of the five ICS principles and the complexity of modern commercial and financial transactions. There is, the Chief Justice says, “a basic defect in Lord Hoffmann’s five point schema. It is not a scheme that can be applied to a substantial range of commercial contractual relationships.” One further aspect of Wholesale Distributors v Gibbons calls for comment. At the end of his judgment, Thomas J. said that the defendant had advanced an interpretation at variance with its actual intention and was seeking to resile from the bargain. Thomas J. then used the word “fraudulent” to describe the case put forward by the defendant, and said that it was “clearly arguable” that the defendant had not acted in good faith. It is difficult to see how these remarks could have been justified. The defendant company’s directors would have been in breach of duty (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

had they caused the defendant to pay rent for the additional eight years if there was no legal liability. They were entitled to obtain a ruling from the court. The meaning of the clause was not obvious. In the Court of Appeal, the defendant’s interpretation had been accepted by one judge; in the Supreme Court, it was initially accepted by one judge, although he eventually changed his opinion. ALAN BERG.* EXCLUDING FIDUCIARY DUTIES:

THE

PROBLEM OF INVESTMENT BANKS

THE decision of the Federal Court of Australia in Australian Securities and Investments Commission v Citigroup Global Markets Australia Pty Ltd [2007] FCA 963 (ASIC v Citigroup) applies the well-understood doctrine that consenting commercial parties may readily exclude or limit in scope and intensity the operation of fiduciary duties (following Hospital Products Ltd v United States Surgical Corp (1984) 156 C.L.R. 41 at 96–116, per Mason J., and New Zealand Netherlands Society “Oranje” Incorporated v Kuys [1973] 1 W.L.R. 1126 at 1130–1132, per Lord Wilberforce). Jacobson J.’s judgment in ASIC v Citigroup raises new questions as to what counts as the minimum of consent necessary to exclude fiduciary obligations in the context of commercial and financial services, where clients may not fully appreciate the possibility that their advisers and financiers will take actions conflicting with client interests. Investment banks present this problem with special force, as the facts of the case reveal. Late in 2004 the Australian transport conglomerate, Toll Holdings Ltd (Toll) was planning a hostile takeover of Patrick Corporation Ltd (Patrick), another large national transport company. This required careful planning and secrecy in order to ensure that the Patrick share price did not rise in anticipation of a takeover; the lower the market price, the higher the premium that Toll would be able to offer Patrick shareholders when it was ready to launch its bid. Citigroup was initially asked by Toll to provide finance for the operation and to help Toll discreetly build up its holdings of Patrick shares. It was suggested to Toll by a Citigroup executive that the financing and trading services Citigroup were likely to be able to provide would increase in scale if the bank’s advisory team was more thoroughly engaged in planning the takeover; in other words, Citigroup was informally tying the full availability of its financial capital to the hiring out of its human capital. In June 2005 Toll agreed to retain Citigroup’s advisory services in addition to its financial and market services. Citigroup stipulated in its contract of engagement as * Tel Aviv. Admissibility; Contracts; Exclusion; New Zealand

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

adviser to Toll that it was to be retained solely “as an independent contractor, and not in any other capacity including as a fiduciary”. The contract further spelt out that Toll “should be aware that Citigroup . . . may be providing . . . financial or other services to other parties with conflicting interests”. Citigroup did not, however, give notice that aside from serving rival clients, it was trading in the securities market on its own account in order to build up its own capital base. The contract did promise due protection of confidential information. Citigroup’s “public side” proprietary trading department worked behind information barriers preventing the flow of confidential or price-sensitive information from the “private side” advisory departments of the bank. However, the existence and mode of operation of the proprietary trading was never explained to Toll. The Toll executives accepted Citigroup’s terms without reservations. In their witness statements the Toll officers revealed that they had not given much thought to the terms or likely effect of the retainer, but simply wished to secure Citigroup’s services as they prepared their assault on Patrick. As Citigroup advisers were helping Toll plan its takeover bid, a Citigroup transport analyst was asked whether a Toll bid for Patrick was likely. He refused to comment. This refusal was interpreted by share market participants as a robust indication that the analyst’s employer, Citigroup, was helping set up just such a takeover, and investors immediately moved to buy Patrick shares, pushing up the price and so reducing the possible share premium that Toll could offer when it launched its bid. The ensuing price shifts attracted the attention of the Australian Stock Exchange. Citigroup’s conflict of interest problem deepened when a trader on the public side of Citigroup noted the information inadvertently released by the Citigroup analyst, and also the ensuing cascade of bids for Patrick shares. Exercising his judgment that quick acquisition of Patrick shares could yield profits when the predicted takeover was announced, the Citigroup trader acquired over one million shares in Patrick worth some $6 million. The Patrick share price rose further as other investors observed the velocity of trading and began chasing the same stock. Executives on the private side of Citigroup became nervous that their public-side trader was joining and reinforcing the market move into Patrick shares, and feared that their own bank’s public trading activity could make the planned Toll takeover difficult or impossible by reducing the premium that could be offered to entrenched Patrick shareholders. After debating whether to offer full disclosure to Toll, the private-side executives of Citigroup decided instead to deal with the problem by crossing the information wall and quietly asking the public trader to desist from trading in Patrick. The trader was taken aside for a cigarette break and simply told without explanation: “Don’t buy any more.” His response was immediately to return to his desk and sell one-quarter of the million shares he had (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

accumulated. In the witness box the trader revealed that he thought his private-side colleagues were in effect warning him to divest his holding of Patrick shares as the price might fall. The trader’s sudden sell-off of the Citigroup holding in Patrick was taken by the market as yet a further sign that Citigroup was preparing a bid on Toll’s behalf, and the Patrick share price continued to rise. Toll adjusted its takeover strategy in response to market inflation of the Patrick share price by measuring its premium against a date preceding the wave of speculation. Toll’s bid eventually succeeded, costing $5.8 billion rather than the predicted $4 billion. In evidence the Toll officers conceded that they might have preferred that their advisers’ trading colleagues in Citigroup had not engaged in buying and selling shares in the company that Toll was preparing to bid for. Toll was aware that Citigroup had a proprietary trading operation, and had simply assumed that information walls would prevent use of confidential information in the course of that trading. They had not further considered that Citigroup public-side traders behind the information wall might engage in independent activity competing directly with Toll’s interests. An essential prelude to understanding the case is that it was not litigated by any party wronged or harmed by any purported breach by the defendant bank, but was brought as a regulatory action. The Australian Stock Exchange referred the case to the Australian Securities and Investments Commission (ASIC), the national regulator, who were already concerned that the investment banks’ wide extent of business would inevitably lead them into conflicts of interest and insider dealing harming confidence in the capital markets. ASIC argued that Citigroup had breached fiduciary obligations owed to its client, and that this raised concomitant breaches of statutory duties restraining conflict of interest, misleading and deceptive conduct, and insider dealing pursuant to s.912A(1)(aa) and ss.1042–1043 of the Australian Corporations Act 2001. ASIC’s key argument was that where a commercial party hires the advisory and financial services of a large investment bank, and so creates what would normally be a fiduciary relationship, the client cannot then give a valid consent to the exclusion of fiduciary duties from the contractual relationship unless that client is apprised of precisely which fiduciary protections are being surrendered and which types of conduct by the bank are thereby being licensed. According to ASIC, the incipient relationship between a client and an advising bank is a relationship of reliance and dependence with fiduciary qualities. This demands full disclosure by the bank of the types of conflicted conduct and profit-taking it is likely to pursue under the negotiated final relationship (for example, proprietary trading in competition with the client). Only then can the client meaningfully consent to such conduct. Analogy was drawn to a solicitor drafting a retainer for a client; fiduciary obligations inhere in the relationship of the negotiating parties independently of the retainer and therefore full (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

disclosure and informed consent is necessary in forming the agreement (following Mahoney J.A. in Law Society of New South Wales v Foreman (1994) 34 N.S.W.L.R. 408 at 435–436; Clark Boyce v Mouat [1994] 1 A.C. 428). Jacobson J. rejected ASIC’s key submission regarding the necessity of disclosure and informed consent; in a commercial relationship between financial adviser and client there is no proto-fiduciary obligation at the time of the execution of the contractual mandate, and fiduciary principles therefore did not inform the negotiation of the mandate. In the present case it was the mandate itself which constituted the relationship between Toll and Citigroup, and since this mandate expressly excluded fiduciary duty, a fiduciary relationship never came into existence. Informed consent or subjective awareness was not therefore a legal requirement; normal contractual notions of objective formation applied. Even if informed consent was required, then Toll’s knowledge of Citigroup’s structure as an investment bank could yield an implied informed consent as to the likelihood that the bank might end up trading in Patrick shares, in competition with the client’s interest (at [293]–[297]). The finding of informed consent was ultimately a question of fact (at [294], following Maguire v Makaronis (1997) 188 C.L.R. 449 at 466). ASIC had (somewhat surprisingly) conceded in the course of argument that the specific statutory duty to avoid conflicts of interest would only be engaged if a fiduciary relationship was present according to general law principles (at [26]). Having found that fiduciary obligations had been eliminated by the contract of engagement, Jacobson J. could therefore clear the bank of breach of the statutory conflict rules; and he likewise found that fiduciary duty had to be present in order to ground any finding of “misleading and deceptive conduct”. Finally he cleared the bank of breach of insider dealing rules through abuse of private information in its proprietary trading, finding that the bank’s information walls between its trading and advisory departments, whilst not perfect, had been adequate; and that the officers of the company had handled information flows within the firm and across the information wall in good faith, so as to protect their client’s interest. The proprietary trader could not himself be described as an “officer of the company” to whom knowledge possessed by the company on the private side could be ascribed, as the trader had too narrow a discretion and lacked a managerial role (at [479]–[501]). Jacobson J.’s decision ultimately turned on the distinction he accepted between per se fiduciaries, whose duties are presumed from the presence of a well-recognised type of relationship, and other fiduciaries whose duties arise from a position of protection and advancement of the other party, a position which gives foundation to a “fiduciary expectation” based on the facts of a specific relationship (at [271]–[275], following Finn, “The Fiduciary Principle” in Youdan (ed.), Equity, Fiduciaries and Trusts (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

(1989), at pp.46–47, and Hospital Products, per Gibbs C.J. at 68, Mason J. at 96). Jacobson J.’s judgment indicated that clients seeking financial services fall into the second class: fiduciary duties will tend to arise in this context only where there is evidence of reliance, confidence, dependence, and so on (Daly v The Sydney Stock Exchange Ltd (1986) 160 C.L.R. 371; cf. Hodgkinson v Simms [1994] 3 S.C.R. 377). If the client when forming a relationship with a financial adviser consents to exclusion of fiduciary duties from the outset, that is a weighty element in the matrix of fact tending to prevent any fiduciary obligation arising; and the consent in such a case can be effective without the client necessarily having to understand the specific types of profit-taking and conflicts of interest that may result from this exclusion. The inference of implied consent will be stronger in a case where a commercial practice involving conflict of interest is intrinsic to the structure of the market (Kelly v Cooper [1993] A.C. 205); no such inference applied to financial advisers since proprietary trading is not an essential or notorious part of their work (at [359]–[360]). However, the fact that large commercial clients who hire investment banks tend to be well-advised, sophisticated market players allows a more ready finding of an implied consent to re-definition of duties (at [296], citing Farah Constructions Pty Ltd v Say-Dee Pty Ltd [2007] HCA 22 at [107]). The immediate policy implication of Jacobson J.’s ruling is that commercial actors must look to their own resources in taking the decision to consent to a lifting of normal fiduciary duties; there is no proto-fiduciary duty requiring a more detailed, better-informed consent. To understand the full implications of this ruling, one must locate the reasons why in the contrasting category of per se fiduciary relationships a more specific consent must be given. To take two disparate examples: a medical patient in a fiduciary relationship with a doctor who relies upon the doctor’s judgment will not readily be found to have consented to the doctor engaging in self-interested conduct (Norberg v Wynrib [1992] 2 S.C.R. 226); cf Breen v Williams (1996) 186 C.L.R. 71; likewise a solicitor cannot win a client’s consent to act for a rival client where that double retainer would involve a real risk of conflict of interest, unless the solicitor explains precisely how the normal duties of disclosure and loyalty will be diminished in the context of their relationship (Beach Petroleum NL v Abbott Tout Russell Kennedy [1999] NSWCA 408; Hilton v Barker Booth & Eastwood [2005] 1 W.L.R. 567, noted (2006) 122 L.Q.R. 1). The explanation for requiring heightened consent is that in such per se relationships the fiduciary duties initially arise through a public policy of protection independent of the parties’ agreement. The professional calling of the doctor or the lawyer and their power over their clients automatically mandates fiduciary loyalty as soon as an undertaking of care is given. The policy question then arises as to why investment banks and other financial advisers who wield power over their clients and who are prone (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

to act in conflict with their clients’ interests are not also candidates for stronger public policy control through a stronger presumption of initial fiduciary status. Jacobson J. adverted to policy arguments along these lines, but ultimately he decided that it was the province of the legislature to require that advisers make fuller disclosures of their conflicting activities to potential clients being asked to waive fiduciary protections (at [602]). In today’s complex global markets, investment banks strive to offer “full spectrum” financial services to corporate and commercial clients, including analysis of asset prices, underwriting and issuing, securitisation, strategic advice including management of mergers, acquisitions and reorganisations, and direct financing. A modern investment bank will also maintain a stock of proprietary capital allowing it to fund client projects directly, and ultimately to cover the bank’s own liabilities. The multiple technical capacities of the large investment bank can be of great benefit to clients; the bank generates valuable information about business opportunities, for example through sophisticated asset-pricing and risk-engineering for a wide extent of businesses; and the bank then disseminates such information to relevant lenders and borrowers, thus facilitating economic activity. The bank’s own capital can also be crucial in some operations; for example in a large-scale takeover, the operation will more readily attract finance where lenders see a major bank taking the lead by injecting its own capital into the project (see further, Morrison and Wilhelm, Investment Banking: Institutions, Politics and Law (2007)). The benefits of all-embracing services in the one-stop investment bank may be considerable; but those benefits are shadowed by an ever-present danger of conflict of interest on the part of the bank, which can result in actions measurably harmful to clients’ interests. ASIC v Citigroup provides a catalogue of such dangers. The solution of erecting “information walls” (“Chinese walls” in popular but now-dating terminology) between the different parts of a bank is now mandated by legislation backed by regulatory enforcement; again, ASIC v Citigroup provides an object lesson in how such walls can be permeable. Jacobson J. emphasised that information barriers can limit but not eliminate conflicts of interest, and that such barriers are only legally valid if clients consent to deal with a fiduciary on that basis (at [308]–[321] and [441]–[456], following D & J Constructions Pty Ltd v Head (1987) 9 N.S.W.L.R. 118 at 123, per Bryson J.; Prince Jefri Bolkiah v KPMG [1999] 2 A.C. 222 at 237–238, per Lord Millett; Asia Pacific Telecommunications Ltd v Optus Networks Pty Ltd [2007] NSWSC 350 at [31]–[41], per Bergin J.). Commercial clients may count the advantages brought by working with a large and skilled investment bank to be so great that the risks of material conflict of interest are deemed to be acceptable. Clients will therefore have an incentive to accept without inquiry the bank’s stipulation that their relationship will not be governed by fiduciary law. The protections (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

offered by the bank’s duties of confidence and the maintenance of internal information walls are deemed to be sufficient. Where the risk of information leakage is thought to be too great, resort may be had to boutique firms offering specialist services such as mergers and acquisitions advice alone. The decision to employ the services of a large investment bank rather than the boutiques, in a case where the client has putative knowledge of how the large banks operate, can therefore be seen as an implied acceptance that damaging conflicts of interest may eventuate. The deeper question for courts, regulators and legislators is whether the risks of conflict and market abuse created by private actors such as investment banks within the financial system are too large to leave solutions to private ordering. The hands-off approach recommended by the decision in ASIC v Citigroup does not strengthen the arm of regulators in their work as gatekeepers protecting the integrity of the financial system, and opens the door to more heavy-handed legislative controls. Jacobson J.’s decision at first instance exonerating Citigroup will not be appealed by ASIC. The case therefore stands as an important precedent; and the attention to both evidence and doctrine in Jacobson J.’s judgment recommends the decision as a perceptive, if troubling, authority. It may be that Citigroup here achieved a mildly pyrrhic victory; and that ASIC can, after all, derive some satisfaction from the course of the litigation. Any well-advised client of an investment bank such as Citigroup will now be warned that whatever the type of legal protection on offer, there are serious risks involved in employing the services of full-spectrum investment banks in sensitive financial business. JOSHUA GETZLER.* UNANTICIPATED FIDUCIARY LIABILITY ONE of the great achievements of Paul Finn J.’s ground-breaking work, Fiduciary Obligations (1977), was a careful delineation of the different obligations owed by fiduciaries. On the 30th anniversary of the publication of that work, the decision of the Supreme Court of Canada in Strother v 3464920 Canada Inc [2007] SCC 24 is an unfortunate reminder of the dangers of a lack of precision in identification of fiduciary obligations and attribution of fiduciary liability. In taking a loose approach to the construction and attribution of the obligations of fiduciaries the majority judgment of the Supreme Court of Canada has dramatically expanded liability in circumstances which fiduciaries will be unlikely to anticipate. In the late 1990s, Monarch Entertainment Corp was involved in the production of American films in Canada. It was financed by a cunning * St Hugh’s College, Oxford. Australia; Conflict of interest; Fiduciary duty; Financial services; Investment banks; Share dealing offences; Takeovers

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

investment scheme which involved investment from Canadian taxpayers who obtained substantial tax benefits. The controlling mind behind the scheme was a rain-maker at the law firm Davis & Co called Robert Strother. Both Davis & Co and Strother were retained by a written agreement with Monarch which prohibited them from acting for any other clients in relation to tax schemes of this type. The retainer expired at the end of 1997 and was not renewed because everyone thought that amendments to Canadian taxation law had defeated the tax benefits of the scheme. Davis & Co continued to act for Monarch on general film production services transactions and unrelated general corporate work. In late 1997 a former employee of Monarch called Paul Darc approached Strother and suggested a revised tax scheme. Strother had discussed one related possibility with a lawyer at a different firm in November 1997 but nothing had come of it. Strother and Darc entered into an agreement in early 1998 to develop a revised tax scheme through a company called Sentinel, with Strother to receive slightly more than half of the profits from any success. Strother doubted that the scheme could work and did not mention his involvement with Sentinel to Davis & Co for six months. When Strother did mention his involvement to the managing partner of Davis & Co, he omitted to mention his financial interest and was forbidden from taking any financial interest. In October 1998 a favourable tax ruling was obtained by Sentinel. Strother resigned from Davis, effective at the end of March 1999, and went to work full time for Sentinel as a 50 per cent shareholder. Monarch only discovered the news of the tax ruling by word of mouth. It immediately severed all ties with Davis & Co and sued both Strother and Davis & Co. Monarch was unable to show that it had suffered any loss, because even if it had been advised in 1998 of the new scheme, it would not have re-entered the market for film production services. Therefore, instead of seeking compensation for loss, Monarch sought disgorgement of all the profits made by Strother and Davis & Co for breach of fiduciary duty and for breach of confidence. Strother had made profits of $32 million and Davis & Co had received legal fees from Sentinel and related enterprises of $9 million. In relation to the fiduciary duty claim, the central issue at trial concerned the scope of the agreement between Monarch and Davis & Co after its written retainer agreement ended in 1997. The trial judge found that the continuing retainer did not require Davis & Co to keep Monarch apprised of relevant developments. Strother was therefore free to be consulted by Darc in January 1998 and Davis & Co was free to act for Sentinel (at [24]). Since there was no undertaking to act for Monarch in these matters, there could not be a breach of loyalty by Strother or Davis & Co. The Court of Appeal for British Columbia allowed the appeal, holding that Strother was in breach of his fiduciary duty of loyalty by creating a conflict between his (2008) 124 L.Q.R., JANUARY ď&#x203A;&#x2122; SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

clients Monarch and Sentinel. They found that Strother also had a conflict between his duty to Monarch and his personal interest in maximising the profit of Sentinel and making sure that Monarch was kept in the dark about the new scheme. The Court of Appeal ordered that Strother disgorge all profits he had made. Davis & Co was held vicariously liable for those profits and directly liable for all the fees it received from Sentinel after January 1998. A majority of the Supreme Court of Canada (Binnie J.; Deschamps, Fish, Charron and Rothstein JJ. concurring) upheld the orders for disgorgement against Strother and Davis & Co, although excluding Davis & Co from liability for the legal fees. The starting point for both the majority and minority judgments in the Supreme Court of Canada was the observation that a lawyer’s fiduciary duty of loyalty is engrafted on to the contract of retainer (at [34], [133]). Strictly, a fiduciary’s duty of loyalty is engrafted on to any undertaking by a fiduciary, not merely a contractual one. The lawyer that acts pro bono, without consideration, has no less an obligation to take care in his work (Hedley Byrne & Co Ltd v Heller & Partners Ltd [1964] A.C. 265) and no less an obligation to do so with loyalty to the client. More importantly, however, although the duty of loyalty is engrafted on to the fiduciary’s undertaking, it must be understood as concerned only with the manner of performance of the fiduciary’s undertaking. The duty requires that the undertaking of the fiduciary must be performed whilst avoiding any interests or duties which conflict with those of the principal and avoiding any personal profit to the fiduciary without consent. Understanding the duty of loyalty as a manner of performance of an undertaking explains several things. First, it explains why the duty of loyalty does not survive the termination of the undertaking or retainer (Prince Jefri Bolkiah v KPMG (a firm) [1999] 2 A.C. 222 at 234–235). Once the retainer is gone there is no obligation upon which to engraft the duty of loyalty. If there is no obligation to perform, a fiduciary’s conduct, however objectionable, is not disloyal. Secondly, it explains why a fiduciary, such as a law firm, can act for clients with competing interests. As long as the undertakings to the clients do not involve competing outcomes, there can be no breach of any duty of loyalty in their performance. For example, a law firm could act for Manufacturing Co A in the purchase of land for a factory at the same time as it acts for a competitor, Manufacturing Co B, in matters relating to day to day business. To put the proposition positively, the performance of undertakings to two principals which involve competing outcomes is disloyal (Hilton v Barker Booth and Eastwood [2005] UKHL 8; [2005] 1 W.L.R. 567) unless the principal gives either express or implied consent (Kelly v Cooper [1993] A.C. 205). This simple point should have provided a simple answer to the issue before the Supreme Court. Before Strother or Davis & Co could be disloyal to Monarch, they needed to have undertaken to Monarch that (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

they would act for it in relation to the film production tax schemes in 1998. A minority of the court (McLachlin C.J.; Bastarache, LeBel and Abella JJ. concurring) held that Davis & Co plainly had not given any such undertaking to Monarch. The written retainer for such work with Monarch had expired in 1997. The continuing work which was being done for Monarch was general “clean-up” and “corporate” services. Strother was only to provide advice if Monarch specifically engaged him to consider it (at [123]). Hence, Davis & Co was free to accept Sentinel as a client and undertake to Sentinel to work on a new tax scheme for film production. In contrast, for the majority, Binnie J. held that the 1998 retainer by Monarch extended to considering and advising upon “tax-assisted business opportunities” (at [40]) so that Davis & Co and Strother were in breach of contract in failing to advise Monarch “that there may yet be life in a modified form of syndicating film production services for tax benefits” and failing to advise Monarch to consult another law firm (at [47]). Binnie J. relied upon oral conversations that Strother had with representatives from Monarch that “alternative” tax-assisted business opportunities might, in the future, be explored (at [40]). This broad approach was justified by reference to the nature of lawyers as “professional advisors, not used car salesmen or pawnbrokers” (at [42]). The implication was that professionals are required to perform obligations beyond those which they have strictly been asked to do. Yet, to have one rule of construction for contracts with professionals such as lawyers and another rule for contracts with other people cannot be justified. The process of construction is concerned with ascertaining objective meaning; considerations of public policy should not intrude. In any event, even if public policy were relevant to issues of construction, it could equally be argued that legal retainers should not be construed broadly because the fiduciary nature of lawyers means that every retainer has engrafted upon it additional, strict duties of loyalty. Alternatively, one could point to a public interest in the services of lawyers being available to the public generally. If retainers were construed in an overly broad manner this would crowd out the market for legal services, particularly in highly specialised areas such as tax scheme concerning production syndication for US films made in Canada. Having adopted this broad approach to construction of the undertaking, the majority should have reached the conclusion that Davis & Co had breached its fiduciary duty. Unlike the example above of Manufacturing Co A and Manufacturing Co B, the undertakings to Monarch and Sentinel involved competing outcomes. There was a limited market for investors for these tax schemes. An analogous example of breach of fiduciary duty recognised by Binnie J. is the case of a law firm which acts for two competitors for a single broadcast licence (at [55], see also Restatement (Third) of the Law Governing Lawyers, Vol.2 at para.121 (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

(2000)). Strangely, Binnie J. concluded that Davis & Co was free to accept Sentinel as a client because Sentinel created a “business opportunity which Monarch could have sought to exploit” (at [52]). But the outcomes of working for both companies were plainly not aligned. It was in the interest of Sentinel that Monarch discovered the new scheme as late as possible. It was in the interests of Monarch to discover it as early as possible. The bizarre consequence of the decision was therefore that although accepting that Strother breached his duty of loyalty, the majority held that Davis & Co was free to accept Strother and Sentinel as clients. Although Davis & Co was exonerated from breach of fiduciary duty, its victory was pyrrhic. It was held vicariously liable under the Partnership Act 1996 (R.S.B.C.) s.12 to disgorge the £32 million profit for which Strother was liable. The conclusion was based on a generous reading of the words of s.12 of the Partnership Act, namely that the other partners are liable for “loss, injury or penalty to the same extent as the partner [in breach]”. That British Columbian Act was itself based on s.10 of the UK Act of 1890 which had sought to codify the common law. For several reasons, it is surprising to think that the common law would ever have concluded that other partners were liable to account for unauthorised profits made by a partner in breach of fiduciary duty. First, when a director makes unauthorised profits in breach of fiduciary duty, the other directors are not liable to disgorge the breaching director’s profits (Regal (Hastings) Ltd v Gulliver [1967] 2 A.C. 134; Ultraframe (UK) Ltd v Fielding [2005] EWHC 1638 (Ch), noted D. Prentice and J. Payne (2006) 122 L.Q.R. 558; cf. CMS Dolphin Ltd v Simonet [2001] 2 B.C.L.C. 704). Secondly, as a matter of principle, it is counterintuitive to require fiduciaries like the partners of Davis & Co to disgorge a profit of which they were never aware and the making of which they had forbidden (at [17]). The asserted rationale for such liability was the purpose of teaching “faithless fiduciaries that conflicts do not pay” (at [77]) but the partners were neither “faithless” nor in receipt of any of the “payment” that they were required to disgorge. The overall effect of the Supreme Court’s decision is two significant expansions of fiduciary liability. First, a fiduciary lawyer’s undertakings will be construed very broadly, beyond the strict promise itself. The fiduciary will therefore be subject to strict duties of loyalty in the performance of undertakings, even though the breadth of those undertakings may not be clear. Secondly, even if a fiduciary performs its obligations loyally, if an unauthorised profit is made by its employee, partner or another for whom the fiduciary is vicariously responsible, the fiduciary will be liable to account for all the profit made by that other person. (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

Commercial fiduciaries have little cause to celebrate such expanded liability in circumstances that are difficult, or impossible, to anticipate and prevent. JAMES EDELMAN.* EQUITABLE THIRD PARTY LIABILITY THE unanimous judgment of Farah Constructions Pty Ltd v Say-Dee Pty Ltd [2007] HCA 22 (May 24, 2007) marks the first time in 30 years that the High Court of Australia has had the opportunity to consider and clarify the equitable liability of third parties to breach of trust or fiduciary duty. Hence, it was also the first opportunity for the High Court to consider the possibility of recognising a strict liability unjust enrichment claim in factual circumstances encompassed by equity’s knowing receipt rule. The first appellant, Farah Constructions Pty Ltd, formed a joint venture with the respondent, Say-Dee Pty Ltd, to re-develop a block of land in Sydney. The director of Farah, Mr Elias, oversaw an unsuccessful development application to the local council, in the course of which he learned that if adjacent properties were purchased, a revised application would most likely secure Council approval. Say-Dee alleged that Farah, through Elias, breached its fiduciary obligations by not imparting this information to Say-Dee and instead arranging for a related company controlled by Elias, Lesmint Pty Ltd, and Elias’ wife and two daughters, to purchase the relevant properties adjoining the joint venture’s land. It was alleged that in doing so Elias acted as his wife and children’s agent and, consequently, that his knowledge of the breach of duty should be imputed to them. The New South Wales Court of Appeal overturned Palmer J.’s decision at first instance that there had been no breach of fiduciary duty by Farah. It also held that Lesmint, Elias’ wife, and his children were liable under the first limb of Barnes v Addy (1874) L.R. 9 Ch. App. 244 (knowing receipt of trust property in breach of a trust or fiduciary duty); and, controversially, that the facts supported a strict liability claim in unjust enrichment. The High Court reinstated Palmer J.’s finding of no breach of fiduciary duty by Farah; but in lengthy obiter dicta went on to consider the Court of Appeal’s reasoning with respect to the recipient liability of the third parties and an alternative submission by Say-Dee that the third parties were knowing assistants to the breach of fiduciary duty by Farah, under the second limb of Barnes of Addy. Importantly, the court continued to frame the legal questions in terms of the two limbs of the so-called rule in Barnes v Addy. * Keble College, Oxford. Canada; Fiduciary duty; Retainers

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

Equitable recipient liability There are at least two ways to interpret the Court of Appeal’s finding that the third parties had knowingly received “trust” property according to the first limb of Barnes v Addy. The first possibility is that the information acquired by Elias was itself property belonging to the joint venture, which was then “received” by the third parties; that is, it was part of the joint venture’s “intellectual stock-in-trade” ([2005] NSWCA 309 at [173], per Tobias J.A.). Alternatively, the land acquired by the third parties as a result of the information was the relevant “trust” property; that is, the third parties acquired “the fruits of the valuable intelligence” obtained by Elias ([2005] NSWCA 309 at [175], per Tobias J.A.). The High Court, correctly in our view, rejected the Court of Appeal’s reasoning in both respects as fundamentally flawed. Recipient liability is the personal liability companion to a proprietary claim contingent upon the application of priority or tracing rules; that is, the beneficiary or principal must have an equitable proprietary interest which can be identified in property received by the third party. In this case, the High Court held, the information acquired by Elias was not property to which a trust could attach (contra Edelman (2006) 122 L.Q.R. 174 at 177). Even if the information were confidential, which it was not, this would not necessarily make it property for the purposes of recipient liability. Nor was the land acquired by the third parties “trust” property as it was never held by the joint venture, or by Elias, and nor was it possible to trace the information acquired by Elias into the properties. The High Court’s approach is consistent with recent English authorities: Satnam Investments Ltd v Dunlop Heywood & Co Ltd [1999] 3 All E.R. 652; Ultraframe (UK) Ltd v Fielding [2005] EWHC 1638; [2006] F.S.R. 17 at [1525], per Lewison J. Further grounds for rejecting recipient liability were that the evidence did not establish that Elias was acting as the agent of his wife and children and, unconvincingly, that his actual knowledge could not be imputed to them. In a surprising aside, the High Court also expressed doubt as to whether recipient liability should apply at all in relation to property which is not the subject of an express trust: this had been “assumed, but rarely if at all decided” (at [113]). Previously, the requirement of “trust” property in the first limb was assumed to encompass all situations where a fiduciary has property of the principal under his or her control (see, e.g. Belmont Finance Corp v Williams Furniture Ltd (No.2) [1980] 1 All E.R. 393 at 405, per Buckley L.J. and Foley v Hill (1848) 2 H.L. Cas. 28 at 35–36, per Lord Cottenham L.C.). (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

Strict liability The New South Wales Court of Appeal held that a restitutionary, unjustenrichment based claim was available to Say-Dee. Such liability is strict, subject to defences, and thus no knowledge needs to be proved where “recipients” have been unjustly enriched at the claimant’s expense. The High Court categorically rejected the use of unjust enrichment to impose strict liability in this context and was highly critical of the lack of analysis by the Court of Appeal. First, the High Court emphasised that long-established authority, including “seriously considered” (at [134]) obiter dicta of the High Court in Consul Development Pty Ltd v DPC Estates Pty Ltd (1975) 132 C.L.R. 373, precluded an intermediate court from taking such a radical step. Thus, the requirement of notice or knowledge as a foundation of recipient liability in equity is too firmly entrenched to be abandoned. Say-Dee had sought to explain the unjust enrichment claim as arising alongside the equitable claim, rather than as a recasting of it (consistently with Birks’s later writing on the subject, e.g. “Receipt” in Birks and Pretto (eds), Breach of Trust (2002), p.213 at p.223). The High Court, however, correctly recognised that this would be to render the first limb of Barnes v Addy, and its notice requirement, otiose (at [134]). Hence, recipient liability should not be outflanked by the recognition of a strict liability unjust enrichment claim. Apart from precedent, a number of conceptual arguments can be advanced in support of this view and the “want of intellectual merit” (at [132]) of a strict liability unjust enrichment claim in this context. Specifically, the High Court rejected the view that there had been any receipt of trust property by the third parties, for the purposes of the knowing receipt rule. Although the High Court did not separately consider the point, it is equally problematic, and for similar reasons, to establish the receipt of any “enrichment’ at the claimant’s expense. What was the enrichment: the information itself? Or the property purchased “using” that information? Clearly, regardless of whether the courts are to impose recipient liability or strict liability under unjust enrichment, on the receipt of property or enrichment, respectively, it is imperative to avoid loose and overly broad meanings of those concepts. The High Court also noted that there was no relevant “unjust factor” justifying restitution. “Breach of fiduciary duty” could not be such a factor, as the third parties were not fiduciaries, and did not have knowledge of any such breach. The only possible candidate was “ignorance”, but the High Court stated that even in England, no case has treated ignorance as a basis for restitution (at [156]). At a normative level, the High Court emphasised the injustice of recognising a strict liability unjust enrichment claim on these facts. The High Court stressed that the unjust enrichment claim should not be used (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

to “cut down traditional equitable protection” (at [153]). The onus is on those advocating the recognition of the strict liability unjust enrichment claim to justify the normative merits of such change to the law (at [148]). It was not clear “how there was any justice in permitting restitution against a defendant who received trust property without notice of that fact” (at [155]). Hence, the assumption by many that a strict liability unjust enrichment claim is both conceptually and normatively preferable to knowledgebased liability in such cases, such as to be almost “foreordained and . . . inevitably correct” (Farah, at [133]), is finally being challenged at the highest judicial levels. This is something to be welcomed; we have extensively expressed our own views on the matter recently, in particular rejecting the burdensome onus that strict liability under unjust enrichment places on innocent recipients to satisfy defences (Dietrich and Ridge, (2007) 31 M.U.L.R. 47). If all this suggests that the divide between the English and Australian law of restitution is ever widening (on which, see Hedley (2004) 28 M.U.L.R. 759), of even more interest is the approach of the High Court to the “unjust enrichment” principle more generally. This is not a concept to be given a free or wide-ranging operation. Unjust enrichment is not to be used as a mechanism to recast the law; at best, it operates in narrow terms, requiring the existence of specific qualifying or vitiating factors (at [150]). Perhaps most importantly, in a probable reference to the Birksian school of unjust enrichment theory, the High Court rejected “a mentality in which considerations of ideal taxonomy prevail over a pragmatic approach to legal development” (at [154]). The court adopted views of Gummow J. in Roxborough v Rothmans of Pall Mall Australia Ltd (2001) 208 C.L.R. 516 at 544, stressing that theory in the common law derives from judicial decisions rather than the other way around (at [154]). It also endorsed Gummow J.’s view that much modern unjust enrichment theory represents a dogmatism that “will tend to generate new fictions in order to retain support for its thesis. It also may distort well settled principles in other fields . . . so that they will answer the newly mandated order of things” (at [151]). For those who may have thought that these views of Gummow J. are essentially his own, Farah puts paid to such arguments. Knowing assistance liability In the Privy Council case of Royal Brunei Airlines Sdn Bhd v Tan [1995] 2 A.C. 378 Lord Nicholls of Birkenhead drew on nineteenth-century case law concerning the liability of third parties who participated in breaches of trust to formulate a general principle of “accessory liability” for breaches (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

of trust and fiduciary duty for which dishonesty on the part of the third party is the requisite element for liability (Fyler v Fyler (1841) 3 Beav. 550; Attorney General v Corp of Leicester (1844) 7 Beav. 176; Eaves v Hickson (1861) 30 Beav. 136). Lord Nicholls used these cases to support his reasoning that a dishonest and fraudulent breach of duty by the trustee or fiduciary is not required and consequently he rejected Lord Selborne’s formulation of the second limb of Barnes v Addy. In contrast, the High Court in Farah, whilst relying on almost the same nineteenth-century case law, has taken the opposite approach to that of Lord Nicholls by expressly differentiating the ways in which third-party participatory liability for breach of trust or breach of fiduciary duty may arise. Thus, whilst the Privy Council’s formulation of accessory liability incorporates various forms of third party involvement in breach of trust or fiduciary duty, namely, inducement, procurement, and assistance, into one general principle with one criterion for liability (dishonesty by the third party), Australian courts are to maintain the distinction between liability under the second limb of Barnes v Addy and liability by way of procuring or inducing a breach of trust. Unfortunately, there is little in the way of principled explanation for why the distinction between assistance by a third party in a breach of trust or fiduciary duty on the one hand, and procurement or inducement by a third party of a breach of trust on the other, should be maintained. Nor is there any explanation of whether thirdparty liability for procurement and inducement applies only in relation to breaches of trust. The immediate consequence of the High Court’s approach is that Lord Selborne’s statement of principle in Barnes v Addy is given far greater weight. Thus, the High Court affirmed that for knowing assistance liability the breach of trust or fiduciary duty by the fiduciary must itself be dishonest and fraudulent before the third party is liable; whereas, this is not necessary in order for a third party who procured or induced a breach of trust to be liable. The extent to which the knowledge criterion for knowing assistance liability in Australia differs in substance from the dishonesty test for English accessory liability continues to be an open question. Further reinforcing its divergence from the English position, the High Court expressly adopted the Baden scale of knowledge (Baden v Société Générale, etc. [1992] 4 All E.R. 161 at 235, per Peter Gibson J.) as being of assistance in interpreting the judgments in Consul as to the third party’s requisite state of knowledge (contra Royal Brunei at 392, where Lord Nicholls suggested that the Baden scale of knowledge is “best forgotten”). This is rather surprising given that one member of the court has referred, extra-judicially, to the scale as “the zenith of complexity”: Heydon and Leeming, Jacob’s Law of Trusts in Australia, 7th edn (2006), para.[1335]. The court used the scale to clarify some uncertainties arising (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

from Consul, and concluded that level (iv) on that scale, “knowledge of circumstances which would indicate the facts to an honest and reasonable man” is sufficient for liability. It is designed to capture the “morally obtuse” defendant who does not recognise an impropriety that would have been recognised by an “ordinary person” (at [177]). ‘‘Knowing assistance” in Australia now has a much narrower scope than English accessory liability. The motivation for this appears to be a concern with the potential for injustice when third-party liability rules developed in relation to traditional express trusts are applied to breach of fiduciary duty in general. Particular reference is made to the situation of “directors, advisers and bankers of [an] insolvent company” (at [179]). A pressing question after Farah, then, concerns the extent to which breach of equitable duties by fiduciaries and trustees should be treated as interchangeable for the purposes of third-party liability. Nor, we would suggest, is this question solely an Australian concern. Unfortunately, whilst the High Court made some attempt to address the conceptual and normative merits, or lack thereof, of an unjust enrichment claim, there is little substantive discussion of the normative questions raised by the court concerning the proper scope of equitable wrong-based liability for third parties to breach of trust or fiduciary duty. For the moment, arguments based upon precedent have won the day; in the longer term, however, a principled and normative analysis, which builds upon precedent in an acceptable way, is required. It is to be hoped that it will not be another 30 years before the High Court has that opportunity. PAULINE RIDGE.* JOACHIM DIETRICH.* FINAL GUIDELINES ON COMPENSATION OF COMMERCIAL AGENTS THE Commercial Agency Regulations, which regulate the relationship that commercial agents have with their principals, implement an EC Directive which aims to reinforce the legal protection of commercial agents. One of the most important protective features of the Directive is to give commercial agents, on termination of the agency relationship, the right to claim a lump sum payment. The Directive allowed Member States to choose between “indemnity” or “compensation for loss” (Art.17(1)), the two concepts drawing on German and French law respectively, but offered limited guidance on the meaning of such terms or how they should be calculated. The Government did not select one or the other and instead left the choice between the two options to the parties (reg.17(1)). In * Australian National University. Australia; Breach of fiduciary duty; Breach of trust; Strict liability; Third parties; Unjust enrichment

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

the absence of choice, compensation applies (reg.17(2)). Although the Directive and the Commercial Agency Regulations offer some guidance as to the application of such concepts, what is actually meant, especially as to compensation, cannot be said to be clear, and merely copying the Directive out does not make things easier. Much depends on the courts clarifying the scope of application of this part of the implementing text by explaining what “damage”, i.e. loss, agents suffer on termination (reg.17(6)) and how to assess the compensation (reg.17(7)). Most cases to date have involved compensation rather than indemnity; it is accepted that the loss for which the agent can claim compensation pursuant to reg.17(6) is specific to the agency relationship and as such different from any existing common law principles. In King v Tunnock Ltd 2000 SC 424 Lord Caplan stated that the agent’s loss is that caused by “the termination of [the agent’s] relations with his principal” and an agency “has commercial value” (at [38]). In Lonsdale v Howard & Hallam Ltd [2006] EWCA Civ 63; [2006] 1 W.L.R 1281, Moore-Bick L.J. stated that the loss suffered was “the loss of the goodwill attaching to the agency business” (at [28]) for which a claim would not otherwise arise. In spite of this, the Scottish and English courts have disagreed on the proper assessment of the compensation. In King v Tunnock Lord Caplan said that reg.17(7) did not “set out the exclusive circumstances giving rise to compensation” (at [42]). Thus compensation was due even “where the principal shuts down the relevant part of his business, or say ceases to trade because the company goes into receivership” and “so far as entitlement to compensation is concerned, the Directive is not troubled with what happens after the date of termination” (at [43]). After stating that the pursuer’s business enjoyed considerable goodwill, he calculated the compensation without taking account of the fact that the principal’s business was reduced following the closure of the part of the business of which the agent was in charge. Although Lord Caplan considered the approach of the French courts and awarded two years worth of commission, he nevertheless insisted that he was applying Scots law. In the Court of Appeal for England and Wales, however, in Lonsdale v Howard & Hallam Ltd [2006] EWCA Civ 63; [2006] 1 W.L.R 1281, Moore-Bick L.J. said compensation should be assessed as if one were valuing a business. Although the value of that business was to be considered at the date of termination, not all common law principles would be disregarded, and therefore events following termination that were likely to affect its value were also relevant. One such event was the fact that the principal’s business was in decline, as it reduced the value of the agency and therefore the value of the compensation. In his analysis, Moore-Bick L.J. refused to consider the French approach even as a broad guideline. Following such uncertainty, leave to appeal to the House of Lords was granted. The House ([2007] UKHL 32; [2007] 1 W.L.R. 2055) upheld (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

the Court of Appeal position and rejected the Scottish approach, and gave guidance on the assessment of compensation. Lord Hoffmann delivered the leading speech (with which Lord Bingham of Cornhill, Lord Rodger of Earlsferry, Lord Carswell and Lord Neuberger agreed). Before deciding how compensation should be determined, he stated that he first had “to decide exactly what the agent should be compensated for” (at [7]). Commenting on the French origin of the option, he said that one could therefore “look at French law for guidance or confirmation as to what [the loss] means” (at [9]). He concluded that French law places a value on the right to be a commercial agent and that the loss the Directive requires commercial agents to be compensated for is the loss they suffer by being deprived of the benefit of the agency relationship (at [10], [11]). This part of the decision is unproblematic since it confirms that the loss for which agents are compensated is a special one. Lord Hoffmann then turned his attention to the issue at the core of the decision, the determination of the compensation, which, in turn, raised three separate questions; (i) whether their Lordships should follow the French method of calculation and award two years’ worth of commission; (ii) whether the question of calculation should be referred to the European Court of Justice (“ECJ”); and if not (iii) what precise method of calculation should be adopted. To the question whether the French system should be followed, he answered, unsurprisingly, with a resounding “no” for three reasons. First, although Art.17(2) of the Directive is based on French law, the EU Commission did not endorse it as a true reflection of community law (at [16]). Secondly, the French and English implementing texts both contain provisions requiring commercial agents to be compensated for the loss suffered on termination: what differs is simply the method by which the loss is calculated. Member States are allowed some discretion over the method of calculation of termination payments (Honyvem Informazioni Commerciali Srl v Mariella De Zotti (Case C–456/04, [2006] E.C.R. I–02879). Thus, the difference in the French and English methods to calculate the loss is not a problem. Finally, Lord Hoffmann noted that whilst in France agencies often changed hands, this did not seem to be the case in the United Kingdom. Consequently, “the difference between French and English practices exist . . . because they are operating in different market practices” (at [18]). To state that English commercial agents operate in different market conditions than their French counterparts without empirical evidence is somewhat controversial. Interestingly however, his Lordship remarked upon the (little) advertised French practice of “droit d’entr´ee” whereby an incoming agent would pay an entrance fee of, usually, two years of commission (at [18]). Judging from the mention of such a practice in a comparative academic study (Sellier, Principles of European Law, Commercial Agents, (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

Franchise and Distribution Contracts (2006), p.103), it seems to exist beyond France. Yet, how widespread it is is unclear. On the question whether it was necessary to refer the matter to the ECJ, Lord Hoffmann also answered “no”. After agreeing that the lack of consensus between the Scottish and the English courts created uncertainty, he added that it was not the meaning of the Directive which was uncertain, but the method by which the compensation is calculated. Since Member States have discretion in this matter, what was uncertain was the way the discretion had been implemented domestically, the task of resolving this uncertainty therefore fell on their Lordships and not on the ECJ. Turning to the most difficult question, that of a precise method for calculating the compensation, Lord Hoffmann held that the level of compensation was to be assessed by reference to the open market sale value of the agency, i.e. by reference to what a “hypothetical purchaser” “would have been willing to pay for similar businesses at the time” (at [12], [28]). Since what had to be valued was “the agent’s expectation that “proper performance of the contract” will provide him with a future income stream” (at [11]), the value of the compensation is based on the value of the agency “on the assumption that it continued”, which represents “the amount which the agent could reasonably expect to receive for the right . . . to continue to perform the duties of the agency and receive the commission which he would have received” (at [21]). By doing so, the court takes account of “what the parties were likely to have expected to happen” after termination (at [39]), which is clearly different from the French approach. To the claim by counsel for the appellant that anything less favourable to commercial agents than the French method of calculation would not give effect to the protective stance of the Directive, Lord Hoffmann considered what an agent would have obtained under the indemnity option since “there is no doubt that this would satisfy the policy of the Directive” (at [19]). Remarking that no indemnity would have been payable “in a case such as the present in which the principal went out of business and therefore derived no benefit from the customers”, he stated that it was therefore “impossible to argue that [compensation] requires a payment of twice gross commission whether the principal has derived any benefit or not” (at [20]). His Lordship then concluded that to value the agency on the assumption that it continued was therefore not inconsistent with the protective stance of the Directive. To use the indemnity option as a rationalisation that the method of calculating the compensation due is consistent with the protective stance of the Directive is, with respect, doubtful, given the different approaches of the two termination payments. Compensation considers whether the agent suffered a loss following the termination of the relationship; in contrast, the indemnity focuses on whether the principal benefited from the agent’s efforts (reg.17(3)(a)). The distinction is crucial; whilst the agent can claim compensation for (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

the loss suffered even if the principal does not acquire new customers, no indemnity will be due if the principal has not gained financially from the agent’s efforts. Furthermore, the indemnity is capped at one year’s remuneration (reg.17(4)), compensation is not. An interesting argument in his Lordship’s reasoning can be seen in his comment that although the EU Commission Report stated that the UK and the French approaches as regards to compensation were different, there was no suggestion that either approach would fail to implement the Directive. The difference in question relates to the fact that given the unfamiliar nature of the concept of compensation to a common lawyer, the UK courts would be likely to have regard to existing common law principles when assessing it, a difference that the EU Commission merely referred to as a “difficulty”. By calculating the compensation due by taking account of what the parties were likely to have expected to happen after termination, the House appears to be having regard to common law principles, which, in view of the report, is apparently not a problem, even if different from the French approach. Continuing on the method of assessment, Lord Hoffmann insisted on the need for the valuation to be done by reference to what is happening in the “real world” (at [13]), He cast a critical eye on the case law to date, rejecting King v Tunnock and stating that the sheriff’s analysis that the goodwill disappeared when the business closed was the correct one. After praising the judge’s approach in Tigana Ltd v Decoro [2003] Eu. L.R. 189 for calculating compensation on net earnings, since they are what matters to the hypothetical purchaser, he nevertheless criticised the final assessment for lack of evidence that anyone would have paid this figure for a comparable business (at [30]). H.H. Judge Overend’s assessment in Smith, Bailey Palmer v Howard and Hallam Ltd [2006] Eu. L.R. 578 was similarly rejected for being based on cost rather than what one would have paid for the agency (at [31]). Emphasising the importance of a clear methodology for the valuation, his Lordship’s final task was to define the necessary factors for assessing the quantum of the compensation. The first is the level of net commissions received by the agent; and when the agent has several agencies, the costs must be fairly attributed to each (at [29]). The fact that the market, and therefore the principal’s business, is in decline is important too, since a buyer would pay less for the agency (at [36]) and there must be a deduction for accelerated receipt (at [13]). Such factors are closer to indemnity than compensation. The last was considered in relation to the intervention of the Australian Winemakers’ Federation whose concern was the event where, following termination, the agent can transfer the goodwill created to another principal. Lord Hoffmann considered that to pay compensation in such circumstances would be unfair since the former principal does not retain the goodwill (at [37]). This argument seems (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

to consider the principal’s gain rather than the commercial agent’s loss, which, as previously mentioned, is closer to indemnity than compensation. However, Lord Hoffmann also added that, if, in such a situation, the “hypothetical purchaser” could not be sure that the customers could continue to trade with him, “he would be unlikely to be prepared to pay much for the agency” (at [38]). Since the agent is the one claiming compensation, the burden of proof of the value of his agency falls on him (at [35]). Expert evidence will therefore be needed, which will be costly. Aware of this problem, his Lordship reiterated the need for a clear methodology for the valuation, arguing that “once it is firmly understood that the compensation is for the loss of the value of the agency, relatively few cases will go to court” (at [35]). In fact, “after a period of experience in such valuations” the courts will be able to apply a “going rate” for what Lord Hoffmann refers to as a “standard case”, hence reducing the need to repeat boilerplate evidence in every case (at [36]). Lord Hoffmann nevertheless emphasised the importance for the judge to be “fairly confident that he is dealing with the standard case” (at [36]). As the first ruling of the House of Lords on the method of calculation of the compensation due, this unanimous decision is of great importance. The recognition of the specificity of the loss the agent suffers on termination and the need for a clear methodology for its assessment is to be welcome. By stating that the loss suffered is that of the agency as an asset to the agent, the court seems to accept that the basis of the payment is the expropriation of the agent’s quasi-proprietary interest and not the principal’s gain. This analysis, in line with French law, also recognises the different emphasis between compensation and indemnity. However, to take account of the fact that the principal’s business is failing as reducing the value of the agency and therefore the value of the compensation blurs the distinction between compensation and indemnity and appears to calculate compensation by reference to existing common law principles. In fact, looking at the state of the principal’s business means that in certain cases, in spite of losing the agency, the agent will be left with no compensation. Member States have some discretion in the matter, yet, given the previously mentioned different emphasis between the two options, it is not clear whether this is correct. Although the EU Commission did not consider that calculating compensation by reference to common law principles was a failure to implement the Directive, doing so may nevertheless undermine the specificity of the loss agents suffer on termination, which may be against the protective stance of the implementing text. By emphasising the domestic discretion in the calculation of compensation, their Lordships accentuate the fact that the Directive stipulated a remedy for a loss but left it to Member States to decide how that remedy is to be calculated. If the effect of (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

such a discretion, is that, on similar facts, domestic courts interpret the Directive differently, this raises wider questions as to the effectiveness of this Directive as a tool for harmonisation of European private law, a problem recognised by the European Commission, who, in its Action Plan on a More Coherent European Contract Law (Com (2003) 68 final) criticised the Directive for creating uncertainty (Com (2003) 68 final at para.18). SEVERINE SAINTIER.* WHEN IS A RISK OF INJURY FORESEEABLE? IN 1980 the High Court of Australia pronounced judgment in a case that has a strong claim to being the most significant authority concerning the law of torts in Australia. That case was Wyong Shire Council v Shirt (1980) 146 C.L.R. 40. The plaintiff in those proceedings suffered quadriplegia in a water-skiing accident. That injury was sustained because the plaintiff interpreted a sign that read “Deep Water” as indicating that the water beyond the sign was deep. In fact, the sign, which had been placed by the defendant Council, marked a channel that had been dredged for the use of water-skiers. While skiing in what he wrongly believed to be deep water, the plaintiff fell and struck his head on the bed of the lake. The issue was whether the risk of injury generated by the ambiguity in the warning sign was foreseeable. Mason J., in a classic passage, declared that a risk is foreseeable unless it is “far-fetched or fanciful” (at 47). This test, derived from the Judicial Committee’s advice in The Wagon Mound (No.2) [1967] 1 A.C. 617, was adopted over a more demanding formula suggested by Barwick C.J. in Caterson v Commissioners for Railways (1973) 128 C.L.R. 99 at 101–102 of “not unlikely to occur”. Unsurprisingly, the High Court proceeded to hold that, from the perspective of the reasonable Council, the risk of injury was foreseeable. The undemanding test selected in Shirt all but removed the requirement of foreseeability as a precondition to liability in negligence. This not only increased the circumstances in which negligently inflicted injuries were compensable but it also signalled a sympathetic attitude towards plaintiffs in the High Court. This stimulated further growth of the tort of negligence. Consequently, it is unsurprising that when the pendulum began to swing back in favour of defendants at around the start of the 21st century (see Luntz, “Torts Turnaround Downunder” (2001) 1 O.U.C.L.J. 95), the Shirt test came in for severe criticism. For instance, in Swain v Waverly Municipal Council (2005) 220 C.L.R. 517 at 549, McHugh J. asserted that * University of Sheffield. Commercial agents; Compensation; Termination

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

the Shirt test was a “pernicious principle [that] has done much damage to the utility of the common law doctrine of negligence” (see also Tame v New South Wales (2002) 211 C.L.R. 317 at 351–357). Similarly, in Koehler v Cerebos (Australia) Ltd (2005) 222 C.L.R. 44 at 64 Callinan J. remarked that “with enough imagination and pessimism it is possible to foresee that practically any misadventure, from mishap to catastrophe is just around the corner” and lamented that “the rule . . . in [Shirt] requires everyone to be a Jeremiah.” Once the retreat of negligence was underway, it did not take long for the High Court to give permission for the correctness of the test for foreseeability as enunciated in Shirt to be contested. The first opportunity came in Paua Nominees Pty Ltd v Miller [2005] H.C.A.Trans. 774. However, during the hearing of that case, the court came to the view that it was an inappropriate vehicle for reconsidering the Shirt test and rescinded special leave to appeal. The next opportunity arose in New South Wales v Fahy (2007) 81 A.L.J.R. 1021. In this matter the court gave full consideration to whether the Shirt test should be retained. In the end, that test escaped intact. However, the decision in Fahy reveals a marked difference of opinion in the High Court as to the appropriateness of that test. It also provides fresh impetus to re-examine the reason why a risk of injury must be foreseeable before the defendant is required to take reasonable care to avoid its materialisation. The facts in Fahy were straightforward. The plaintiff police officer and another officer with whom she had been assigned to work, Senior Constable Evans, attended a medical centre where an assault victim was being treated. The plaintiff entered the room where the victim was being seen by a doctor. Mr Evans waited outside. While in the treatment room the plaintiff tried, among other things, to reduce the bleeding from a very serious wound. The stress to which the plaintiff was subjected in this situation resulted in her suffering a psychiatric injury. The plaintiff contended that this injury would not have been sustained had the New South Wales Police Service taken reasonable care to ensure that the “partner system”, whereby paired officers would work together and support each other, was carried into effect on this occasion. The gravamen of her complaint was that Mr Evans had, without good reason, abandoned her. The plaintiff succeeded at first instance in the District Court of New South Wales. Although her damages were reduced by the Court of Appeal on account of her failure to mitigate her loss, the Court of Appeal unanimously affirmed the correctness of the District Court’s ruling on liability ((2006) 155 I.R. 54). However, a majority of the High Court overruled the decisions below and held that the defendant was not liable. The difference of opinion in the High Court did not concern whether the risk of psychiatric injury was foreseeable. It was accepted that it (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

was, regardless of the applicable test. Rather, the High Court divided over whether there was a breach of duty. As this point is less interesting than the issue of foreseeability, only a few words will be said about it. In brief, Gummow and Hayne JJ. thought that the reasonable employer would not have instructed police officers who were assigned to work in pairs to stay together unless operational circumstances required otherwise because such an instruction would be so lacking in precision that it could not be effectively implemented. Their Honours also considered that the proposed instruction would impermissibly qualify the statutory obligations of police officers. Callinan and Heydon JJ. concluded that the postulated instruction was impractical as the situations in which police officers would become separated in the course of their duties were very numerous. Furthermore, their Honours thought that the Police Service, in providing training to police officers and taking other measures to combat stress, had taken reasonable care for the plaintiff in the circumstances. The minority, constituted by Gleeson C.J., Kirby J. and Crennan J. disagreed, principally on the basis that the trial judge’s conclusion was open to him on the evidence and was therefore not susceptible to appellate disturbance. We can now turn our attention to the issue of foreseeability. Gleeson C.J., Gummow and Hayne JJ. and Kirby J. held that the Shirt formula should be retained. Gleeson C.J. and Gummow and Hayne JJ. remarked that they had not heard any convincing reason for doubting the correctness of that formula. In contrast, Kirby J. went to considerable lengths to commend the Shirt test. His Honour argued that the entrenched position of that test in negligence law coupled with considerations of accident prevention were reasons for affirming it. While the foregoing Justices expressed concern at the fact that the courts had sometimes jumped from a finding that a risk was foreseeable to the conclusion that the defendant was negligent without pausing to ask whether the defendant acted reasonably in the circumstances, they considered that this was a flaw in the application of the Shirt test rather than a problem with that test itself. Crennan J. did not feel that it was necessary to express any view on the correctness of the Shirt test since, no matter how the risk of psychiatric injury was analysed, it was incontestably foreseeable. However, Callinan and Heydon JJ. took aim at the test. Although their Honours considered that Fahy was an unsuitable case for replacing that test, presumably because the risk in question was so obvious, they said that its replacement was desirable as it is so easily satisfied that it seriously detracts from the traditional learning, memorably expressed by Lord Atkin in Donoghue v Stevenson [1932] A.C. 562 at 580, that “. . . liability for negligence . . . is no doubt based on a general public sentiment of moral wrongdoing for which the offender must pay”. Callinan and Heydon JJ. did not reach (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

a firm conclusion on what should replace the Shirt formula. However, they expressed a preference for a test of whether the risk in question was “significant enough in a practical sense” (at [226]). This test had been advanced by Walsh J. sitting at first instance in The Wagon Mound (No.2) [1963] S.R. (N.S.W.) 948 at 957. It is somewhat surprising, considering the present hostile attitude of the High Court towards the tort of negligence, that the Shirt test survived the challenge to it. As noted above, in recent years, the High Court has been busy restricting the circumstances in which the tort of negligence affords a remedy. This is particularly obvious in relation to so called “diving cases”. The high-water mark in such actions is the decision in Nagle v Rottnest Island Authority (1993) 177 C.L.R. 423. In that case, the High Court held that a man who suffered quadriplegia when he dived off a rock platform into water without checking its depth could recover damages from the authority responsible for managing the area on the grounds that it was negligent in failing to warn that the water was shallow. While the High Court has not expressly declared that Nagle was wrongly decided, it has gone to considerable lengths to confine it and to emphasise factors militating against liability (see, e.g. Vairy v Wyong Shire Council (2005) 223 C.L.R. 422). The High Court has also lately rejected attempts to extend the reach of the tort of negligence. For instance, the court has refused to recognise actions for “wrongful life” (Harriton v Stephens (2006) 226 C.L.R. 52), declined to find an exception to the rule that principals are not liable for the negligence of independent contractors where the contractor acts as the “representative” of the principal (Sweeney v Boylan Nominees Pty Ltd (2006) 226 C.L.R. 161) and rebuffed efforts to impose a non-delegable duty of care on highway authorities with respect to road users (Leichhardt Municipal Council v Montgomery (2007) 81 A.L.J.R. 686; (2007) 233 A.L.R. 200). In view of this restrictive climate and the trenchant attacks made on the Shirt test, the retention of that test is rather unexpected. This may signal that the High Court believes that the common law of negligence now strikes an appropriate balance between the interests in bodily security and freedom of action. However, one cannot help but wonder whether the test would have been retained had the risk in Fahy been less obvious. On a different note, it is disappointing that the High Court did not, on the whole, grapple with the issue of why a risk of injury must be foreseeable before a defendant is required to take reasonable precautions against its materialisation. Although one would expect that the reason for this requirement would cast light on when a risk ought to be regarded as foreseeable, only Callinan and Heydon JJ. addressed (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

this matter. As mentioned above, their Honours embraced the conventional judicial understanding that the requirement of foreseeability, like the other elements of the tort of negligence, is intended to bring about a degree of correspondence between liability and moral blameworthiness. They thought that the Shirt test was so expansive that it failed to advance this aim and that a more rigorous test was needed. There is no doubt that there is a close relationship between foreseeability and blameworthiness. According to a commonly held view, for a person to be to blame for bringing about an undesirable outcome, it is necessary (but not sufficient) that he had an opportunity to avoid that outcome (see Perry, “Risk, Harm, and Responsibility” in Owen (ed.), Philosophical Foundations of Tort Law (1995), p.321 at pp.339–345; cf. Honor´e, Responsibility and Fault (1999), at p.18). It follows that unless the outcome in question is foreseen, no such opportunity exists. However, Callinan and Heydon JJ.’s reasoning runs into difficulty because liability in negligence departs from notions of moral blameworthiness in so many profound ways that simply tightening the test of foreseeability would hardly reduce the vast chasm that exists between liability and culpability (regarding this gap, see Atiyah, The Damages Lottery (1997), at pp.33–38 and Goudkamp, “The Spurious Relationship between Moral Blameworthiness and Liability for Negligence” (2004) 28 M.U.L.R. 342). The reality is that the tort of negligence is so far out of kilter with considerations of blameworthiness that nothing short of a wholesale reworking of that tort would be sufficient to bring liability and culpability into step. As it is unlikely in the extreme that the courts would undertake such a reworking (assuming it to be within their powers), minor adjustments such as that proposed by Callinan and Heydon JJ. simply keep alive the unhelpful fiction that negligence law is based on culpability. Finally, it should be noted that although the issue of when a risk of injury is foreseeable has long attracted the interest of courts of final appeal, this is a matter that will gradually become less important in Australia. Governments in all Australian States, prompted by the 2001–2002 “insurance crisis”, enacted legislation, applicable to most personal injury actions, that modifies the breach element of the tort of negligence. That legislation provides that a defendant is not negligent in failing to take particular precautions unless the following conditions are satisfied: (1) the risk of injury was “foreseeable”, (2) the risk was “not insignificant” and (3) the reasonable person in the defendant’s position would have taken those precautions (see, e.g. Civil Liability Act 2002 (NSW) s.5B). Although the extent to which it is more difficult to show that a risk of injury is “not insignificant” than it is to establish that it is “not far-fetched or fanciful” is debatable, it is now harder to make out a breach of duty than had previously been the case. Consequently, as cases (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

to which the common law still applies are gradually finalised, the position at common law will become of less concern to insurers. The motivation to challenge it will slowly disappear. JAMES GOUDKAMP.* REFINING THE DUTY OF CARE IN SINGAPORE EVER since Lord Atkin espoused a general principle upon which to found a claim in negligence (Donoghue v Stevenson [1932] A.C. 562), courts around the common law world have endeavoured to formulate a precise, universal test for duty of care with varying degrees of success. At one stage, the English courts applied the Anns two-stage test (Anns v Merton London Borough Council [1978] A.C. 728), based on proximity and policy, which was later rejected in favour of the Caparo three-stage test (Caparo Industries Plc v Dickman [1990] 2 A.C. 605), based on reasonable foreseeability, proximity and what is fair, just and reasonable. Pure economic loss cases have sometimes been resolved by other, apparently independent tests, including assumption of responsibility and incrementalism (Customs and Excise Commissioners v Barclays Bank Plc [2004] EWCA Civ 1555; [2005] 1 W.L.R. 2082; cf. Customs and Excise Commissioners v Barclays Bank Plc [2006] UKHL 28; [2007] 1 A.C. 181, where the House of Lords expressed unease with this multiple approach but declined to clarify the law). Canadian and New Zealand courts have applied a variant of the Anns test, while the Australian courts have abandoned the search for a universal test. Apart from the tension between the various tests, another equally significant contest is between the underlying methodologies to be applied in developing the law of negligence: a choice between a “general principles” approach, where the court ignores earlier cases and merely applies the test to the instant facts (favoured in Anns and generally supporting a more expansionary approach) and an incremental approach where the court reasons by analogy from existing categories, using the test to take small steps rather than giant leaps forward (favoured in Caparo and generally supporting a more restrictive approach). The Singapore Court of Appeal in Spandeck Engineering (S) Ptd Ltd v Defence Science & Technology Agency [2007] SGCA 37, after an extensive review of Commonwealth and domestic jurisprudence, as well as academic writing, held that a universal test for duty of care could and should be formulated. The court essentially modified the test from Anns and grafted it onto the incremental methodology of Caparo. The Anns test was refined to limit the legal test for duty to proximity and policy, with reasonable foreseeability being relegated to a purely preliminary factual inquiry. * Visiting Fellow, Faculty of Law, University of Wollongong. Australia; Foreseeability; Personal injury claims

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

Briefly, the facts in Spandeck were that the claimant had entered into a contract with the Singapore Government to redevelop a facility for the Ministry of Defence. The defendant, a division of the Ministry of Defence, was appointed to supervise the project, and amongst other things, was responsible for certifying interim payment in respect of the claimant’s works under the project. The contractual structure, typical of its kind, was designed to avoid any direct relationship between the defendant and the claimant and to shield the Government from any liability to the claimant arising out of the defendant’s fault. The contract also provided for arbitration in case of disputes. Problems occurred after the claimant commenced work; the defendant alleged that the claimant was underperforming, and the claimant alleged that the defendant was undercertifying its works, thus not facilitating its progress payments. Eventually, the claimant sued the defendant in tort for economic losses incurred as a result of the defendant’s alleged negligence. The Court of Appeal upheld the trial judge’s decision dismissing the claim on the ground that there was no proximity between the parties. Proximity was determined by examining whether there was voluntary assumption of responsibility and reasonable reliance, as well as by considering the contractual matrix linking the three parties. On one reading, it may be that the Court of Appeal in Spandeck perhaps overemphasised the factual basis of voluntary assumption of responsibility, given that there is a long line of House of Lords authorities from Smith v Eric S Bush [1990] 1 A.C. 831 to Customs and Excise Commissioners reaffirming that “voluntary assumption of responsibility” is a legal inquiry based on an objective test. It is not whether the defendant in fact assumed or disclaimed responsibility (although that would be a material consideration), but whether the court, for the purpose of the duty inquiry, would deem the defendant to have assumed responsibility. Given the vulnerable position of the claimant in Spandeck, the absolute discretion vested in the defendant to certify payment and the disclaimer by the Government of any liability for the fault of the defendant, it is arguable that a proximate relationship between the parties, akin to that which was established in Smith, could have been found to exist. In support of its finding on proximity, the Court of Appeal in Spandeck relied on Pacific Associates Inc v Baxter [1990] 1 Q.B. 993, which it held was indistinguishable on the facts. A close reading of Pacific Associates suggests that the principal reason the Court of Appeal rejected a duty of care was due to the third limb of the Caparo test, i.e. it would not be fair, just and reasonable for obligations that had been deliberately structured under a contractual arrangement to be undermined by the imposition of additional tortious duties. Indeed, Russell L.J. rejected the duty solely on this ground, while Purchas and Ralph Gibson L.JJ. went further and rejected the duty on the first two limbs of the Caparo test as well. (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

However, their analyses suggest that their rejection of duty was largely based on the absence of reasonable foreseeability as a legal criterion. While it was factually foreseeable that the engineer’s negligence could have resulted in economic loss to the claimant, both judges were of the view that it was not reasonable to expect the engineer to foresee that the claimant would be so affected given the existence of the contractual structure which was clearly intended to remove such reliance (Pacific Associates at 1011, per Purchas L.J., 1031, per Ralph Gibson L.J.). The relegation of reasonable foreseeability from the test of duty of care is the controversial aspect of Spandeck. Reasonable foreseeability has long been at the heart of duty of care jurisprudence (Palsgraf v Long Island Railroad Co 284 N.Y. 339 (1928); Bourhill v Young [1943] A.C. 42). While it may have less value in the well-established areas of negligence relating to positive acts causing physical injury, it remains a useful device to control the duty of care where only non-physical damage is caused, e.g. pure economic loss and psychiatric injury. Pacific Associates itself shows how reasonable foreseeability was used as a normative inquiry to deny duty of care. Treating reasonable foreseeability as a purely factual inquiry, as opposed to a normative one based on reasonableness, also raises the question of whether an objective or subjective approach is to be employed. The reasonable foreseeability test in pure economic loss and psychiatric injury cases is more narrowly construed and, in effect, is closer to the remoteness test. It is not enough for the defendant to foresee that the claimant may be affected by the defendant’s conduct; the defendant must foresee that the claimant may suffer some form of economic loss or psychiatric injury for a duty of care even to be considered. Contrary to Pacific Associates, Spandeck held that it was not necessary for the defendant to have foreseen economic loss to the claimant, which inquiry the court said belonged to the realm of remoteness (at [89]). Reconceptualising reasonable foreseeability in this manner inevitably renders it redundant as a legal test. Indeed, it may well be that it is not necessary to have three layers of normative inquiry—reasonable foreseeability, proximity and policy: in that sense Spandeck is a bold decision that balances robust pragmatism and doctrinal integrity. Reasonable foreseeability and proximity have had a symbiotic relationship in the neighbour principle, which is the foundation of the modern tort of negligence. Returning to where it began, Lord Atkin held that a duty of care was owed to all our “neighbours”, whom we could reasonably foresee would be affected by our conduct (Donoghue v Stevenson, above at 580). He then defined “neighbours” as those who are “so closely and directly affected” by our conduct—this is the precursor to proximity. To determine the existence of a duty of care by ignoring reasonable foreseeability and (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Notes

merely focusing on proximity would be contrary to Lord Atkin’s formulation, which has two distinct, albeit greatly overlapping elements. First, it is only our legally recognised neighbours to whom we owe a duty of care, i.e. those who satisfy the proximity requirement. Secondly, it is only when it is reasonably foreseeable that our “neighbour” may be affected that a duty of care is owed to that neighbour. Reasonable foreseeability was more than a mere factual inquiry to decide the instant case; it was a control mechanism to define a recognised category of duty. Spandeck’s aim of articulating a universal test for duty is laudable and may well herald a universal approach that is doctrinally sound and pragmatic. Nevertheless, perhaps the test proposed could do with the minor refinement of reinstating the normative reasonable foreseeability inquiry. The Caparo three-stage process of reasonable foreseeability, proximity and what is fair, just and reasonable, analysed sequentially, provides a workable universal framework. Each element within the universal framework should continue to evolve to meet the unique requirements of the distinct categories of negligence. Thus, the reasonable foreseeability inquiry in the context of economic loss will be different to that with respect to physical injury. The proximity inquiry with respect to psychiatric injury will be different to that with respect to public authority liability. At the end of the day, the framework remains universal—but internally, it will be sufficiently nuanced to apply across all areas of negligence. KUMARALINGAM AMIRTHALINGAM.*

* National University of Singapore. Duty of care; Foreseeability; Proximity; Singapore

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

LEGISLATION THAT WOULD “PRESERVE” THE COMMON LAW: THE CASE OF THE DECLARATION OF INTENTION

“FIRST CITIZEN: What is the word reform? What does it mean? SECOND CITIZEN: Marry, it means leaving things as they are; I like it not.” Oscar Wilde, The Duchess of Padua (1883) Act II. PART 11 of the Criminal Justice Act 2003 made telling, and generally welcome, changes to the rule against hearsay. As the Court of Appeal has confirmed in Singh,1 ss.114 and 115 have given criminal, if not civil,2 evidence a new definition of hearsay, which excludes (at least, for the most part)3 those implied assertions whose prior inclusion by the House of Lords in Kearley 4 prompted such a barrage of criticism. Similarly, numerous situations which would have previously fallen within the hearsay rule’s purview for one reason or another now lie beyond its grasp.5 The Act, in what one court has designated “the labyrinthine recesses” of ss.116 and 117,6 has also hollowed out and (more likely than not) clarified statutory exceptions to the rule previously elaborated in ss.23–26 of the Criminal Justice Act 1988. Additionally, in s.118 the legislature chose to retain eight common law exceptions to the rule against hearsay. The fourth of these concerns res gestae. The purpose of this paper is to examine this not unimportant common law exception to the hearsay rule, and to explore one perplexing feature of the statute’s oddly wrought provision. At a parochial level the paper’s concern will be the relatively narrow issue of teasing out the true meaning of one contested feature of the res gestae doctrine, as set forth in s.118(1), r.4. 1 [2006] EWCA Crim 660; [2006] 1 W.L.R. 1564; (2006) 170 J.P. 222. 2 The definition of “hearsay” in civil proceedings is to be found in the Civil Evidence Act 1995 s.1(2)(a).

See generally, Cross and Tapper on Evidence, 11th edn (2007), esp. pp.625 et seq. 3 The court acknowledged that “the interrelationship between sections 114 and 115 is deeply obscure.” It held that, in addition to being admissible as ‘implied assertions’ under the Act’s new definitional provisions, the contested evidence was “also admissible under s.118(1) rule 7, as statements by an admitted co-conspirator against another party to the enterprise,” and for good measure, “[a] third possible route to admissibility is provided by section 114(2)(d) [sic]”: [2006] 1 W.L.R. 1564 at [14]–[15]. This is confusing. The Court of Appeal first tells us that the evidence falls outside the definition of hearsay, then buttresses its argument by suggesting two statutory exceptions to the rule against hearsay through which the evidence might also be admissible. It is very perplexing. Either the evidence is hearsay or it is not. The court surely cannot have it both ways. 4 [1992] 2 A.C. 228. 5 See, e.g. N(K) [2006] EWCA Crim 3309; (2007) 171 J.P. 158. 6 Kordansinki [2006] EWCA Crim. 2984; [2007] 1 Cr. App. R. 17; (2007) 171 J.P. 206 at [64], per May L.J.

JANUARY 2008] Legislation that Would “Preserve” the Common Law 47 However, interpretation of r.4 is inextricably intertwined with a perplexing question, of far deeper import: namely, what exactly does it mean when the legislature “preserves” a rule of common law. I. PARLIAMENT’S PRESERVATION OF THE COMMON LAW DOCTRINE OF RES GESTAE Section 118(1), r.4, so far as is relevant for present purposes, provides as follows: “Preservation of certain common law categories of admissibility (1) The following rules of law are preserved. ... Res gestae 4. Any rule of law under which in criminal proceedings a statement is admissible as evidence of any matter stated if— ... (c) the statement relates to a physical sensation or a mental state (such as intention or emotion).” (Emphasis added) Cutting to the chase, it will be seen that both s.118(1) and the heading refer expressly to the “preservation” of the common law categories of admissibility. At first blush, this would suggest that the existing common law rules governing res gestae are retained and that the ensuing rule restates that exception, as understood at the date of entry into force of the 2003 Act. However, r.4(c) refers to one retained common law exception as including a statement relating to “a mental state (such as intention . . .).” Evidence lawyers will immediately recognise that this provision presents a difficulty. The precise difficulty is that prior to this enactment, to put it kindly, at common law there existed profound doubt as to whether a declaration of intention actually was admissible in criminal cases under this particular common law exception. Almost certainly, the more orthodox view was that under English law res gestae were confined to declarations affecting present state of mind and did not extend to declarations of future intention. If that is so, s.118(1), r.4 poses a real interpretative puzzle. Was r.4(c) intended to encompass declarations of intention? If so, given that the section is expressed—and, indeed, more than once—in terms of “preservation” of certain existing “categories of admissibility”, what actually is the status of a party’s hearsay declarations of intention under the Criminal Justice Act 2003?7 7 In notes accompanying a presentation delivered in Leeds on May 19, 2005, David Ormerod wrote that “118(1) 4 c confirms that statements as to future intentions are to be treated as res gestae, removing

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

II. DECLARATIONS OF INTENTION AT COMMON LAW:

REMINDER

The res gestae exception relating to contemporaneous state of mind is of comparatively long standing. Such statements, it has traditionally been emphasised, must refer to contemporaneous states of mind. In general, this has meant that courts have tended only to admit statements of present state of mind—statements like “I am afraid”, “I am bemused”. Significantly, declarations of intention have regularly been excluded from the ambit of this exception where the purpose has been to prove that the intention was in fact carried out. Thus, in Wainwright 8 a murder victim’s statement that she intended to visit the defendant, W, was rejected by Sir Alexander Cockburn, C.J. on the ground that “it was only a statement of intention which might or might not have been carried out”.9 Similarly, in Thomson 10 a statement made by the victim of an illegal operation, indicating that she intended to perform the operation upon herself, was ruled inadmissible for the defence. After some initial hesitation, in Thomson Lord Alverstone, C.J. delivered a ruling that was both decisive and confident: one senses that by the late nineteenth century the courts considered that the question was closed.11 Moreover, it was settled that the prohibition applied without fear or favour to both prosecution and defence evidence. It is true that one particular nineteenth-century authority might at first have suggested that such statements could sometimes be adduced in evidence; but, on closer examination, this authority probably does no such thing. In Buckley 12 the court admitted, without demur, the words of a murdered policeman, who, before going out on patrol, had reported to his superior officer that he had heard that B was up to his old tricks again and that he was going out looking for him. The officer’s statement was received as evidence implying that PC Green had carried out his declared intention and that he had subsequently had a fateful encounter with B. In Buckley Lush J. did not in fact stipulate on what ground the police officer’s declaration of intention was admitted. Lush J. did step out of court briefly to confer with his brother judge, Mellor J. (as judges were wont to do at the time) before ruling on the matter. Significantly, Lush J.’s judgment makes no reference to the contemporaneous state of mind hearsay exception to the rule against hearsay, which could have been anticipated had that formed the basis of his decision. It is highly the ambiguity at common law” (The Sweet & Maxwell Criminal Justice Act Update Conference). He cited Buckley (1893) 13 Cox C.C. 293 as authority for the common law ambiguity. As will become clear in this paper, I hesitantly differ from David Ormerod, probably in my reading of both the history and the content of this area of law, as well as in my construction of the 2003 Act. 8 (1875) 13 Cox C.C. 171. 9 (1875) 13 Cox C.C. at 172. 10 [1912] 3 K.B. 19. See also Pook (1871) 13 Cox C.C. 172n. 11 “In our opinion there is no principle upon which this evidence is admissible any more than any other hearsay evidence”: [1912] 3 K.B. at 21. 12 (1873) 13 Cox C.C. 293. cf. fn.6, above.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 49 likely that, in 1873, Lush J. would have opted for the uncontroversial course of admitting the police officer’s statement under a quite distinct common law exception to the rule against hearsay—which has not been retained by s.118 of the Criminal Justice Act 2003—that admitted the declarations of deceased persons made in the course of duty.13 This would certainly explain why the inspector’s evidence in Buckley, as recorded in the law reports, focused tightly on the reporting procedures followed by his officers before they ventured out on patrol.14 Buckley, I would contend, in all likelihood has nothing whatever to do with the contemporaneous states of mind exception to the hearsay rule.15 Moghal 16 is another authority, sometimes taken to suggest that declarations of intention do in fact fall within this exception to the hearsay rule. Moghal, however, is another problematical case. M and his mistress, S, had initially been jointly indicted for the murder of S’s former lover. The murder could only have been committed by the two of them jointly, or by one or other of them. Owing to what might be generously described as unusual circumstances, S had successfully applied for a separate trial and been acquitted of murder, having run the defence that it was M who had committed the fell deed. At his subsequent trial, M, in turn, had cast the blame on S. At M’s trial the Crown conceded that it was S who had performed the actual killing and that, indeed, she had previously threatened to do exactly that. One contested issue on appeal arose from the circumstance that M had been wrongly discouraged from adducing evidence of a tape recording, made at a family conference roughly eight months before the murder, in which S announced her intention of killing the victim and accurately prophesied his death within months. This would have shown S’s motive to kill as well as the true character of her relationship with her “reluctant, cowardly and passive” lover, M. Scarman L.J. held that the tape recording would have been admissible because S’s feelings and state of mind before and at the time of the killing were relevant facts at M’s trial. Although contemporaneity is a matter of degree, the court was clear that the tape recording of S’s truculent and overbearing declarations were “relevant to [M’s] defence” that S alone had committed the killing and, therefore, admissible. The judgment, it has to be said, is not especially sure-footed or clear. It conveys a strong impression that, along with the Crown’s acknowledgement that it was S who had actually stabbed the victim—and 12 other obliging Crown concessions—Scarman 13 See, e.g. Cross on Evidence, by Sir Rupert Cross and Colin Tapper, 6th edn (1985) p.573, fn.9; Colin Tapper, “Hillmon Rediscovered and Lord St Leonards Resurrected” (1990) 106 L.Q.R. 441 at p.461, fn.80; Ian Dennis, The Law of Evidence (2002), p.595. 14 e.g. “If a constable was going on a particular duty he should report to me, if convenient. Green did make a report to me on that occasion of his intended duty that night. That report was verbal.” 15 This could explain why a Chief Justice and a Lord Chief Justice of England each omitted to refer to Buckley in their rulings in Wainwright and Thomson respectively. 16 (1977) 65 Cr. App. R. 56.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

L.J. was straining to show that every allowance was being made to assist M’s case even though, in the ultimate resort, his conviction would be allowed to stand. Professor Dennis, I would suggest, hits the nail square on the head: “It is very doubtful whether [Moghal ] can be taken as authority for any proposition of law. Thomson [and, indeed, Wainwright for that matter] was not cited, the dictum regarding the admissibility of the pre-act declaration was subsequently doubted by Lord Bridge in Blastland,17 and in any event the prosecution had conceded that the declarant was the principal party in the murder.”18 Moghal is of deeply dubious authority, therefore.19 In post-2003 terms, Moghal, if anything, resembles the sort of case which provoked the Law Commission, in its report on the hearsay rule, to argue that for certain defendants an inclusionary discretion, now enacted in s.114(1)(d) of the Criminal Justice Act 2003, was an essential evidentiary perquisite.20 At common law it can be seen that the admissibility of declarations of intention was, at best, unclear at English common law; the dominant view was that such evidence was inadmissible under the res gestae exception presently under discussion. Nor was England the only jurisdiction to encounter difficulty in determining whether to admit statements of intention in order to suggest that the intention was carried through. Canada, which operates a more flexible approach to the admission of hearsay than English law prior to the Criminal Justice Act 2003, has been similarly troubled by this situation. In 2000 its Supreme Court had to consider the admissibility of declarations of intention under this res gestae exception. In Starr 21 that court divided 5:4. Whereas four judges would have admitted such evidence, the majority held that evidence of a murder victim’s declaration of intention had been wrongly admitted by the lower courts. The victim, C, had been the target of a gangland “hit”. C had earlier told a witness, G, that he had to “go and do an Autopac scam with (the defendant)”, the implication being that C could be assumed to have carried out his stated intention and to have met up with the defendant. 17 [1986] A.C. 41 at 60. 18 Dennis, The Law of Evidence, fn.13 above, p.595, fn.20. 19 Interestingly, when the decision first appeared, commentators seem not to have spotted that Moghal

had wrought this great change in the status of declarations of intention: [1977] Crim. L.R. 373. Another dictum that one could mention for the sake of completeness, for which no particular authority was cited, is Beldam L.J.’s remark in Gilfoyle: “When the intentions or state of mind of a person making the statement are relevant to a fact in issue, hearsay evidence is admissible” ([1996] 1 Cr. App. R. 302 at 321). 20 Evidence in Criminal Proceedings: Hearsay and Related Topics (Law Com. No.245, 1997), para.8.133: “Our purpose is to allow for the admission of reliable hearsay which could not otherwise be admitted, particularly to prevent a conviction which that evidence would render unsafe.” But cf. Finch [2007] EWCA Crim 36 at [26], per Hughes L.J.: “It is not . . . the law that every reluctant witness’s evidence automatically can be put before the jury under s.114.” 21 [2000] 2 S.C.R. 144.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 51 The majority gave as its principal reason for excluding G’s evidence its inherent unreliability. More specifically, they found that C’s declaration had been made in “circumstances of suspicion” when C might have had a motive for lying to G, who happened to be a former girlfriend. Academic literature In the literature there is universal agreement that the evidential status of declarations of intention is at best moot. Some writers have simply designated them the “present tense” exception to the hearsay rule. Murphy, for instance, emphasises that past physical and mental states have not normally been admissible under this hearsay exception, without feeling a need to mention that future intentions fall outside the exception as well.22 When it comes to declarations of intention which are being used to prove that the intention was carried out, Ian Dennis notes not only that it is “surprising how little authority there is on this question, given its apparent importance” but also that “The English cases on the point are not consistent, but the majority support the view that such a declaration is not admissible to prove performance of the act.”23 Roberts and Zuckerman cautiously state: “English precedents are divided on the question whether a declaration of intention can be admitted to prove the intention was subsequently carried out.”24 The late Sir Richard May’s take on the matter was that there were “conflicting authorities as to the scope of the rule” and that, in likelihood: “in the absence of any modern authority defining the scope of this rule, it is thought that such evidence would require a high degree of probative force to be admitted today.”25 Because this cluster of cases proves so problematical and their reconciliation inevitably tentative, many writers have sought the answer in overseas authorities. Most frequently, they have pored over the conflicting judgments of the Australian High Court in Walton v R.26 Judges in other jurisdictions have sometimes displayed a greater willingness to allow such evidence to be presented to the tribunal of fact either by taking declarations of intention out of the realm of hearsay altogether, by treating them 22 Murphy on Evidence, 9th edn, (2005), pp.239–40. 23 The Law of Evidence, 2nd edn (2002), p.594. 24 Criminal Evidence, 2nd edn (2004), p.651. 25 May and Powles, Criminal Evidence, 5th edn (2004), para.8.60, p.218. 26 (1989) 166 C.L.R. 283.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

(mysteriously) as analogous to “conduct”, or by judging their admissibility in terms of their likely reliability. Thus, one leading text, Phipson, having swiftly outlined how “the English case law on this point is in disarray”,27 proceeds to analyse the disparate views expressed by Australian judges on the question in Walton v R. itself, as well as in the subsequent case of Bull.28 Choo, too, feels that “a sensible approach” might be to follow Mason C.J.’s view in Walton,29 although it needs to be recognised that Walton has not found favour with every English evidence scholar.30 Fortunately, none of this matters much in the context of the present study. For a reason which will be spelled out presently, it seems unnecessary to conduct a detailed review of the overseas case law. For the time being, suffice it to say that English law is at best uncertain about, at worst hostile towards, the admissibility of declarations of intention where the implication is that the intention was carried out. The better view, I would contend, is that hitherto these declarations have fallen outside the hearsay exception under discussion. III. THE SHAPE OF THE PROBLEM By now, it will be apparent that two conjoined questions have to be considered: (i) What is actually “preserved” by s.118(1), r.4? This involves making sense of the relevant case law in the context of the 2003 Act. And (ii) What, if anything, may the courts do in the future either to modify the scope of the rule—possibly, making the rule conform to likely parliamentary expectations—or to develop and re-model the preserved rule, as and when need arises? These questions are far from easy. Nevertheless, before addressing them, it might briefly be mentioned that, paradoxically, the Criminal Justice Act 2003 can be said to have dispelled one (relatively minor) uncertainty concerning declarations of intention. No longer is it possible to dodge the bullet by maintaining that a declaration of intention does not constitute 27 Phipson on Evidence, 16th edn (2005), para.31.34. The same term is employed by Cross and Tapper on Evidence, 11th edn, (2007), p.613. 28 (2000) 201 C.L.R. 443. 29 Evidence (2006), §4.2.1.3. See also A. Choo, Hearsay and Confrontation in Criminal Trials (1996, p.135. In fact, this approach is now problematical under the 2003 Act: to the extent that Mason C.J.’s argument in Walton was that the defendant’s declaration of intention to meet and kill the victim, as reported to the court by his wife, ought actually to have been viewed as original evidence which had “independent evidentiary value in proving the author’s intentions, those intentions being a fact in issue or a fact relevant to a fact in issue” ((1989) 166 C.L.R. 283 at 288), the Chief Justice’s thesis takes such statements outside the realm of hearsay altogether. Were this analysis accepted in England, the words “such as intention” in s.118(1), r.4(c) would be rendered redundant. That makes no sense. 30 e.g. C. Tapper, “Hillmon Rediscovered and Lord St Leonards Resurrected” (1990) 106 L.Q.R. 441.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 53 hearsay evidence at all, because that is exactly how it is introduced in s.114(1)(b) of the 2003 Act. So far as relevant, the provision reads: “Admissibility of hearsay evidence (1) In criminal proceedings a statement not made in oral evidence in the proceedings is admissible as evidence of any matter stated if, but only if — ... (b) any rule of law preserved by section 118 makes it admissible . . .” (Emphasis added) The declaration of intention, then, is one of a catalogue of exceptions to the re-drafted hearsay rule. If it has only clearly achieved one thing, Parliament has conclusively ruled out the notion that a declaration of intention is original evidence, beyond the clutches of the rule against hearsay—an idea encouraged by Mason C.J., for example, in Walton v R.31 IV. WHAT IS THE LIKELY EFFECT OF S.118(1), R.4? What, then, is the likely effect of s.118(1), r.4? If one accepts that, according to the English case law, it is very doubtful whether declarations of intention fell within the common law’s understanding of res gestae, has the 2003 legislation now resolved such doubts one way or another? The Law Commission’s Final Report on Hearsay The ancestry of s.118(1), r.4(c) can be retraced directly to cl.6(5)(c) of the Law Commission’s Draft Bill on hearsay evidence, which was appended to its final report, Evidence in Criminal Proceedings: Hearsay and Related Topics. The wording of this provision is not utterly dissimilar from that used in the Act.32 In retrospect, the somewhat truncated account of the law given in the body of the Law Commission’s report, however, may not have been altogether helpful. The report advocated retention of the common law exception, “if the statement relates to . . . a mental state (such as intention . . .).”33 It did discuss the content of this branch of the res gestae doctrine. However, discussion was restricted to one footnote, which made it seem 31 (1989) 166 C.L.R. 283 at 288. Arguably, this positive achievement of s.114(1)(b) is more than offset by the puzzling phenomenon of traditional common law exceptions being retained in the 2003 Act but being applied to a newly-defined species of hearsay. 32 Clause 6(1) reads: “The rules of law to which this section applies are preserved.” Clause 6(5) then provides that “[t]his section also applies to any rule of law under which in criminal proceedings a statement is admissible as evidence of any matter stated if . . .(c) the statement relates to . . . a mental state (such as intention or emotion).” 33 LC No.245, above fn.26, para.8.126.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

that the only relevant authority on the point was the Court of Appeal’s decision in Moghal. In fairness, the report did add that Moghal had been “doubted” in the House of Lords in Blastland, “but on the grounds that it was an isolated declaration of intention made six months before the murder and was thus insufficiently relevant.”34 Otherwise, the scope of the exception was not really ventilated in the report. What may have been a well-meaning desire on the part of the reporters either to spare readers the Calvary of trying to make sense of the contorted case law, or even to impose a particular interpretation of the exception—the latter suggestion is, of course, just wild speculation—complicates the picture further. Parliament clearly intended to preserve the existing common law exception because it said so. If one wished to wrangle, Parliament could more or less plausibly be said to have taken the content of that exception from the Law Commission’s report, even if the exact wording of the Commission’s proposed Bill did not survive the legislative phase. But if the Law Commission was in error in its description of that exception—which may well have been the case—which version ought Parliament to be taken to have enacted : the orthodox or the heterodox? The Explanatory Notes accompanying the Criminal Justice Act 2003 Although this author has repudiated the authority of these wretched documents,35 ex abundanti cautela brief reference might be made to the Act’s Explanatory Notes. They shed little light on the matter. Paragraph 411—which, incidentally, uses the word “preserve” on no less than six occasions—repeats what we already know: namely, that s.118 “preserves a number of common law exceptions to the old rule against the admission of hearsay evidence”, simply noting that “ ‘res gestae’ will be admissible.” Paragraph 412 is marginally more informative. It explains that “one justification for this exception is that reported words which are very closely connected to a relevant event are reliable accounts and should therefore be admissible in certain circumstances.”36 Then, the note continues: “Such statement may be admitted if one of the following conditions is met: ... the statement relates to a . . . mental state, such as an intention or emotion.” 34 LC No.245, above fn.26, para.8.125, fn.173. 35 e.g. “Explanatory Notes and Statutory Interpretation” (2006) 170 J.P. Jo. 124. 36 Pursuing my vendetta against the use of these notes as authoritative aids to construction, I would

have been fascinated to have heard the Home Office lawyers’ account of the other justifications for this hearsay exception at which they broadly hint in this annotation.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 55 The assumption seems to be that what has been “preserved” is an exception that does indeed incorporate declarations of intention. This is nevertheless an offbeat reading of the old law because, as we know, the very reason for common lawyers’ suspicion of such statements was precisely, to adopt the language of the Explanatory Notes themselves, that these reported words were not necessarily very closely connected to a relevant event and, therefore, could not be treated as reliable accounts of what actually occurred. The Explanatory Notes, then, carry us no further forward.37 Rather, they appear to compound the confusion. Two close legislative precedents The legislative drafting device of “preserving” common law rules is not unique to s.118 of the Criminal Justice Act 2003. Section 118 is drafted along broadly similar lines to s.7 of the Civil Evidence Act 1995, which, in turn, superseded s.9 of the Civil Evidence Act 1968. Section 7 of the 1995 Act provides: “(2) The common law rules effectively preserved by s.9(1) and (2)(b) to (d) of the Civil Evidence Act 1968, that is, any rule of law whereby in civil proceedings— (a) published works dealing with matters of a public nature (e.g., histories, scientific works, dictionaries and maps) are admissible as evidence of facts of a public nature stated in them, (b) public documents (e.g., public registers, and returns made under public authority with respect to matters of public interest) are admissible as evidence of facts stated in them, or (c) records (e.g., the records of certain courts, treaties, Crown grants, pardons and commissions) are admissible as evidence of facts stated in them, shall continue to have effect. (3) The common law rules effectively preserved by s. 9(3) and (4) of the Civil Evidence Act 1968, that is, any rule of law whereby in civil proceedings— (a) evidence of a person’s reputation is admissible for the purpose of proving his good or bad character, or (b) evidence of reputation or family tradition is admissible— (i) for the purpose of proving or disproving pedigree or the existence of a marriage, or 37 The issues considered in the present paper seem not to have arisen during discussion in Standing Committee (now, Public Bill Committee) B, 15th and 16th sittings, both held on January 28, 2003.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

(ii) for the purpose of proving or disproving the existence of any public or general right or of identifying any person or thing, shall continue to have effect in so far as they authorise the court to treat such evidence as proving or disproving that matter. Where any such rule applies, reputation or family tradition shall be treated for the purposes of this Act as a fact and not as a statement or multiplicity of statements about the matter in question.” Does the civil enactment—which, let it be noted, retains several of the same rules of common law as feature in s.118(1) of the criminal statute—shed light on this notion of “preservation” of rules of common law? I am not persuaded that it does. First, apart from the very occasional glancing reference,38 the formulation of s.7 seems not to have given rise to any dispute in the courts. Therefore, there are no judicial intimations of what “preservation” might signify. Secondly, as is the case with the overwhelming majority of the eight common law rules retained in s.118, in many instances the common law rules “preserved” in civil cases by s.7 represent undisputed exceptions to the rule against hearsay that have been settled, and fixed in their courses, for centuries.39 What is entirely distinctive about s.118(1), r.4 is that, to this day, the admissibility of declarations of intention is anything but settled at common law. Thirdly, there is an argument that in the Civil Evidence Act 1995 s.7(2) and (3) the legislature forsook the wording of the Civil Evidence Act 1968, which had spoken in terms of “preserving” certain rules of common law. In place of the concept of “preservation”, the 1995 Act confined itself to stating that such rules had been “effectively preserved” and “shall continue to have effect”. Fourthly and finally, there is what appears at first to be a material difference between the drafting of s.7 and s.118. Critically, in s.7(4) the Civil Evidence Act 1995 provides that:

38 e.g. Martin v Myers [2004] EWHC 1947 (Ch) at [17], per Nicholas Strauss, Q.C. 39 The hearsay rule, in the words of one writer, only “began to mature in its modern form during the

late eighteenth century. The majority of that jumble of exceptions to the rule seems to have been in place before the advent of the nineteenth century—virtually all save the res gestae exception”: T.P. Gallanis, “The Rise of Modern Evidence Law” (1999) 84 Iowa L. Rev. 499 at 533. Statutes do not always take the common law as a point of reference in an identical manner. If, say, one looks at the Occupiers’ Liability Act 1957, which legislated on the duties occupiers owe to visitors, s.1(2) of that Act provides that “The rules so enacted . . . shall not alter the rules of the common law as to the persons on whom a duty is so imposed or to whom it is owed; and accordingly . . . the persons who are to be treated as an occupier and as his visitors are the same . . . as the persons who would at common law be treated as an occupier and as his invitees or licensees.” This identifies which classes of person are affected by the new enactment, and it seems clearly to have been envisaged that those well-established classes of person would not be extended or reduced following the Act’s entry into force.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 57 “The words in which a rule of law mentioned in this section is described are intended only to identify the rule and shall not be construed as altering it in any way.”40 One might seek to argue that since s.7(4) of the Civil Evidence Act expressly precludes any suggestion that the rules, as described in the statute, have to the slightest degree been extended or modified whereas the Criminal Justice Act does not do so, the latter enactment does not prevent the courts from developing the rules. Therefore, even if declarations of intention have been incorporated into r.4 because Parliament was mistaken as to the true scope of the common law rule, it is open to the courts to set matters “right” if that is what they choose to do. This fourth argument has some initial appeal. It is not obvious, however, that Parliament’s intention can be clearly read in this way. The meaning of s.7(4) itself is unclear. On the one hand, the subsection might signify that not only does the Act not alter the preserved rules but it is not anticipated that the courts will modify them either. According to this construction, the common law rules are forever fixed in amber, like Martial’s bee.41 On the other hand, s.7(4) might be read so as to leave it open to the judges to develop or contract the designated rules within the framework of the law in their accustomed manner, a living organism fitted snug within the greater structure. According to this reading of s.7(4), nothing really can be read into Parliament’s omission to incorporate an identical or equivalent provision into s.118 of the Criminal Justice Act 2003. Parliament might even have assumed (wrongly, I would submit) that, as a result of the success enjoyed by s.7(4)—which has not been the subject of any meaningful litigation—the “preservation” of rules of common law is an established device which now neither requires legislative explanation nor amplification. Of course, were we to imagine that the judges today took it into their heads to extend one of these preserved rules—for example, by determining that anything published by whomsoever on the internet constituted “published works dealing with matters of a public nature” within the meaning of s.7(2)(a) of the 1995 Act or of s.118(1), r.1 of the 2003 Act, which is framed in identical terms—this question would erupt into life in both civil and criminal domains. The Civil Evidence Act 1995 states that it retains rules “effectively preserved” by s.9 of the Civil Evidence Act 1968. Section 9, it might be noted, was more felicitously expressed than the equivalent provisions in the 1995 and 2003 statutes. The 1968 Act, so far as pertinent, ran:

40 A similar provision, s.9(6), used to be found in the now repealed Civil Evidence Act 1968. 41 “Et latet et lucet Phaethontide condita gutta, ut videatur apis nectare clusa suo.” “The bee is enclosed,

and shines preserved in amber, so that it seems encased in its own nectar”: Martial, Epigrams IV, 32.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

“(1) In any civil proceedings a statement which, if this Part of this Act had not been passed, would by virtue of any rule of law mentioned in subs.(2) below have been admissible as evidence of any fact stated therein shall be admissible as evidence of that fact by virtue of this subsection.”42 Under the 1968 Act, then, courts were spared the need to wrestle with the word “preserve”.43 That the legislature uses the term “preserve” in both the later enactments may in fact complicate our problem as it suggests that what was meant was something other than simply the retention of the rules of common law which would have rendered admissible certain categories of hearsay but for the intervention of the legislature. The 1995 Act’s declaration, at two points, that the common law rules were “effectively preserved” adds to one’s puzzlement, as it suggests that there may be two separate processes: “effective preservation”, as described in s.9(1) and (3) of the 1968 Act, and “preservation”, as employed in s.118(1) of the 2003 Act. “ ‘Curiouser and curiouser!’ cried Alice (she was so much surprised, that for the moment she quite forgot how to speak good English).”44 This leads us to another short, connected point. The common law rules, preserved in virtually identical language by both the civil and criminal enactments, are of sufficient antiquity as to hail from an era when both civil and criminal law shared evidential principles. The fact that they have for long enjoyed this identity of form and have been retained by Parliament in similar, if not identical, language could suggest that whatever construction we were to place upon the preservation of the one, ex hypothesi a similar construction ought likewise to be imposed upon the other. The internal evidences Whilst there exists a case for saying that r.4 may quite possibly have been enacted upon the assumption that declarations of intention fell within the res gestae doctrine, the language of s.118 does not ineluctably point to such a conclusion. Rule 4 simply speaks of the preservation of

42 s.9(3) of the 1968 Act is drafted in similar terms. 43 See, e.g. Humberside CC v DPR (an Infant) [1977] 1 W.L.R. 1251. Judges did occasionally use the

term in their judgments, however: e.g. Hewgill v Hewgill, unreported, October 31, 1984, per Purchas L.J.; Kuwait Oil Tanker Co SAK v Al Bader et al., unreported, December 17, 1998, per Moore-Bick J. In H v H (Minor) (Child Abuse: Evidence) [1990] Fam. 86 Croom-Johnson L.J. remarked: “Section 9 preserves certain of the old common law exceptions to the hearsay rule, but the words in s.1(1) ‘but not otherwise’ prevent the introduction of any others” (emphasis added). 44 Lewis Carroll, Alice’s Adventures in Wonderland (1866), Ch.2. One does just wonder whether the use of the word, “effectively”, was an error of English. Perhaps the draftsman meant to say “in effect preserved”? Yet, even that expression conveys strange connotations.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 59 “any rule of law under which in criminal proceedings a statement is admissible as evidence of any matter stated if . . . the statement relates to a . . . mental state (such as intention . . .)” The fact that Parliament chose the casual, possibly tentative, wording, “any rule of law under which”—as opposed, say, to “those rules of law that”—looks almost designed to leave open the possibility that there exists no such rule. Uncertainty is factored into the formula. Against this backdrop, the equally casual expression, set in parentheses, “such as”, looks as much like a vague, suggested line of inquiry as a reference to an existent rule of law. The very language of the Act, then, does little to dispel our perplexity. In fact, the real puzzle is what other “mental states” might Parliament have had in mind when it inserted the phrase, “such as”. Evidence lawyers, once again, are left scratching their heads, mystified. (So, probably, are behavioural scientists.) More important still, one might imagine, is the circumstance that what the Act explicitly “preserves” is the “rule of law” under which such statements are admissible. The natural meaning of this expression seems to enjoin us to identify the true scope of the common law rule as it existed prior to the enactment of the Criminal Justice Act 2003. Back to square one, then. Moreover, it may just be possible to say that had the legislature meant to be prescriptive—either altering the content of a rule or even settling a moot point—one could have legitimately anticipated that Parliament would have made this explicit. After all, when particular restrictions apply to a preserved rule, Parliament has taken this step. Thus, the common law rule preserved in s.118(1), r.2 of the 2003 Act, according to which “evidence of a person’s reputation is admissible for the purpose of proving his good or bad character”, is accompanied by the following note: “Note The rule is preserved only so far as it allows the court to treat such evidence as proving the matter concerned.” Overseas solutions One hypothesis can be fairly confidently discounted. We have seen that a number of the leading texts, despairing of making sense of the English case law on declarations of intention, have resorted to overseas authority. Not infrequently, they have been drawn to the approach espoused by Mason C.J. in his judgment in Walton v R.45 It can safely be assumed that Parliament was not endeavouring to “preserve” the common law of Australia or to subscribe to some romantic notion of either a “common 45 (1989) 166 C.L.R. 283.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

law of the Commonwealth” or of the common law of any other of our former colonies. It is the English rules of common law, without frills or furbelows, that have been preserved. Hypotheses advanced by these text writers, therefore, will have no bearing on our question: What can s.118(1), r.4 legitimately be taken to mean? Attaining an evidential fixity To the extent that our question is capable of any authoritative answer, a key consideration is likely to be what we identify as the role of s.118 within the overall scheme of c.II of Pt 11 of the Criminal Justice Act 2003, which is entitled “Hearsay Evidence”. Here, too, it would seem that two contradictory philosophies present themselves. On the one hand, it could be contended that the object of c.II was to fix, once and for all, in a complete, clear, codified form the mazy precedents and scattered enactments that had previously made up that notoriously protean rule against hearsay. It presented a fresh, authoritative definition of what actually constitutes hearsay evidence, together with an exclusive list (“if, but only if”) of exceptions to the rule against hearsay, including “any rule of law preserved by s.118”.46 True, in some instances the provisions of c.II are expansively drafted. However, one interpretation would be that the statute was designed to instil certainty into what, traditionally, was one of the most fraught areas of English law. This would imply that the intention was to nail down those common law rules, selected for retention and enumerated in s.118.47 The legislature’s expectation would not have been that the courts would blithely continue developing their “common law”. Once incorporated into the statutory schema, common law method no longer would have a valid place in fashioning future law. “Preservation”, following this approach, would mean that a court construing a provision like s.118(1), r.4 would approach the task in the same way that it would in construing any other statutory enactment. The text becomes the matrix; the law is no longer permitted to drift from precedent to precedent. It is noticeable, after all, that s.118(1), r.4 chooses to speak not of rules of common law but refers, pointedly perhaps, to “rules of law” simpliciter. Twice. The language of the statute virtually implies that Parliament’s wish was to denature the common law’s creations. An opposing view, however, would construe the Act’s preservative provisions from a purposive angle. The argument would run, the 2003 Act was meant to distance the rule against hearsay from that divorce between the reliability of evidence and its admissibility, which had previously done 46 Criminal Justice Act 2003 s.114(1)(c). 47 “With the exception of the rules preserved by this section, the common law rules governing the

admissibility of hearsay evidence in criminal proceedings are abolished”: Criminal Justice Act 2003 s. 118(2).

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 61 little to enhance the reputation of the Law of Evidence. Evidence lawyers well recall the words of no lesser judge than Lord Reid in Myers v DPP: “The whole development of the exceptions to the hearsay rule is based on the determination of certain classes of evidence as admissible or inadmissible and not on the apparent credibility of particular evidence tendered. No matter how cogent particular evidence may seem to be, unless it comes within a class which is admissible, it is excluded. Half a dozen witnesses may offer to prove that they heard two men of high character who cannot now be found discuss in detail the fact now in issue and agree on a credible account of it, but that evidence would not be admitted although it might be by far the best evidence available.”48 The Law Commission’s objective had been, subject to satisfactory safeguards, to liberate English law from this astonishing philosophy and to seek to admit relevant evidence whenever its admission coincided with the interests of justice. Although the Government made the sweeping claim during the passage of the Criminal Justice Bill that it had remodelled its provisions along the more “progressive” lines sketched out in Sir Robin Auld’s Review,49 the basic structure proposed by the Law Commission was in fact retained in the Act. To receive declarations of intention as an exception to the hearsay rule, it could be argued, is to fulfil the purpose of the legislature by admitting evidence on the ground that it is relevant—although this will not invariably be true. The only problem with this second approach is that it places considerable strain on the natural meaning of the words “preserve” and “preservation”, whose connotations are pointedly conservative. V. TOWARDS A CONCLUSION In practice, does this issue greatly matter? For practical purposes, it could be said that it is of little import whether or not declarations of intention qualify under s.118(1), r.4. After all, if a statement is inadmissible under this preserved common law exception, it might remain open to a court to admit a declaration of intention under the inclusionary discretion provided for in s.114(1)(d) of the 2003 Act, which reads: “In criminal proceedings a statement not made in oral evidence in the proceedings is admissible as evidence of any matter stated if, but only if— ... 48 [1965] A.C. 1001 at 1024. 49 Review of the Criminal Courts of England and Wales (London: HMSO, 2001). See D. Birch,

“Hearsay: Same Old Story, Same Old Song?” [2004] Crim. L.R. 556 at 558–559.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

(d) the court is satisfied that it is in the interests of justice for it to be admissible.” True, admissibility under s.114(1)(d) is not guaranteed. Before a court can admit evidence under this inclusionary discretion, even if it is now excused from reaching a publicly declared conclusion on each and every one of those nine and more matters,50 it must still weigh all nine factors listed in s.114(2), along with any others it considers relevant. However, should the court determine that it will be in the interests of justice to admit the declaration of intention—and, let it be noted, the courts are resorting to s.114(1)(d) quite readily51 —in its ruling it cannot neglect to state the logical steps that have led it to resort to s.114(1)(d). As Leveson J. observed in Maher v DPP: “Although the purpose of the hearsay provisions . . . was undeniably to relax the previously strict rules against the admission of hearsay, it is important to underline that care must be taken to analyse the precise provisions of the legislation and ensure that any route of admissibility is correctly identified.”52 This minimal demand for intellectual rigour ensures that our problem cannot simply be ignored. Alternatively, as Colin Tapper has pointed out,53 the evidence of intention in most of the cases under consideration would in likelihood be admissible under s.116 of the 2003 Act. Section 116, which deals with cases where witnesses are unavailable, so far as relevant, provides: “(1) In criminal proceedings a statement not made in oral evidence in the proceedings is admissible as evidence of any matter stated if— (a) oral evidence given in the proceedings by the person who made the statement would be admissible as evidence of that matter, 50 Taylor [2006] 2 Cr. App. R. 14. See, e.g. M and others [2007] EWCA Crim 219 at [17]–[26]. 51 See further, R. Munday, “The Judicial Discretion to Admit Hearsay Evidence” (2007) 171 J.P. Jo.

276. 52 (2006) 170 J.P. 441 at [26]. Translated into terms of s.114(1)(d), if it exercises its inclusionary discretion a court must have attributed adequate weight to all the factors enumerated in s.114(2): see McEwan v DPP [2007] EWHC 740 (Admin); (2007) 171 J.P. 308, where the Divisional Court determined that justices had failed to give any, or any sufficient, weight to the prejudicial effect that the defendant’s inability to challenge a supposedly sick witness’s evidence would have had on the former’s case. As Latham L.J. observed in S [2007] EWCA Crim 335 at [15], “it is unfortunate that [the judge] failed to deal with the matters set out in [s.114(2)] which are required before a statement under section 114(1)(d) can be admitted. However clear it may be to a judge that the document in question should be admitted in the interests of justice, it is nonetheless incumbent upon him to consider the matters set out in subs.(2), and in particular, in the present context, the provisions requiring the judge to consider under (h) the amount of difficulty involved in challenging a statement or (i) the extent to which that difficulty will be likely to prejudice the party facing it.” 53 Cross and Tapper, fn.2 above, p.614, fn.216.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 63 (b) the person who made the statement (the relevant person) is identified to the court’s satisfaction, and (c) any of the five conditions mentioned in subsection (2) is satisfied. (2) The conditions are— (a) that the relevant person is dead; (b) that the relevant person is unfit to be a witness because of his bodily or mental condition; (c) that the relevant person is outside the United Kingdom and it is not reasonably practicable to secure his attendance; (d) that the relevant person cannot be found although such steps as it is reasonably practicable to take to find him have been taken; (e) that through fear the relevant person does not give (or does not continue to give) oral evidence in the proceedings, either at all or in connection with the subject matter of the statement, and the court gives leave for the statement to be given in evidence.” If the witness (i) could have given the evidence in person were they available as a witness, (ii) is identifiable to the satisfaction of the court, and (iii) fulfils any one of the five conditions laid down in s.116(2), the court has to treat the hearsay statement as admissible. If this is correct, the practical impact of the point discussed in this paper is likely to be relatively small. Correcting an obvious drafting error? Casting about for solutions, the question might arise whether the courts can treat s.118(1), r.4 as an obvious drafting error which they would then be entitled to rectify. In Inco Europe Ltd v First Choice Distribution the House of Lords reviewed this judicial power, affirming that where it is clear that the language employed by the draftsman has failed to give effect to Parliament’s plain intention, it is permissible to read words into an enactment to give effect to that undoubted intention.54 As Lord Nicholls of Birkenhead explained: “I am left in no doubt that, for once, the draftsman slipped up . . . The language used was not apt to achieve [the desired] result. Given that the intended object of [the provision] is so plain, the paragraph should be read in a manner which gives effect to the parliamentary intention . . . The role of the courts in construing legislation is not 54 [2000] 1 W.L.R. 586.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

confined to resolving ambiguities in statutory language. The court must be able to correct obvious drafting errors.”55 His Lordship added: “This power is confined to plain cases of drafting mistakes. The courts are ever mindful that their constitutional role in this field is interpretative. They must abstain from any course which might have the appearance of judicial legislation. A statute is expressed in language approved and enacted by the legislature. So the courts exercise considerable caution before adding or omitting or substituting words.”56 Lord Nicholls stipulated that, before exercising this power, a court had to be “abundantly sure” of three matters: (i) the intended purpose of the relevant provision; (ii) the fact that, through inadvertence, the draftsman had failed to give effect to the intended purpose of that provision; and (iii) “The substance of the provision Parliament would have made, although not necessarily the precise words Parliament would have used, had the error in the Bill been noticed.”57 The third requirement, his Lordship stressed, in particular, was “of crucial importance. Otherwise any attempt to determine the meaning of the enactment would cross the boundary between construction and legislation.”58 Once in a while the courts employ this faculty to correct drafting oversights. In one recent case, R. (CPS) v Bow Street Magistrates’ Court,59 it was obvious that Sch.2 to the Identity Cards Act 2006 had inadvertently repealed the provision of the Forgery and Counterfeiting Act 1981 which made it an offence to possess a false passport before the replacement offences of possessing false identity documents under ss.25 and 26 of the 2006 Act had been brought into force. Instead of the smooth transition sought for, the draftsman had created an embarrassing hiatus. Forbes J. declared himself “abundantly sure” that the purpose of the legislation was to secure a smooth transition, and also that he was “satisfied” that Lord Nicholls’s two further conditions were both met so a degree that justified the court in correcting the drafting error.60

55 [2000] 1 W.L.R. at 592. 56 [2000] 1 W.L.R. at 592. 57 [2000] 1 W.L.R. at 592. 58 [2000] 1 W.L.R. at 592. 59 [2007] 1 W.L.R. 291 60 [2007] 1 W.L.R. at [41]–[43].

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 65 In all probability, in the case of declarations of intention the issues are insufficiently clear-cut to satisfy Lord Nicholls’ three criteria and to justify a court correcting an “obvious drafting error”. Despite the likelihood that s.118(1), r.4 was intended to implement the law as understood by the Law Commission, the legislature did elect to use (on more than one occasion) the unusually strong—and, to some intents, untested—term, “preserve”, in respect of the retained common law rules. Further, as has been seen, r.4 can be read in a permissive manner, almost leaving open whether declarations of future intention do in fact fall within the res gestae exception to the rule against hearsay—which, in the uncertain state of the law existing at the date of the enactment, may have been an understandable posture for an informed legislature to have adopted. Is there a deeper moral to this story? It seems plausible that courts eventually will have interpretative fun and games with the declaration of intention. It is also possible, however, that a wider lesson can be drawn from use of the device the legislature has implanted in the Civil Evidence Acts, and now in the Criminal Justice Act, too. One is bound to ask: are common law and legislation actually incompatible to such an extent that the former cannot be successfully absorbed into the latter by the simple device of “preservation”? “Preservation” leaves a lot of moot questions. More generally, we may discover that the respective outlooks of these two forms of law-making are so radically different from one another that an unadorned concept of “preservation” makes no sense.61 Even in a system such as our own, where the common law’s exclusive bailiwick—not to mention its very method—shrinks and mutates day by day as the legislature extends its sphere of influence and as an oral system of law acquires habits ever more reminiscent of written systems,62 it hardly needs stating that common law and statute still operate in wildly divergent ways. The obvious property of statutory enactments is that they possess a fixity. No matter what, would-be interpreters of legislation find themselves anchored to a written matrix, inexorably compelled to engage every time with the settled text, with judicial precedents providing an additional sheet 61 In a nutshell, my great concern is not unlike that expressed in Lord Hope of Craighead’s thoughtful concurring opinion in the Pitcairn Island sexual abuse case, Christian v R. [2007] 2 W.L.R. 120 at [49]. There his Lordship wondered “whether the legislative technique that was adopted by the Governor was sufficiently well adapted to conditions on Pitcairn for it to be possible for us to say that the [Sexual Offences Act 1956] was in force there at the relevant time . . .” Is Parliament’s technique in s.118, r.4 adequate to fulfil its probable intention to make declarations of intention admissible under the res gestae exception? (I might add that the quality of Lord Hope’s opinion does not quell my unease at this newfangled practice of delivering concurring opinions in the Judicial Committee: see R. Munday, “Coming of Age in the Privy Council” (2006) 170 J.P. Jo. 244.) 62 See, notably, Lord Rodger of Earlsferry, “The Form and Language of Judicial Opinions” (2002) 118 L.Q.R. 226.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

anchor. Human rights law and a few rogue precedents apart, statutory law is pretty much nailed down. Common law is of another kidney. The steady, yet unpredictable accretion of judicial decisions, reminiscent of a random accrual of polyps on a coral reef, is capable of protracted growth, sometimes with no need for courts ever to return to a rule’s point of origin, imagining that it could be accurately pinpointed. Over time the law may variously appear to expand or contract, to decay, to conjure novel doctrines out of virtual thin air, to re-invent itself in fresh configurations, even to drift. The common law, in short, is anything but fixed. In some ways, it is the very antithesis of legislation. Well-known examples of large-scale common law activity abound: the sudden emergence of a tortious liability for negligent misrepresentation63 ; the tussle for territory, resembling a grinding of tectonic plates, currently being acted out between an imperialist tort of negligence and a beleaguered rule in Rylands v Fletcher 64 ; the “recognition” of new heads of contractual damages65 ; the emergence of broad restitutionary doctrines; and so on. In the law of evidence we have the doctrine of res gestae itself. In Bedingfield 66 the hearsay exception relating to the admissibility of excited utterances was given so narrow an ambit that Sir Alexander Cockburn C.J., who had delivered the ruling, was subjected to almost unparalleled criticism for his day.67 Just under a century later, Lord Wilberforce’s obiter views expressed in the Privy Council in Ratten v R.68 supplied a secure launching pad from which the Court of Appeal69 and, ultimately, for the House of Lords in Andrews 70 could transform the doctrine into something closely resembling the formulation presently enshrined in s.118, r.4(a) of the Criminal Justice Act 2003. The common law rules selected for “preservation” in the Civil Evidence Acts and the Criminal Justice Act 2003 are of two basic sorts. On the one hand, the majority of “preserved” rules—such as the rule that character may be proved by evidence of reputation71 —have been more or less stable for considerably in excess of a century. It may seem unlikely that the evidence of reputation rule would suddenly erupt into life from its nighcryogenic state—although one could have imagined that, had it not been for the legislation, in a bid to instil greater realism into the law affecting 63 Hedley Byrne & Co Ltd v Heller Partners Ltd [1964] A.C. 465. 64 For its latest manifestation, see Transco Plc v Stockport MBC [2004] 2 A.C. 1. 65 e.g. Ruxley Electronics & Construction Ltd v Forsyth [1996] A.C. 344 at 360, per Lord Mustill (“the

consumer surplus”). 66 (1878) 14 Cox C.C. 341. 67 See R. Munday, “The Judge who Answered his Critics” [1987] C.L.J. 303. 68 [1972] A.C. 378. 69 e.g. Turnbull (1985) 80 Cr. App. R. 104. 70 [1987] A.C. 281. 71 See Civil Evidence Act 1995 s.7(3)(a) and Criminal Justice Act 2003, s.118(1), r.2, preserving the ruling in Rowton (1865) Le. & Ca. 520.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 67 character evidence, an adventurous court might just have contemplated the rule’s abrogation. Since the legislature has “preserved” the rule, that avenue now looks to be barred. But following the entry into force of the legislation, may the courts undertake further “refinement” of the reputation rule? For instance, could they impose a corroboration requirement, making admissibility of evidence of reputation dependent upon confirmation from a second, independent source? Could the courts impose a stringent leave requirement for this highly questionable species of evidence, beyond anything presently contemplated in respect of evidence of non-defendants’ bad character under s.100 of the Criminal Evidence Act 2003? Could the courts restrict the delivery of such evidence to the testimony of members of designated religious orders, elected local representatives and suitably accredited community leaders? As I have tried to show, not all preserved common law rules are completely settled. Rules that have been subject to recent dispute and even significant recent extension, such as the res gestae exception to the hearsay rule, are also “preserved” in the 2003 Act. We have seen that their more volatile nature can present an additional puzzle. As well as having to determine whether these rules, too, may continue to develop as conventional common law entities within what is meant to be a rigid statutory framework, the legislature’s description of the rule may be at variance with the common law construct that it seeks to “preserve”, making it difficult—or even impossible—to ascertain exactly what has been “preserved”. As G.K. Chesterton often reminded us, “it is futile to talk of reform without reference to form”,72 for “reform implies form”.73 In the case of declarations of intention, our fault was that we never really knew the form before we set about the “reform”. Ought the legislature to be taken, therefore, not only to have “preserved” the genus, say, of res gestae in these provisions, but also to have kept that “preserved” common law organism alive in captivity? Is there an assumption that the res gestae exception will continue to grow organically, even though the doctrine has been absorbed into a would-be comprehensive statutory framework? Do the courts treat interpretation of the retained provisions as a conventional exercise in statutory construction or as the province still of traditional judge-made law? This is largely unknown territory. Legislative technique The draftsman employs a range of techniques when seeking to integrate common law survivals into a new legislative setting. Sometimes the statute 72 The Superstition of Divorce (1920), p.3. 73 Orthodoxy: The Romance of Faith (1904), preface.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

will simply supplement the common law. In this spirit, the Terrorism Act 2000 laid down: “A power conferred on a person by virtue of this Part— (a) is additional to powers which he has at common law or by virtue of any other enactment, . . .” (s.105) Sometimes the Act will specify particular features of the common law that are to be kept in the new law by specifying that they are “not affected” by the legislation. The Criminal Procedure and Investigations Act 1996 s.21 retains some common law elements in the new statutory rules governing disclosure, but discards others: “(1) Where this Part applies as regards things falling to be done after the relevant time in relation to an alleged offence, the rules of common law which— (a) were effective immediately before the appointed day, and (b) relate to the disclosure of material by the prosecutor, do not apply as regards things falling to be done after that time in relation to the alleged offence. (2) Subs.(1) does not affect the rules of common law as to whether disclosure is in the public interest.”74 The Public Order Act 1986 s.40(4), too, provides: “Nothing in this Act affects the common law powers in England and Wales to deal with or prevent a breach of the peace.” In similar vein, other statutes like the Contempt of Court Act 1981 s.6 will tell us that certain rules of common law are not “prejudiced” by the new enactment: “Nothing in the foregoing provisions of this Act— (a) prejudices any defence available at common law to a charge of contempt of court under the strict liability rule; . . .”75 The common law may actually be employed as a yardstick for determining the future scope of a statute’s provisions. Thus, in delineating the rights, duties, etc. of enforcement officers in relation to High Court writs of execution, the Courts Act 2003 Sch.7 para.4 lays down: “(1) This paragraph applies in relation to writs directed to one or more enforcement officers under paragraph 3. 74 See also Merchant Shipping Act 1995 s.60(10): “Nothing in the regulations or done in pursuance of the regulations shall be construed as affecting any power to institute, prosecute, entertain or determine proceedings (including criminal proceedings) under any other enactment or at common law.” 75 See also Criminal Justice and Public Order Act 1994 s.34(5).

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 69 (2) The relevant officer has, in relation to the writ, the duties, powers, rights, privileges and liabilities that a sheriff of a county would have had at common law if— (a) the writ had been directed to him, and . . .” Lately, when “codifying” the rules of common law and equity that hitherto have governed director’s duties, on more than one occasion the Companies Act 2006 has had to invoke those very common law rules and equitable principles. The legislature saw its task as setting forth the “(perhaps unstated) general principles” that were thought to have lain behind the declarations of principle expressed in individual judgments of the courts. As the Department of Trade and Industry explained in the Act’s Explanatory Notes: “In the company law field, the principles being applied will frequently be taken from other areas, in particular trusts and agency. It is important that these connections are not lost and that company law may continue to reflect developments elsewhere. Frequently the courts may formulate the same idea in different ways. In contrast legislation is formal. It is not easy to reconcile these two approaches but the draft sections seek to balance precision against the need for continued flexibility and development.” (para.305) To foster this approach—and, perhaps, also demonstrating keener awareness of the difficulties posed when attempts are made to marry up statute and common law than was shown in the case of s.118 of the Criminal Justice Act 2003—in treating of the civil consequences of various breaches of general duties owed by directors, the Companies Act 2006 adopts unusual phraseology. Section 178(1), for example, enacts that “the consequences of breach (or threatened breach) of ss.171 to 177 are the same as would apply if the corresponding common law rule or equitable principle applied.” For the same reason, s.170, the major provision that outlines the scope and nature of directors’ general duties, declares, inter alia: “(3) The general duties are based on certain common law rules and equitable principles as they apply in relation to directors and have effect in place of those rules and principles as regards the duties owed to a company by a director. (4) The general duties shall be interpreted and applied in the same way as common law rules or equitable principles, and regard shall be had to the corresponding common law rules and equitable principles in interpreting and applying the general duties. (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

(5) The general duties apply to shadow directors where, and to the extent that, the corresponding common law rules or equitable principles so apply.” It would serve little purpose to go on multiplying the examples. The point is made. Although Parliament from time to time assimilates existing elements of common law into new legislative frameworks, it does not normally use the term “preserve” or “preservation” in order to bring about this result. Parliament has other tried and tested techniques. Its choice of uncharacteristically inflexible language in s.118 of the Criminal Justice Act 2003 is singular. Given that one must begin by assuming that all the legislature’s words are meant to have significance, it would seem that comparison with the methods customarily employed when it is sought to insert portions of common law into a legislative initiative only strengthens the suggestion that when the legislature resorted to the concept of “preservation”, it had in mind something more stiff and uncompromising than was present in other enactments in which elements of the common law have been retained. The moral of the tale If this tale has a moral, it may be the following. The technique of “preservation” of common law rules pursued by Parliament—and, so far as I am aware, until now the practice of declared “preservation” (and “effective preservation”, too, whatever that may be) has been confined to evidence legislation—requires either that the retained rules be clearly settled or that the draftsman use less unyielding terms than “preserve” and “preservation” (which emphatically denote a static situation) and leave the courts more obvious room for prudential manoeuvre. It is not even as though the declaration of intention necessarily presents a case of unique difficulty. One can, for example, foresee trouble ahead in respect of s.118(1), r.7, which preserves “Any rule of law under which in criminal proceedings a statement made by a party to a common enterprise is admissible against another party to the enterprise as evidence of any matter stated.” This common law rule, too, is notoriously slippery; and appears to be still in course of evolution.76 ‘‘Preservation” looks to create a mongrel law, neither fish nor fowl. It may be that, when the time comes, our courts will simply bulldoze 76 Sir John Smith, “Proving Conspiracy” [1996] Crim. L.R. 386; Sir John Smith, “More on Proving Conspiracy” [1997] Crim. L.R. 333. See recently, Platten [2006] EWCA Crim 140; [2006] Crim. L.R. 920. Rose L.J., it will be recalled, alluded casually to this rule in Singh [2006] 1 W.L.R. 1564 at [15]: see fn.3, above.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008] Legislation that Would “Preserve” the Common Law 71 through this problem, adopting an outlook to statutory construction akin to omnia praesumuntur rite et solenniter esse acta. They may just conclude tout court that declarations of intention are admissible as an exception to the rule against hearsay. However, the very fact that a discussion such as the present is possible and that questions of this degree of opacity can beset us, thanks to a few stray words in a public Act of Parliament, strongly suggests that the legislative technique of “preservation” of rules of common law is flawed.77 Probably, the technique is either best avoided altogether or only employed when the relevant rule of common law can be exactly identified and all risk of its being further developed is safely passed. RODERICK MUNDAY.*

77 That said, it could be mentioned that at least one other legislature has found itself constitutionally entangled when it selected an unfortunate method of integrating the common law into an overarching, legislative structure. Although the impugned legislation operated in a quite different setting, it does show how fraught this enterprise can prove. In Western Australia v Commonwealth (1995) 183 C.L.R. 373, the High Court of Australia was required to consider, inter alia, the validity of s.12 of the Native Title Act 1993. This provision sought to incorporate the common law principles affecting native title, in their entirety, into a comprehensive statutory regime. It did so by providing: “Subject to this Act, the common law of Australia in respect of native title has, after 30 June 1993, the force of a law of the Commonwealth” (emphasis supplied). As Mason C.J., Brennan, Deane, Toohey, Gaudron and McHugh JJ. warned, “If s.12 be construed as an attempt to make the common law a law of the Commonwealth, the attempt encounters some constitutional obstacles” (at [146]). The principal obstacle derived from the nature of common law itself. As Brennan J. had explained almost a decade earlier in Giannarelli v Wraith (1988) 165 C.L.R. 543 at 584, “In the view of a court sitting at the present time, earlier decisions which are not binding upon it do not necessarily represent the common law of the earlier time, though they record the perception of the common law which was then current.” Thus, the judges reasoned, if a court, because it perceives the common law to be different from what it was earlier perceived to be, so declares it, then effect will be given to that declaration as truly representing the common law. The majority went on to assert, “the ‘common law’ must be understood either as a body of law created and defined by the courts or as a body of law which, having been declared by the courts at a particular time, may in truth be—and be subsequently declared to be—different. Whether the common law be understood by reference to its source in judicial reasons for decision or by reference to its content as developing from time to time, there are objections to its being treated as a law of the Commonwealth” (at [147]). If the “common law” is taken as the body of law created and defined by the courts, s.12 “attempts to confer legislative power upon the judicial branch of government. That attempt must fail either because the Parliament cannot exercise the powers of the courts or because the courts cannot exercise the powers of the Parliament” (at [148]). On the other hand, if s.12 is construed as importing the common law as an organic, developing but unwritten body of law, under the terms of the 1993 Act the content of the common law would, in the ordinary course of events, change from time to time according to the changing perceptions of the courts. Such changes would occur without reference to Parliament in which is reposed the power to make special laws for the people of a race as the Parliament deems necessary which would be unlawful under the Act, which required that laws deemed necessary for the people of a race must have been considered by the Parliament. The judges therefore concluded that to construe s.12 as an attempt to make the “common law” a law of the Commonwealth, was invalid either because it purported to confer legislative power on the courts or because the enactment of the common law relating to native title found no constitutional support in the statute. “A ‘law of the Commonwealth’, as that term is used in the Constitution, cannot be the unwritten law. It is necessarily statute law, for the only power to make Commonwealth law is vested in the Parliament” (at [151]). See further, Mr Justice Gummow, “Statutes: 2005 Sir Maurice Byers Lecture” (Summer 2005/2006) N.S.W. Bar Association Bar News 30. * Fellow of Peterhouse, Cambridge. Common law; Hearsay; Intention; Res gestae

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

EXPLAINING RESULTING TRUSTS THE most difficult question one can ask about resulting trusts is why they arise. The word “result” comes from the Latin resalire or resultare, meaning “to spring back”. When used to qualify the word “trust”, it tells us that the trust “springs back”, which is only to say that its beneficiary is the person who made the transfer of rights now forming the subject-matter of the trust. But it does not tell us why. There have been two recent attempts to explain the incidence of resulting trusts. Lord Browne-Wilkinson has said that they arise because of a presumption that the transferor intended them so to do.1 Birks and Chambers, by contrast, argue that they are the law’s response to “nonbeneficial transfers”.2 This latter thesis is particularly important because it is extrapolated from to generate resulting trusts in the generality of unjust enrichment cases.3 This article will argue that neither thesis is correct. Of the three trusts traditionally classified as resulting,4 two arise because of a legal presumption that a trust was declared by the transferor in his own favour. Like the express trust, they are species of consensual trusts. The third, however, arises where there is no proof, by evidence or presumption, of a declaration of trust in favour of its beneficiary. In that respect, it arises by operation of law. But exactly why the law should impose a trust in such circumstances has yet to be satisfactorily explained, though one thing this cannot be is as a response to a “non-beneficial transfer”. I. THE TRADITIONAL CATEGORIES OF RESULTING TRUSTS Because the word “resulting” looks to the identity of the beneficiary and not the reason why the trust arises, resulting trusts inevitably cut across the categories with which they are traditionally aligned, express, implied, and constructive trusts, which do ask that question. In order to keep them in check, judges and textbook-writers artificially limit their incidence to three situations: 1 Westdeutsche Landesbank Girozentrale v Islington LBC [1996] A.C. 669 at 708. 2 P. Birks, “Restitution and Resulting Trusts” in S. Goldstein (ed.), Equity and Contemporary Legal

Developments (1992), pp.335–373; “Trusts Raised to Reverse Unjust Enrichments: The Westdeutsche Case” [1996] R.L.R. 3; R. Chambers, Resulting Trusts (1997); “Resulting Trusts in Canada” (2000) 38 Alberta L.R. 378; “Resulting Trusts” in A. Burrows and Lord Rodger of Earlsferry, Mapping the Law: Essays in Memory of Peter Birks (2006), pp.247–264. 3 The only such claims not generating resulting trusts would be those involving transfers of rights under contracts later frustrated or terminated for breach: Birks, fn.2 above (1992), pp.346–359; Chambers, fn.2 above (1997), pp.143–170. 4 These are enumerated immediately below. For reasons of space, other cases, e.g. those involving surpluses and so-called Quistclose trusts, will not be discussed.

JANUARY 2008]

Explaining Resulting Trusts

i. where it is proved by evidence5 that rights were voluntarily transferred by A to B inter vivos and A is not the father (or other person standing in loco parentis) of B or the husband of B. Here, a presumption arises which, unless rebutted, causes B to hold the rights so conveyed on trust for A. We will call this a “voluntary conveyance resulting trust”; ii. where it is proved by evidence that A inter vivos paid C in whole or in part to convey rights to B, and A is not the father (or other person standing in loco parentis) of B or the husband of B. Here, a presumption arises which, unless rebutted, causes B to hold the rights so conveyed on trust for A, either in whole or in part.6 We will call this a “purchase-money resulting trust”; iii. where it is proved by evidence that A conveyed rights to B either inter vivos or post mortem, to hold on declared trusts which are void because they lack objects, offend the rule against perpetuities, or both. Here, B will hold the rights so conveyed on resulting trust for A. We will call this a “failed trust resulting trust”. There are at least four differences between the voluntary conveyance and purchase-money resulting trusts and the failed trust resulting trust. First, as we have seen, voluntary conveyance and purchase-money resulting trusts do not arise where the transfer or provision of purchase-money is made by a father to his child or a husband to his wife. In such cases, a “presumption of advancement” is said to arise. However, since no fact is there proved by presumption, it is more accurate to describe this only as a situation where the resulting trust presumption does not apply.7 Secondly, as we have also seen, neither voluntary conveyance nor purchase-money resulting trusts arise in the case of transfers by will.8 Thirdly, voluntary conveyance and purchase-money resulting trusts do not arise where the right transferred is an interest under a trust.9 And fourthly, no voluntary conveyance resulting trust arises where the interest conveyed is something different from that held by the transferor. Failed trust resulting trusts, by contrast, can arise in all four situations.

5 Wherever the phrase “proved by evidence” is used, it should be taken to include proof by admission. 6 A resulting trust also traditionally arose where A and B contributed unequally to the purchase price

and the title was conveyed to A and B as joint tenants, whereby A and B held as equitable tenants in common in proportion to their contributions (Lake v Gibson (1729) 1 Eq. Cas. Abr. 290). In Stack v Dowden [2007] UKHL 17, a majority of the House of Lords held that this rule no longer applied in the case of “matrimonial or quasi-matrimonial homes”. 7 “It is called a presumption of advancement but it is rather the absence of any reason for assuming that a trust arose”: Martin v Martin (1959) 110 C.L.R. 297 at 303 (Dixon C.J., McTiernan, Fullager, and Windeyer JJ.). 8 Rogers v Rogers (1733) Cas. T. Talbot 268. 9 Spence, The Equitable Jurisdiction of the Court of Chancery (1846), p.454.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

II. PRESUMPTIONS: USE AND ABUSE It is a matter of agreement on all sides that the first two resulting trusts arise because of the operation of a presumption. We need, therefore, to understand how presumptions work.10 This is especially important for, as we will see, the main protagonists in this debate demonstrate some fundamental misunderstandings in this regard. (a) “True” presumptions Presumptions properly-so-called form part of the law of proof. Generally speaking, facts can be proved by admission, judicial notice, or evidence. In the absence of admission and judicial notice, the general rule is that facts must be proved by evidence, the burden of proving those facts lying on the party alleging them to have occurred. Very occasionally, however, proof by evidence of one fact, the “basic” or “primary” fact, gives that party to the litigation the benefit of another fact, the “secondary” fact, without any need to adduce evidence in proof. In such cases, the fact is proved by presumption. The burden then lies on the other party to adduce evidence to rebut the presumption.11 If they do not, the tribunal of fact must find the secondary fact proved. Such presumptions, variously known as “legal presumptions” or “presumptions of law”, are described in Cross on Evidence as the only “true presumptions”.12 They usually arise because the existence of the secondary fact is the most likely inference to draw from proof by evidence of the basic fact.13 “Presumptions”, said Murphy J. in Calveley v Green, “arise from common experience . . . If common experience is that when one fact exists, another fact also exists, the law sensibly operates on the basis that if the first is proved, the second is presumed. It is a process of standardized inference.”14 A well-known legal presumption is that of legitimacy. Proof by presumption that a child is the legitimate issue of a husband and wife 10 The best account is P. Murphy, Murphy on Evidence, 10th edn (2007), pp.632–643. See also I.H. Dennis, The Law of Evidence, 3rd edn (2007), pp.508–513; H. Malek Q.C. (ed.), Phipson on Evidence, 16th edn (2005), paras 6-16–6-31; Sir John Smith, Criminal Evidence (1995), pp.47–53; C. Tapper, Cross and Tapper on Evidence, 11th edn (2007), pp.146–151. 11 The amount of evidence required in rebuttal depends on whether the presumption is “persuasive” or “evidential”. In the case of persuasive presumptions, the evidence in rebuttal must show on a balance of probabilities the untruth of the fact proved by presumption. In the case of evidential presumptions, the burden will be discharged if sufficient evidence is adduced to make the existence of the fact proved by presumption a live issue. 12 Cross and Tapper on Evidence, fn.10 above, p.144. 13 This is not the only reason. They sometimes arise because of the seriousness of the allegation (possibly the explanation of the presumption of legitimacy: Murphy, fn.10 above, p.635), or to resolve evidential impasses (e.g. the statutory presumption in s.184(1) of the Law of Property Act 1925 regarding the order of death of joint tenants in circumstances where it is uncertain which survived the other). 14 (1984) 155 C.L.R. 242 at 264 (Murphy J.). The contrast drawn between proof and presumption is unfortunate. It is more accurate to talk of proof by evidence and proof by presumption.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

occurs on proof by evidence that the child was born to the wife, that it was born during lawful wedlock or within the normal period of gestation after wedlock had ended, and that the husband was alive at the date of conception.15 It is then for the other side to adduce evidence in rebuttal showing that the husband could not be the father. However, though known as the “presumption” of legitimacy, it is vital to appreciate that legitimacy is not the secondary fact proved, for legitimacy is a legal conclusion from proved facts. When we say that a child is presumed legitimate, what we mean is that it has been proved by presumption, not evidence, that the husband impregnated the mother and that it was this impregnation which led to her giving birth. (b) “False” presumptions Unfortunately, the word “presumption” is a much misused term, and there are at least four senses in which it used to describe something different from the legal presumption outlined above. None are “true” presumptions, for they either involve no issue of proof, or, when they do, no question relating to proof of a secondary fact on proof by evidence of a primary fact. In this sense, they are all misnomers. (i) Presumptions indicating the location of the burden of proof The most common misuse occurs in statements indicating the location of the burden of proof in a civil or criminal trial. So, for example, there is said to be a “presumption” of innocence in favour of the defendant in a criminal trial.16 However, as Cross points out: “There is no basic fact at all, and the presumption does no more than express the incidence of the relevant burden.”17 In truth, therefore, a rule of law as to the incidence of the burden of proof is being restated in adjectival terms with the language of presumption adding nothing. (ii) Presumptions describing rules of substantive law A second misuse occurs when the word “presumption” is used to describe, not a rule of evidence, but of substantive law. So, for example, there is said to be a “presumption” that a child under the age of 10 years cannot commit a crime,18 and that a criminal conviction is conclusive evidence in certain civil proceedings that a crime has been committed.19 Such “presumptions” are often known as “irrebuttable presumptions of law”, but this immediately shows that they are not presumptions at all. 15 The Banbury Peerage Case (1811) 1 Sim. & S. 153; The Poulett Peerage Case [1903] A.C. 395. 16 R. v Keogh [2007] EWCA Crim 528; [2007] 1 W.L.R. 1500. 17 Cross and Tapper on Evidence, fn.10 above, p.144. 18 Children and Young Persons Act 1933 s.50. 19 Civil Evidence Act 1968 s.13(1).

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

Being incapable of dislocation by contrary evidence, they tell us nothing whatever about proof. (iii) Presumptions of fact A third misuse lies in the use of the word “presumption” to describe a rule providing that upon proof by evidence of one fact, a tribunal of fact may find a second fact proved in the absence of sufficient evidence to the contrary. Examples of such “presumptions” include the “presumption of intention”, that a person intends the natural consequences of his acts,20 and the “presumption of guilty knowledge”, that an accused proved by evidence to have been found in possession of goods recently stolen the acquisition of which he cannot explain is “presumed” to be guilty of handling stolen goods.21 These are variously called “presumptions of fact” or “provisional presumptions”. But though close to legal presumptions, they are different in that they are not concerned with the allocation of the burden of proof, the tribunal of fact not being compelled, as it is with legal presumptions, to find the fact in question proved.22 They have therefore been described as “no more than rational inferences to be drawn in the light of experience and common sense . . . The law does no more than recognize factual probability.”23 (iv) Presumptions as rules of construction A fourth misuse occurs in the context of the construction of words in statutes,24 wills, deeds, and other instruments. Thus, there are said to be “presumptions” that: a consolidating statute is not intended to effect a change in the law25 ; statutes are not intended to operate retrospectively26 ; the operation of statutes is confined to the United Kingdom27 ; the benefit of a covenant restrictive of the user of land is annexed to each and every part of the dominant tenement28 ; and a contracting party intends to deal with the person in front of him.29 Even the rules laid down in the Sale of Goods Act 1979 for the passing of property under a 20 R. v Steane [1947] K.B. 997; Criminal Justice Act 1967 s.8. 21 R. v Schama and Abramovitch (1914) 11 Cr. App. R. 45; R. v Garth [1949] 1 All E.R. 733. 22 For this reason, the “presumption of fact” is described by Phipson (fn.10 above) as a misnomer:

para.6-17. 23 P. Carter, Cases & Statutes on Evidence, 2nd edn (1990), p.77. 24 A whole chapter is devoted to the “presumptions” used in the construction of statutes in G. Dworkin, Odgers’ Construction of Deeds and Statutes, 5th edn (1967), pp.387–416. 25 H v H [1968] 3 All E.R. 560 at 566 (Sir Jocelyn Simon P). 26 Ward v British Oak Insurance Co Ltd [1932] 1 K.B. 392 at 397 (Scrutton L.J.). 27 Jefferys v Boosey (1854) 4 H.L.C. 815 at 970 (Lord Brougham). 28 This is how most textbooks present it: e.g. C. Harpum, Megarry & Wade: The Modern Law of Real Property, 6th edn (2000), p.1031, fn.1. However, in the case relied on, Federated Homes Ltd. v Mill Lodge Properties Ltd [1980] 1 W.L.R. 594, Brightman L.J. correctly spoke, not of a presumption, but of a rule of construction. 29 Ingram v Little [1961] 1 Q.B. 31 at 61 (Pearce L.J.).

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

contract of sale, despite themselves never using the word, are said by some textbook-writers to involve “presumptions”.30 None, however, are true presumptions, for in none is a question of proof of facts in issue. III. THE TRADITIONAL EXPLANATION OF “PRESUMED” RESULTING TRUSTS Because the voluntary conveyance and purchase-money resulting trusts are universally accepted as resting on a presumption, they are collectively, though somewhat unfortunately known as “presumed” resulting trusts. Two questions arise. Is the word “presumption” being used in its true sense? If so, what is the fact proved by presumption? (a) The type of presumption in “presumed” resulting trusts The presumption in play in the “presumed” resulting trust is undoubtedly a legal presumption. This is illustrated by cases where: (i) the presumption is dispositive of the result where no contrary evidence is adduced; (ii) evidence is admitted to rebut the presumption; and (iii) courts refuse to invoke the presumption where there is no gap in the evidence. (i) Presumption dispositive of the result In The Venture,31 Percy Stone proved by evidence that he contributed £550 towards the £1,050 purchase price of a title to a yacht which had been conveyed by the vendor to Percy’s brother, Andrew. When Andrew died intestate, his widow claimed possession of the yacht, alleging that Percy had only provided the £550 by way of loan. Percy alleged that the money was put up by way of partnership, and that title to the yacht had been held by his brother on trust for them both. Neither party adduced any evidence to substantiate these allegations. The registrar held that Percy had failed to make out his case and gave judgment for the widow. The Court of Appeal held this approach to be fundamentally misconceived. “On its being proved that Percy Stone had advanced a certain part of the purchase money, the presumption of law arose that he was beneficially entitled to a corresponding share in the yacht. It was for the plaintiff to displace that presumption by bringing evidence to the contrary; but she has entirely failed to bring any such evidence. The Court must therefore give effect to the presumption, and must hold that, as the defendant paid a part of the purchase money, he acquired an interest in the yacht . . ..”32 30 So, e.g. Bridge describes them as “presumptive rules of intention”: M. Bridge, The Sale of Goods (1997), p.62. 31 [1908] P. 218; (1907) 77 L.J.P. 105. 32 (1907) 77 L.J.P. 105 at 108 (Farwell L.J., giving the judgment of the court). The corresponding passage in [1908] P. 218 at 230, is essentially the same, though the phrase “presumption of law” is rendered only as “presumption”.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

The same view of the operation of the presumption is found in Mellish L.J.’s judgment in Fowkes v Pascoe,33 where the presumption is described as a “rule of law” which “must prevail” even though “the court might not believe that the fact was in accordance with the presumption”.34 (ii) Presumption rebutted by contrary evidence Decisions in which contrary evidence has been effective to rebut the presumption are legion. Thus, in Dyer v Dyer, Eyre C.B. said: “it is the established doctrine of a court of equity that this resulting trust may be rebutted by circumstances in evidence.”35 An example is Fowkes v Pascoe itself.36 The facts proved by evidence were that the testatrix had an only child, a son. He predeceased her, leaving a widow, who remained living with the testatrix and who remarried from her house. The widow’s son and daughter from her second marriage were treated by the testatrix as her own flesh and blood. During her lifetime, the testatrix bought a number of annuities in the joint names of herself and her surrogate grandson, by this time a young man, which annuities at the time of her death were valued at some £7,000. As the successor of joint tenants, the issue was whether the surrogate grandson held them outright or on resulting trust for the testatrix’s estate. At first instance, Sir George Jessel M.R. held that a presumed resulting trust arose in favour of the testatrix’s estate, which trust had not been rebutted.37 His decision was reversed on appeal, Mellish L.J. saying: “. . . if there is evidence to rebut the presumption, then . . . the Court must go into the actual facts. And if we are to go into the actual facts, and look at the circumstances of this investment, it appears to me utterly impossible . . . to come to any other conclusion than that the . . . investment was made for the purpose of gift and not for the purpose of trust. It was either for the purpose of trust or else for the purpose of gift; and therefore any evidence which shews that it was not for the purpose of trust is evidence to shew that it was for the purpose of gift.”38

33 (1875) L.R. 10 Ch. App. 343. The facts of this case are discussed immediately below. 34 (1875) L.R. 10 Ch. App. 343 at 353. See also Rathwell v Rathwell (1978) 83 D.L.R. (3d.) 289 at

303, where Dickson J. described the presumption as a “matter of law”. Similar comments are found in Stack v Dowden [2007] UKHL 17; [2007] 2 W.L.R. 831, where Lord Neuberger (at [123]) described the presumption as effecting a shift in the burden of proof and Lord Walker spoke of it in matrimonial cases as no longer operating as a “legal presumption” (at [31]). Though Baroness Hale described it as “not a rule of law” (at [60]), her Ladyship would only seem to be there saying that the presumption is capable of rebuttal. 35 (1788) 2 Cox. Eq. Cas. 92 at 93. 36 (1875) L.R. 10 Ch. App. 343. 37 (1875) L.R. 10 Ch. App. 343 at 345, n.1. 38 (1875) L.R. 10 Ch. App. 343 at 353.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

(iii) Presumption cannot be resorted to where there is no gap in the evidence An example of a case where the court refused resort to the presumption where the evidence was complete is Goodman v Gallant.39 A title to a house was vested in the claimant’s husband on trust for himself and the claimant as joint tenants. The marriage broke down and the husband left. The claimant and her lover, the defendant, successfully negotiated to jointly buy out the husband’s interest. A conveyance was executed by the husband, claimant, and defendant, reciting that the title was conveyed to the claimant and defendant “upon trust for themselves as joint tenants”. They later fell out, and the claimant sought a declaration that she had a three-quarter share under a purchase money resulting trust. Her claim failed. There was no room, said the court, to invoke a presumption where the words accompanying the transfer were proved by evidence.40 (b) What is the fact proved by presumption in a presumed resulting trust? At the outset, it must be noted that just as the fact proved by presumption in the case of legitimacy is not legitimacy, so too the fact here cannot be “resulting trust”, for “resulting trust” is similarly a legal response to proved facts. It is accepted on all sides that in times past, the fact proved by presumption was that the transferor declared a use, the forerunner of the modern trust, in his own favour. In the fifteenth, sixteenth, and seventeenth centuries, it was standard practice for those with fee simple estates in land to put them in use for themselves. The device of a “feoffment to uses” was also employed in purchases of such titles.41 To have a fee simple title in use for oneself had a number of advantages.42 First, it allowed a will of the title to be made. Secondly, it avoided reliefs and wardship. Thirdly, it allowed the title to be used for the payment of debts on death.43 Fourthly, it avoided both dower44 and curtesy.45 Fifthly, it enabled the creation of settlements of land of greater complexity than was possible at law.46 And sixthly, it avoided escheat 39 [1986] Fam. 106. 40 [1986] Fam. 106 at 110–111 (Slade L.J.). 41 Rather than the vendor enfeoffing the purchaser and the purchaser enfeoffing his trusted friends,

the purchaser simply instructed the vendor to enfeoff his feofees direct, who would then hold to the purchaser’s will. 42 See generally, A.W.B. Simpson, A History of the Land Law, 2nd edn (1986), pp.173–207. 43 Though the common law rule was that debts died with the debtor, there was a concern that the deceased would suffer eternal damnation unless they were repaid. 44 Chaplin v Chaplin (1733) 3 P. Wms. 229 at 233–234; F.W. Saunders, An Essay on the Nature and Laws of Uses and Trusts (1791), p.109. 45 Saunders, fn.44 above, pp.109–110. 46 Simpson, fn.42 above, p.175.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

and forfeiture for treason, the latter being especially problematic in the turbulent times of the late fifteenth century.47 In fact, so common did feoffments to uses become, that it was said that “few men be sole seised on their land”.48 But because it was not usual to include words of trust on the face of the conveyance, and because such transfers were gratuitous, it was difficult to distinguish them from outright transfers. The solution lay in the use in certain cases of a presumption of a declaration to uses. Thus, in Cook v Fountain, Lord Nottingham L.C. spoke of trusts where the declarations “appear either by direct and manifest proof, or violent and necessary presumption. These last are commonly called presumptive trusts; and that is, when the Court, upon consideration of all circumstances presumes there was a declaration, either by word or writing, though the plain and direct proof thereof be not extant.”49 The presumption of declaration of trust was not, however, universally applied. As the same judge later explained in Grey v Grey: “A feoffment to a stranger, without consideration, raised a use to the feoffor; but a feoffment to the son, without other consideration, raised no use by implication to the father, for the consideration of blood settled the use in the son, and made it an advancement.”50 It is important to appreciate that the fact proved by presumption was, as Lord Nottingham makes clear, a declaration to uses, not merely that the feoffor possessed at the time of the transfer an unexpressed intention to create a use. Were it otherwise, we would be dealing with a rule of substantive law, not procedure, for an unexpressed intention to create a trust when proved by evidence does not generate a trust. As Megarry J. said in Re Vandervell’s Trusts (No.2), “the mere existence of some unexpressed intention in the breast of the owner of the property does nothing: there must at least be some expression of that intention before it can effect any result.”51

47 “During the contests between the houses of York and Lancaster, as it was the constant practice to attaint the vanquished, almost all the lands in England were conveyed to Uses”: W. Cruise, An Essay on Uses (1795), p.30. 48 Christopher St Germain, Doctor and Student; or Dialogues between a Doctor of Divinity and a Student in the Laws of England (1528–1531), ii, c.22. 49 (1676) 3 Swanst. 585 at 591 (emphasis supplied). The word “violent” is here used in the sense of “strong” or “compelling”. In modern terminology we would call it a persuasive presumption: fn.11 above. The presumption nowadays is only evidential: Pettitt v Pettitt [1970] A.C. 777 at 814 (presumption “readily rebutted by comparatively slight evidence”: Lord Upjohn). 50 (1677) 2 Swanst. 594 at 598. 51 [1974] Ch. 269 at 294. Though in Westdeutsche Landesbank Girozentrale v Islington LBC [1996] A.C. 669 at 708, Lord Browne-Wilkinson spoke of a presumption of “intention”, the point was not there in issue.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

IV. HAS THE PRESUMPTION CHANGED? Though admitting that the fact proved by presumption in this early period was a declaration of trust, both Birks and Chambers argue that this is no longer the case. The fact proved by presumption today, they say, is a “non-beneficial transfer”.52 It is said that the presumption changed around 1660, when trusts replaced uses. The reason was the enactment of two statutes, the Wills Act 1540 and the Tenures Abolition Act 1660,53 the former allowing fee simple titleholders to leave their interests by will, the latter abolishing a number of feudal incidents. Their combined effect, says Chambers, was to remove the incentive for title-holders to create trusts for themselves. The argument, however, is weak, for the two statutes addressed only some of the reasons causing titleholders to create trusts for themselves. As we have seen, such trusts were also used to avoid dower and curtesy, to allow for the creation of more sophisticated settlements than was possible at law, and to make the testator’s title to land available to pay his debts. These purposes still subsisted after 1660.54 Moreover, there are a number of cases since 1660 in which the judges saw themselves operating the resulting trust no differently from the resulting use. Thus, in 1788, in Dyer v Dyer, we are told that the purchasemoney resulting trust operates “on a strict analogy” with the old law on uses.55 And in Fowkes v Pascoe,56 the Court of Appeal in 1875 still spoke in terms of a presumption of transferors declaring trusts for themselves. We saw this in the dictum of Mellish L.J. discussed above57 ; it is also found in the judgment of James L.J., who asked whether it was: “. . . possible to reconcile with mental sanity the theory that [the deceased] put £250 into the names of herself and her companion, and £250 into the names of herself and Defendant, as trustees upon trust for herself ? What trust—what object is there conceivable in doing this? If this case were tried before a jury, no Judge could withdraw the facts of the contemporaneous purchases and of their repetition from the consideration of a jury, and, in my opinion, no

52 See the passages set out below, text to fnn.77–79. 53 Chambers, fn.2 above (1997), p.20. 54 So, e.g. J.A. Strahan and G.H.B. Kendrick, A Digest of Equity (1905), p.177 state that the practice “of taking conveyances of freeholds in the names of other persons than the purchaser to bar dower” continued until the passing of the Dower Act 1833. See also Lord Nottingham’s list of “mischiefs” carried on by trusts in the place of uses: D.E.C. Yale (ed.), Lord Nottingham’s “Manual of Chancery Practice” and “Prolegomena of Chancery and Equity” (1965), p.48. 55 (1788) 2 Cox. Eq. Cas. 92 at 93. 56 (1875) L.R. 10 Ch. App. 343. 57 Text to fn.38.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

jury would or could be found who would hesitate to say that the thing was done by way of gift and not trust.”58 Nor did the content of the presumption change in the twentieth century. Thus, in Pettitt v Pettitt, Lord Diplock criticised the presumption of resulting trust as an inference of fact “which an earlier generation of judges drew as to the most likely intentions of earlier generations of spouses belonging to the propertied classes of a different social era.”59 If the presumption had really changed, however, why is it described as anachronistic? Further, in Rathwell v Rathwell, Dickson J. said: “Resulting trusts are as firmly grounded in the settlor’s intent as are express trusts, but with this difference—that the intent is inferred, or is presumed as a matter of law from the circumstances of the case.”60 Given that the only intention which counts is one which is expressed,61 for Dickson J. at least, the content of the presumption in 1978 was still a declaration of trust. Finally, there is the decision of the House of Lords in Westdeutsche Landesbank Girozentrale v Islington LBC.62 Westdeutsche transferred substantial sums of money to Islington pursuant to contracts which in subsequent litigation were held to be ultra vires the local authority.63 It sought to recover the money and compound interest thereon. Its claim to interest turned on whether the money was held by the authority as trustee at the point of receipt. Adopting Birks’ argument, it argued that because the payments were made pursuant to a void contract, they had been paid gratuitously, and thus gave rise to a presumed resulting trust. Since that trust arose because of a “presumption of non-beneficial transfer”, and because there was no evidence that the money had been paid by way of gift, the resulting trust was unrebutted. The argument was rejected in limine.64 The presumption of resulting trust, said Lord Browne-Wilkinson, 58 (1875) L.R. 10 Ch. App. 343 at 348–349 (emphasis supplied). See also Standing v Bowring (1886) 31 Ch.D. 282, where the question was whether the presumption of resulting trust had been rebutted on the evidence. Lindley L.J., at 289, framed the inquiry thus: “The Plaintiff . . . rested her case on equitable grounds, and sought to establish a trust in her favour. No trust will suffice short of an absolute trust for herself. But it is impossible to impose such a trust on the Defendant, when the evidence conclusively shews that she never intended to create any trust of the kind ” (emphasis supplied). 59 [1970] A.C. 777 at 824. While it is true that his Lordship immediately before spoke of presumptions of “fact”, the point was not in issue. And the mistake is easily made, for though we are dealing with “inferences” of fact, the “inference” arises from a presumption of law. The same error is present in the statement of Chambers that “The presumptions of resulting trust and advancement are presumptions of fact, that fact being the intention of the person who has provided property to another”: Chambers, fn.2 above, (1997), p.11. 60 (1978) 83 D.L.R. (3d.) 289 at 303. 61 Text to fn.51. 62 [1996] A.C. 669. 63 Hazell v Hammersmith & Fulham LBC [1992] 2 A.C. 1. 64 Chambers seeks to downplay the significance of Westdeutsche, saying: “. . . the only issue on appeal was the claimant’s right to interest. With no examination of the underlying unjust enrichment, the case

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

“is rebutted by evidence of any intention inconsistent with such a trust, not only by evidence of an intention to make a gift.”65 It must, however, be admitted that there are dicta which appears to support the contrary view. In Nelson v Nelson,66 a mother provided the purchase price for a freehold title to land, the conveyance being made to her children so as to enable her to make fraudulent claims to state benefits. The title was later sold, and the mother sought a declaration that the children held the proceeds of sale for her on trust. In the High Court of Australia, Deane and Gummow JJ. spoke of the evidence showing that she “had no intention to confer any beneficial interest” over the title or the proceeds of sale on her children.67 But too much should not be made of this, for the precise nature of the fact proved by presumption was not in issue. The only question was whether the “presumption” of advancement applied in the case of a gratuitous transfer from mother to child, and, if it did, whether the mother’s illegal conduct precluded her from adducing evidence in rebuttal. Also said to support a change in the content of the presumption is an obiter dictum of Lord Millett in the Privy Council decision in Air Jamaica Ltd v Charlton,68 which concerned the distribution of the surplus of a pension trust which failed as offending the rule against perpetuities. The Privy Council held that the surplus was prima facie held on resulting trust for the contributors, the employer and the employees. Clause 4 of the trust deed, however, provided that: “No moneys which at any time have been contributed by the [employer] under the terms hereof shall in any circumstances be repayable to the [employer].” Did this clause prevent a resulting trust arising? As we will see,69 the House of Lords in Vandervell v IRC 70 held that the failed trust resulting trust arises, not because of the operation of a presumption, but by operation of law. Proof by evidence of a contrary intention was therefore irrelevant. Applying this reasoning, Lord Millett described the argument that no resulting trust could arise as

reveals little about why proprietary restitution was unavailable”: Chambers, fn.2 above (2006), p.257. What cannot be denied, however, is that the House did address and reject the initial premise of Professor Birks’ thesis. It mattered not, therefore, that the precise ground of recovery was not identified, for it could have made no difference to the final result. 65 [1996] A.C. 669 at 708. It must be admitted that Lord Browne-Wilkinson had immediately before described the presumed resulting trust as involving a “presumption that [the transferor] did not intend to make a gift to [the transferee]”. But this must be a slip, for it is precisely the argument of Birks which his Lordship immediately goes on to reject. Moreover, as we will see below (text to fnn.100–114), an intention not to make a gift, since it describes a legal conclusion from proved facts, cannot form the subject-matter of a presumption. Unfortunately, the statement probably misled Baroness Hale (at [60]) and Lord Neuberger (at [114]) in Stack v Dowden [2007] UKHL 17; [2007] 2 W.L.R. 831. 66 (1995) 184 C.L.R. 538. 67 (1995) 184 C.L.R. 538 at 549. 68 [1999] 1 W.L.R. 1399. 69 Text to fnn.130–139. 70 [1967] 2 A.C. 291.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

“. . . wrong in principle. Like a constructive trust, a resulting trust arises by operation of law, though unlike a constructive trust, it gives effect to intention. But it arises whether or not the transferor intended to retain a beneficial interest—he almost always does not—since it responds to the absence of any intention on his part to pass a beneficial interest to the recipient. It may arise even where the transferor positively wished to part with the beneficial interest, as in Vandervell v Inland Revenue Commissioners . . ..”71 It is said that this statement now represents the “prevailing view” of resulting trusts.72 This, however, is doubtful. Air Jamaica v Charlton provides nothing more than an unremarkable application of a decision of the House of Lords; it did not attempt to say anything new about resulting trusts. And in any case, since only a failed trust resulting trust was in issue, it says nothing of the content of the presumption operating in the other types of resulting trust. This is not to say that the presumption of a declaration of trust should continue. Presumptions represent a “consensus of judicial opinion disclosed by reported cases as to the most likely inference of fact to be drawn in the absence of evidence to the contrary.”73 This article does not seek to defend the continuation in the modern day of a presumption of declaration of trust where the common experience is that citizens do not generally create trusts for themselves. The point is only that no change in the content of the presumption has yet occurred in the view of the courts. We should finally note the argument of Chambers that a presumption of declaration of trust is inconsistent with the modern law on certainty of intention: “Why should the presumed intention to create a trust be effective to do so? It would contradict the well established requirement of certainty of intention to create a trust. As du Parcq L.J. stated in Re Schebsman, ‘unless an intention to create a trust is clearly to be collected from the language used and the circumstances of the case, I think the court ought not to be astute to discover indications of such an intention.’ ”74 This objection, however, confuses questions of substance and procedure. What du Parcq L.J. was concerned with in Re Schebsman 75 was the construction of words proved by evidence to decide whether they amounted to a declaration of trust. Exercises of construction, as we have 71 [1999] 1 W.L.R. 1399 at 1412. 72 Chambers, fn.2 above (2006), p.248. 73 Pettitt v Pettitt [1970] A.C. 777 at 823 (Lord Diplock). 74 Chambers, fn.2 above (1997), p.34. 75 [1944] Ch. 83.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

seen, are not presumptions at all.76 Indeed, if the same judge had been asked to construe the meaning of the words proved by presumption in a presumed resulting trust, scil., “I give you these rights to hold for me on trust”, he could not but have found that they amounted to a declaration of trust. V. A PRESUMPTION OF “NON-BENEFICIAL” TRANSFER? History apart, how do Birks and Chambers arrive at a presumption of “non-beneficial transfer”? Though it forms the crux of their argument, their reasoning on this point is brief. In talking of the voluntary conveyance resulting trust, Birks says: “. . . when it is thinking in terms of resulting trusts, equity approaches gifts with suspicion. Outside relationships, few and narrow, within which gift-giving is to be expected—relationships, that is, which generate the presumption of advancement—it presumes that the ‘gift’ is not one. In short the apparent donor gets his gift back unless, when the matter is double-checked, his intention to give is evident . . . He gets his property back, it jumps back to him under a resulting trust, unless there is affirmative proof that he intended it not to . . . Equity . . . starts from the assumption that an apparent gift—that is, a transfer without consideration—must result back to the apparent donor, and then, before it will allow the transferee to retain the benefit of the transfer, it requires positive evidence outside the fact of the transfer without consideration showing that the donor did indeed intend to donate.”77 Chambers speaks in similar terms: “Equity tends to be suspicious of gifts and often asks the recipient of an apparent gift to prove that it was intended as a gift. The failure to do so means that it will be held in trust for the apparent donor. In other words, the apparent gift creates a presumption of resulting trust.”78 And later: “The essential fact of intention, common to all cases of resulting trust, is the provider’s lack of intention to benefit the recipient.”79 Five separate propositions seem to be involved here: that gratuitous transfers are “apparent gifts”; that equity is suspicious of gifts; that 76 Text to fnn.24–30. 77 Birks, fn.2 above (1989), pp.343–344. 78 Chambers. fn.2 above (1997), p.11. 79 Chambers, fn.2 above (1997), p.21.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

equity consequently presumes that “apparent gifts” are “not gifts”; that the presumption of “not-gift” is a presumption of “non-beneficial transfer”; and that a “presumption” of “non-beneficial” transfer triggers a trust in favour of the transferor. Each proposition, however, as will be demonstrated below, is false. (a) Gratuitous transfers outside the range of advancement are “apparent gifts” If we are to see the transfers which trigger the first two types of resulting trust as “apparent gifts”, then we need to know the sense in which the word “apparent” is used. Unfortunately, this is nowhere spelt out. However, one thing we can say is that it is not used as a synonym for “presumed”, for we would then be talking about presumed gifts, not presumed trusts. Perhaps what is meant is its ordinary dictionary meaning, “seeming”, “appearing to the senses or mind, as distinct from what really is; not real”.80 But this is not its meaning either, for the gratuitous transfers with which we are concerned are not “seeming gifts”, merely ambiguous transfers. In the cases where the presumption applies, the facts proved by evidence are equivocal, for though the transfer might have been outright, it could also have been on trust. The evidential gap is closed, at least in cases outside the range of advancement, by the court finding as a matter of presumption that the transferor declared a trust in his own favour. Nor can “apparent gift” mean a transfer having one of the attributes of a gift, viz. that no value was given in exchange for the transfer, as they could just as well be called “apparent trusts”, for these too involve gratuitous transfers. It is therefore not correct to describe gratuitous transfers outside the range of advancement as “apparent gifts”. (b) Equity is suspicious of gifts Because in the majority of cases what is proved by evidence is a gratuitous transfer to a person outside the relationships of advancement, which proof triggers the “presumption” of resulting trust, not the “presumption” of advancement, it is easy to see how the “presumed” resulting trust might be thought to reflect a presumption by equity “against generosity”.81 But even if it was correct to see it in this way, it is not the same thing as saying that equity is “suspicious” of gifts. If it is really true that equity entertains a suspicion of gifts, how then do we explain the many situations in which this “suspicion” does not arise? Why, for example, are “apparent gifts” of life estates, entails, or interests 80 J.A Simpson and E.C.S. Weiner, The Oxford English Dictionary, 2nd edn (1989), Vol.I, p.563. 81 J. Hackney, Understanding Equity and Trusts (1987), p.149; Stack v Dowden [2007] UKHL 17 at

[60] (Baroness Hale).

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

under trusts not “suspicious”? Why are post mortem “apparent gifts” not suspect? And most importantly, why is the “suspicion” not present when words of gift are proved by evidence to have accompanied the transfer? These difficulties disappear if we think instead in terms of presumptions. In the first two cases, there is potentially a declaration of trust. But because in neither does it make sense for the transferor to have created a trust for himself, no presumption of declaration arises. And in cases where it is proved by evidence that an outright transfer was made, the presumption is not triggered for the simple reason that no ambiguity remains. It is therefore wrong to couch the presumption in terms of a “suspicion”, for in so doing, we leave the realm of presumptions altogether. But even if it was right to talk of a “suspicion” of gifts, a further difficulty arises. Exactly why should equity be suspicious of “apparent” gifts? Unfortunately, no authority is provided for this assertion. Indeed, it seems only to have been arrived at by working backwards from the existence of the “presumed” resulting trust itself. Thus, we are told: “The true nature of the presumption of resulting trust is revealed by the evidence which confirms or rebuts the presumption. The courts ask whether the apparent donor intended to make a gift and not whether he or she intended to create a trust.”82 However, as we saw from Fowkes v Pascoe,83 this would not seem to be the case. Moreover, even if it was, it does not follow that because proof by evidence of a gift rebuts the presumption, the presumption must be one of “not-gift”. The question at issue in these cases is whether the transferee took the rights outright or on trust, and proof by evidence of an intention to make a gift simply shows that it was not a transfer on trust. The idea that we are dealing with a “suspicion” of gifts has led Chambers to state that courts of equity use the “presumption” of resulting trust as a “safeguard against the unintended loss of assets”.84 However, this provides no clue as to the origins of the “suspicion”, for if it really is unintended losses which concern the courts, why is the concern not present when the transfer is between father and child or husband and wife? Moreover, it cannot be completely unintended losses which are its concern, for there is in every case in which a resulting trust arises an effective conveyance of rights. In reply, it might be said that by “unintended” is meant the sort of vitiation of intent found in cases of unjust enrichment. This is clearly the view of Birks, who wrote: “It is as though equity cynically supposes that a transferor who receives 82 Chambers, fn.2 above (2000), p.394. See also Birks, fn.2 above (1996), p.18. 83 See the passage from James L.J.’s judgment cited above, text to fn.58, and that of Lindley L.J. in

Standing v Bowring (1886) 31 Ch.D. 282 in fn.58. 84 Chambers, fn.2 above (2000), p.386.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

no value for his transfer must necessarily be labouring under a mistake or some other defect of judgment.”85 Unfortunately, why equity should jump to this conclusion is not then explained. Indeed, even if it did, we still need an explanation why transfers proved by evidence to contain words of gift are not viewed in this way. Moreover, it is hardly the most likely inference to draw from the facts proved by evidence, for common experience shows that most gratuitous transfers are not mistaken or otherwise defective. Once again, these difficulties disappear if we think in terms of presumptions, not suspicions. Furthermore, a search of the equity textbooks for anything remotely resembling a “presumption” that transferors labour under defects of judgment when making voluntary dispositions turns up nothing. The nearest we get is the equitable doctrine of undue influence, where, on proof by evidence of certain facts, a presumption arises that a gift was procured by undue influence. That it is this which is referred to is evident from the description of one of the situations attracting the presumption, a large gift from a parent to a child, as a “situation of which equity ought to be suspicious”.86 The two doctrines, however, are distinct. First, the presumption of undue influence, at least outside certain fixed categories of relationship,87 only arises where there is proof by evidence that the donor reposed trust and confidence in the donee.88 Secondly, the presumption is not confined to gifts: a conveyance for consideration can still attract its operation.89 Thirdly, the presumption applies to the grant of lesser interests, which is never the case with “presumed” resulting trusts.90 Fourthly, it is not restricted to transfers outside the relationships of advancement.91 And fifthly, and most tellingly, proof by evidence of words of gift does not prevent the presumption arising. Indeed, many of the cases involve transfers where it is admitted that the transfer was made as a gift.92 How can a court “presume” that a transfer was not a gift when it is admitted on all sides that it was? It cannot be right, therefore, to equate the presumption of resulting trust with the presumption of undue influence, or, more generally, with a presumption of vitiated consent to the transfer.

85 P. Birks, An Introduction to the Law of Restitution, rev. edn (1989), p.64. 86 Chambers, fn.2 above (2000), p.386. 87 e.g. parent/child (Turner v Collins (1871) 7 Ch. App. 329); solicitor/client (Wright v Carter [1903] 1 Ch. 27); spiritual leader/follower (Allcard v Skinner (1887) 36 Ch.D. 145). 88 Described by Lord Browne-Wilkinson in Barclays Bank Plc v O’Brien [1994] 1 A.C. 180 at 189–190, as “Class 2B undue influence”. 89 e.g. Huguenin v Baseley (1807) 14 Ves. 723; Tufton v Spurni [1952] 2 T.L.R. 516. 90 Goldsworthy v Brickell [1987] Ch. 378. 91 Simpson v Simpson [1992] 1 F.L.R. (transfer from husband to wife). 92 As, e.g. in Allcard v Skinner (1887) 36 Ch.D. 145.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

(c) Equity presumes that apparent gifts are “not gifts” Even if it was correct to say that a proved by evidence gratuitous transfer to a person outside the relationships of advancement is an “apparent gift”, and further, that equity is “suspicious of gifts”, it cannot be the case that courts presume that the transfer was “not a gift”, for this is to confuse inferences of fact with inferences of law. In this respect, a “presumption” that the transferor “did not make a gift” is no different to a “presumption” of “legitimacy” or “resulting trust”, for whether a transfer will be recognised as effecting a gift is a question of the application of legal rules to proved facts.93 We still need to know the fact, proved by evidence or presumption, which makes the transfer “not a gift”. The same confusion, between inferences of fact and law, can be seen in the fact that no case relied on by Chambers in support of his argument as to the nature of the fact presumed involves a true presumption.94 All are decisions in which the totality of the facts was proved by evidence; the only issue was the legal inference to be drawn from those facts. Take, for example, Dullow v Dullow, where it was proved by evidence that a mother paid for a title to land to be conveyed to her two sons. Hope J.A. said that despite the fact that she was “lacking in any familiarity with the elementary concepts of property ownership” and “did not have any clear view of what legal results flowed from her intentions”, “[her] intention was to reserve a beneficial interest in the properties for herself during her lifetime”.95 This intention entitled her to a “resulting trust of a life estate”96 : “Cases of this sort support the view that the presumption of resulting trust is of an intention to create a trust. However, the other view can also explain every one of these cases, because the provider’s intention to keep any portion of the beneficial interest necessarily means that he or she does not intend to pass that interest to the recipient.”97 But even if this “other view” was right, it is a legal conclusion from proved facts, not proof of a fact itself. Indeed, the result in the case could not have been reached through the application of the presumption, for proof by evidence that the mother provided the whole purchase price would have given her a trust of the fee simple, not merely an interest for life. 93 A useful summary of the law of inter vivos gifts can be found in the entry entitled “Gifts” by P. Pettit in Lord Mackay of Clashfern (ed.), Halsbury’s Laws of England (2004), Vol.20(1), pp.1–55. See also E.L.G. Tyler and N.E. Palmer, Crossley Vaines’ Personal Property, 5th edn (1973), pp.299–321; R. Chambers, “Conditional Gifts”, in N.E. Palmer and E. McKendrick, Interests in Goods, 2nd edn (1998), pp.429–60; J. Hill, “The Role of the Donee’s Consent in the Law of Gift” (2001) 117 L.Q.R. 127. The rules on post mortem gifts can be found in the standard works on succession. 94 The argument is made in Ch.1 of his book: fn.2 above, (1997). 95 (1985) 3 N.S.W.L.R. 531 at 541. 96 Chambers, fn.2 above (1997), p.21. 97 Chambers, fn.2 above (1997), p.21.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

The same confusion is evident in the next class of case relied on, that concerned with “ignorance”, where the claimant’s money is, without his consent, used by the defendant to purchase rights for himself. In the seven cases cited,98 the court found a resulting trust in the claimant’s favour. In none, it is said, could an intention to create a trust be inferred. The ignorance of those whose money was used is said to show that the “only possible inference, capable of giving rise to a resulting trust, is that they did not intend to benefit the recipients of the property.”99 But even if it could be said of the respective claimants that they had “no intention to benefit” the defendants, it is once again a legal inference from facts proved by evidence, not the proof of an additional fact through the operation of a presumption. (d) The “presumption” of “not-gift” is a presumption of a “non-beneficial transfer” A further difficulty concerns the description of the subject-matter of the “presumption” of “not-gift” as involving a “non-beneficial transfer”. What precisely is a “non-beneficial transfer” is nowhere explained. What we do know, however, is that the word “beneficial”, when spoken of rights, is commonly used in two different senses. The first is as an adverb, to describe how a right is held. The second is as an adjective, to describe a particular type of right. Only the first usage is correct. Rights can be held in one of three ways: outright; on trust; or as security for the performance of an obligation. To say that a person holds a right beneficially is only to say that he does not hold it on trust or as security. So, for example, in Vandervell v IRC,100 Lord Upjohn said that it was a question of intention101 whether the taxpayer and the trustee company “intended that the option should be held by the trustee company beneficially or as a trustee . . .”102 But when it is said of a person with an interest over, say, land that they have a “beneficial interest” in that land, this can only mean that they hold in respect of that land a right of a type known as “beneficial”. However, though there are legal and equitable rights over land, there is no third category of “beneficial” rights. Similarly, though I may hold a title to a chattel outright (beneficially), 98 Ryall v Ryall (1739) 1 Atk. 59; Lane v Dighton (1762) Amb. 409; Williams v Williams (1863) 32 Beav. 370; Merchants Express Co v Morton (1868) 15 Gr. 274; Sharp v McNeil (1981) 36 O.R. (2d.) 473; El Ajou v Dollar Land Holdings Plc [1993] 3 All E.R. 717. 99 Chambers, fn.2 above (1997), p.21. 100 [1967] 2 A.C. 291. 101 The “question of intention” to which his Lordship refers is discussed below, text to fnn.120–122, 133–137. 102 [1967] 2 A.C. 291 at 315. There was no question in the case of the particular right being held as security.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

on trust, or as security, I have no right to that chattel which can be described as “beneficial”. And though I may hold the benefit of a chose in action outright (beneficially), on trust, or as security, there is again no “beneficial interest” I have against the debtor separate from my right to sue. “Beneficial”, therefore, cannot be contrasted with legal or equitable, because those terms describe the two different jurisdictions which make up the common law; there is no third jurisdiction called “beneficial”. Unfortunately, it is this second sense which seems to have been adopted. Thus, Birks tells us that: “there is a resulting trust whether the transfer is made upon an initially valid trust or not, whenever it is found that the transferor did not intend that, in the events which have happened, the transferee should hold beneficially and, further, did not indicate any intention as to any other person or persons in whom the beneficial interest should be located.”103 Thus, though he starts by speaking of a transferee holding “beneficially”, Birks immediately slips into talk of a “beneficial” interest. Examples of the second usage can also be found in Chambers’ work. Thus, he describes the content of the presumption as being “that the provider did not intend to give the benefit of that property to the recipient”,104 and that the cases cited in support of his argument for the content of the presumption are ones concerning resulting trusts “in which the provider did not intend to benefit the recipient and yet could not have intended to retain the beneficial interest . . .”105 Perhaps “beneficial” is here being used as a synonym for “equitable”? A person entitled under a trust is known as a beneficiary, and is sometimes described as having a “beneficial” interest. So, and again in Vandervell v IRC,106 Lord Upjohn spoke of a body to which shares had not yet been transferred as having “no legal or beneficial interest in the shares”.107 But this is not how Chambers uses the word, for he is at pains to stress that an equitable interest under a trust is not the same thing as a beneficial interest, and that the phrase “beneficial interest” can cover both legal and equitable interests.108

103 Birks, fn.2 above (1992), p.352. 104 Chambers, fn.2 above (1997), p.19. 105 Chambers, fn.2 above (1997), p.21. See also Chambers (1997) at p.25: “Her intention to create a trust was significant because it proved the lack of intention to transfer the beneficial interest.” 106 [1967] 2 A.C. 291. 107 [1967] 2 A.C. 291 at 313. 108 Chambers, fn.2 above (1997), p.55.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

Might it then be that a “non-beneficial” transfer is nothing more than a transfer not intended as a gift?109 The difficulty is that this would lead to resulting trusts arising in situations in which no one would say they should. We know, for example, that no resulting trust arises in the case of a transfer proved by evidence to have been made for valuable consideration. Yet when a vendor conveys rights pursuant to a contract of sale, it is clear that he has no intention of making a gift to his purchaser. In the same way, though a citizen paying their validly demanded tax bill can hardly be said to be making a gift, no resulting trust arises. “Benefit” cannot therefore be equated with “gift”, and “not-benefit” with “not-gift”. A different meaning seems to emerge from Chambers’ discussion of Dullow v Dullow,110 where, as we saw,111 an intention by the mother to create a trust in her own favour was said to be equally consistent with the view that she did not intend to benefit her sons. This might be seen to show that the word “benefit” is being used as a synonym for “hold outright”. Such usage would cure the problem of resulting trusts in cases of contracts of sale or payments of taxes, for though such transferors do not intend to make gifts, they do intend the recipients to take outright (beneficially). This usage is also implicit in the discussion of the failed trust resulting trust. In such cases, he says, the fact proved by evidence, that the transferee was to take the rights as trustee, shows that he was not meant to take beneficially.112 An intention not to benefit a person is thus equated with an intention to vest in him rights not to be held outright. However, we immediately run into a further problem, for we know that both Birks and Chambers consider cases of mistaken payments to involve non-beneficial transfers. Indeed, their whole thesis rests on a perceived correspondence between the traditional categories of resulting trusts and the generality of cases of unjust enrichment. It is, however, difficult to see how the typical unjust enrichment scenario can be said to involve a “non-beneficial transfer”, where “non-beneficial” means “not outright”. Take the well-known case of Kelly v Solari.113 An insurance company claimed to have paid out £197 10s on a life insurance policy to the widow of the assured having forgotten that the policy had lapsed. Assuming it was indeed mistaken, it was held that it could recover the sum from the widow in a common law action for money had and received. But it can hardly be said that the company did not intend the right to be held by the widow outright. It thought it owed her the money and its intention 109 “All resulting trusts arise in response to the same type of event: the receipt of an asset by someone who was not intended to be its beneficial owner. In other words, there was a lack of intention to make a gift’: Chambers, fn.2 above (2000), p.387. 110 (1985) 3 N.S.W.L.R. 531. 111 Text to fn.97 above. 112 Chambers, fn.2 above (1997), p.43. 113 (1841) 9 M. & W. 54.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

in transferring the right to her was to discharge its perceived obligation. One does not discharge a debt by transferring money to ones creditor to hold “not outright”. Thus, the plaintiff’s intention can have only been to constitute the widow an outright holder of the right, and if “beneficial” is to be equated with “outright”, then there clearly was an intention to “benefit” her.114 Of course, the company was acting under a mistake, but it was a mistake as to its motive for the transfer, not the capacity in which the rights were to be held. But even if the law did recognise a notion of “non-beneficial transfer” distinct from a transfer on trust or as security, it is difficult to see how the “fact” of such transfer can form the subject-matter of the presumption triggered by proof by evidence of a gratuitous transfer. We have seen how presumptions “arise from common experience”.115 However, the most likely inference to draw from a gratuitous transfer is hardly that it was “non-beneficial”, that the transferee was not meant to take the transferor’s “beneficial” interest. Assuming again that such an interest can exist, and, further, that a “non-beneficial” transfer can form the subject-matter of a presumption, this “fact” would seem no more likely than a declaration of trust for the transferor. Indeed, it seems to contradict the facts proved by evidence, for if it really was the transferor’s intention “not to benefit” the transferee, why did he make the transfer at all? (e) The presumption of non-beneficial transfer triggers a trust in favour of the transferor The final claim is that the law responds to proof by presumption of the “fact” of “non-beneficial” transfer by raising a trust in favour of the transferor. Of course, the trust response is irresistible if the fact proved by presumption is a declaration of trust, for we know that the law responds to proof by evidence of that fact with a trust, and proof by presumption can yield no different result. But why is the raising of a trust a response to proof by presumption of a “non-beneficial” transfer? The reason is to be found in Chambers’ discussion of the failed trust resulting trust, of which he says: “Because the beneficial interest has not been completely disposed of and the trustees were not intended to enjoy the benefit of the legal ownership for themselves, the resulting trust arises.”116 114 cf. The Rt Hon. Sir Peter Millett, “Restitution and Constructive Trusts” (1998) 114 L.Q.R. 399 at

412. 115 Calveley v Green (1984) 155 C.L.R. 242 at 264 (Murphy J.). Though it is acknowledged that this is not the only reason why legal presumptions arise (fn.13 above), no other reason has been put forward here. 116 Chambers, fn.2 above (1997), p.54.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

But this explanation only works if there is such a thing as a “beneficial interest” separate from the legal interest, and, as we have seen, this is not the case. If there is no such interest, its disposal can never be described as “incomplete”. VI. LORD BROWNE-WILKINSON’S EXPLANATION OF FAILED TRUST RESULTING TRUSTS The rule that a transfer accompanied by a declaration of trust which fails to declare effective trusts generates a resulting trust in favour of the transferor is sanctioned by many years of authority.117 Typical is Morice v Bishop of Durham, where the testatrix left her residuary estate to the Bishop of Durham “on trust for such objects of benevolence and liberality as the Bishop in his own discretion should most approve of.” The court held that the objects were uncertain and the intended trust void, with the consequence that the rights were held by the bishop on trust for the testatrix’s estate.118 In an obiter dictum in Westdeutsche Landesbank Girozentrale v Islington LBC, Lord Browne-Wilkinson said that this trust was no different from the other resulting trusts and arose because of the operation of a presumption: “Both types of resulting trust are traditionally regarded as examples of trusts giving effect to the common intention of the parties. A resulting trust is not imposed by law against the intentions of the trustee (as is a constructive trust) but gives effect to his presumed intention.”119 But while it is true that the language of presumption can be found in some of the failed trust resulting trust cases, it is not a true presumption at all. There is sometimes an initial question whether, in the events which have happened, the transferee was intended to take outright the rights over which there was no effective declaration of trust or whether the trust was intended to extend to the whole. This is a matter of construction which, if answered in favour of the transferee, prevents a resulting trust arising. The question is unfortunately often described as involving a “presumption” of resulting trust, though one which can be “rebutted” by “proof” that the transferee was intended to take absolutely.120 117 An early case is Lloyd v Spillett (1740) 2 Atk. 148. 118 (1804) 9 Ves. 399. 119 [1996] A.C. 669 at 708. His Lordship’s references to the common intentions of the parties and

the presumed intentions of the trustee are odd. As we have seen, the fact proved by presumption is a declaration of trust by the transferor. 120 See, e.g. M.M. Bigelow, Story’s Commentaries on Equity Jurisprudence as Administered in England and America, 13th edn (1886), para.1199. In re West [1900] 1 Ch. 84 at 89, Kekewich J. said that there was a “presumption that a gift in trust is not a beneficial gift”.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

However, as we have seen, this is not a presumption at all, for no additional facts are proved.121 All that is involved is a question of the construction of facts already proved by evidence.122 In truth, the raising of the resulting trust is a question of intention only in a negative sense, and then not as a matter of presumption. The initial exercise of construction merely decides whether the transfer was or was not outright and not whether the transferor intended to create a trust for himself in the latter event. Moreover, the content of the words accompanying the transfer having been proved by evidence, it is impossible in such circumstances to invoke a presumption to prove that the uttered words had some different content. In Lord Browne-Wilkinson’s defence, it might be said that the resulting trust arising on failure does so because of a “presumption” that, in the events which have happened, the transferor would have wanted the rights to be so held. But this does not work either. The first difficulty is that, as we have seen, an unexpressed intention is never enough to create a trust.123 The second is that in all likelihood the transferor did not address his mind to the question. Indeed, it was for this reason that Harman J. in Re Gillingham Bus Disaster Fund rejected presumed intent as the explanation for such trusts. The resulting trust could not “rest on any evidence of the state of mind of the settlor, for in the vast majority of cases no doubt he does not expect to see his money back: he has created a trust which so far as he can see will absorb the whole of it.”124 Even if not contemplated, might it be said, adopting the economists’ fiction that people are selfish, that this is what the transferor would have wanted had he thought about it?125 This explanation, however, fares no better than the last, for once we admit that we are dealing with a fictitious intent, not only do we step outside the realm of presumptions, but we immediately encounter the problem that the House of Lords has forbidden courts from creating trusts based on speculative intent. As Lord Morris of Borth-y-Gest said in Gissing v Gissing: “When the full facts are discovered the court must say what is their effect in law. The court does not decide how the parties might have ordered their affairs: it only finds how they did. The court cannot 121 Text to fnn.11–14. This point seems not to be appreciated by Chambers: fn.2 above (1997), pp.45–46, 51, 66. 122 This point is also made in D. Waters, M. Gillen and L. Smith, Waters’ Law of Trusts in Canada, 3rd edn (2005), p.107. 123 Text to fn.51 above. 124 [1958] Ch. 300 at 310. The decision foreshadowed that of the House of Lords in Vandervell v IRC [1967] 2 A.C. 291, discussed immediately below. 125 cf. Restatement of the Law of Trusts (1935), p.1247.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

devise arrangements which the parties never made. The court cannot ascribe intentions which the parties in fact never had.”126 Moreover, insurmountable difficulties arise in terms of rebuttal if we see the failed trust resulting trusts as resting on a presumption. If the “presumption” is that a trust was declared in the transferor’s favour, the only facts capable of rebutting it will be ones showing that the transferee was to take outright or as security. Yet we already know, having undertaken the process of construction discussed above,127 that this particular transferee was intended to do neither. Indeed, the resulting trust only arises, the trust only fails, because this conclusion was reached. There can therefore be no facts capable of rebutting the “presumption”, and an irrebuttable presumption is, as we have seen,128 no presumption at all. VII. VANDERVELL V IRC That we are not here dealing with a presumption was the ratio decidendi of the House of Lords in Vandervell v IRC.129 Vandervell wished to donate a substantial sum of money to the Royal College of Surgeons to establish a chair of pharmacology. He was the beneficiary of a trust of a parcel of shares in a private company (the company), of which he was also managing director and principal shareholder. The trustee was National Provincial Bank Ltd. There was also a settlement for his children, the trustee of which was Vandervell Trustees Ltd (the trustee company). On Vandervell’s instruction, the shares were transferred by the bank to the college, his intention being that the college take the shares outright. The plan was for the company to declare a dividend on the shares sufficient to fund the chair. The reason the donation was structured in this way was to ensure that a substantial part of the total sum, 33.33 per cent, was given, not by Vandervell himself, but by the public purse. The dividends were to be gross dividends from which tax at 33.33 per cent would be deducted at source. That tax would then be recovered by the college as a charity. But the scheme would only work if Vandervell divested himself of all rights in respect of the shares. A failure to do so would mean that, by virtue of s.415 of the Income Tax Act 1952, he would be liable to surtax (a higher rate tax of 50 per cent) on the dividends declared. Unfortunately for Vandervell, his accountant advised that it be arranged for the college 126 [1971] A.C. 886 at 898. Similar statements can be found in the speeches of Viscount Dilhorne (at 900) and Lord Diplock (at 904–905). Apparent statements to the contrary in Stack v Dowden [2007] UKHL 17; [2007] 2 W.L.R. 831 fall foul of the doctrine of precedent, for no invocation of the 1966 Practice Statement (Practice Statement (Judicial Precedent) [1966] 1 W.L.R. 1234) was there made. 127 Text to fn.120–122 above. 128 Text to fnn.18–19 above. 129 [1967] 2 A.C. 291.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

to grant to the trustee company an option to buy back the shares once they had served their purpose. For estate duty purposes, it was thought prudent that the company be made public before Vandervell’s death and therefore inadvisable for a parcel of shares to be left outstanding in the hands of a third party. There was, however, no settled plan as to what was to happen to the shares on their return. Two possibilities were that they be applied to the children’s settlement or to a new settlement in favour of the company’s employees. One thing was certain, however: the shares would not go back to Vandervell, for that would ruin the tax efficiency of his scheme. The dividends were duly declared and paid, but the Inland Revenue charged Vandervell surtax on the amounts, arguing that the option to purchase was held by the trustee company for him on resulting trust, and that he had therefore failed to divest himself entirely of any interest in the shares. To reach that conclusion, three issues had to be decided. First, who was the grantor of the option? Secondly, in what capacity did the trustee company take the option? And thirdly, on the assumption that Vandervell was the grantor and that the trust company took the option on trust for undeclared objects, who was the beneficiary of the failed trust resulting trust? A majority of their Lordships130 held that though the college was in terms the grantor of the option, the real grantor was Vandervell. Though it had been asked to grant the option, “this was a matter of courtesy; at this time the college had no legal or beneficial interest in the shares and they could only comply with it. . . . [I]n law . . . it was [Vandervell] who procured the college to grant the option to the trustee company.”131 But because no words of trust were expressed in the grant to the trustee company, it was not immediately obvious how it took the option, especially when it was known that its objects went beyond merely acting as a trustee. The Court of Appeal132 resolved the issue by the use of a “presumption” of resulting trust: the grant of the option being gratuitous, the trustee company was “presumed to have acquired the benefit of the option upon a resulting trust in favour of the settlor.” Although that presumption could be rebutted by proof by evidence that Vandervell intended the trustee company to hold the option outright or on some other trusts, it had not been rebutted, for “the evidence shows that the settlor did not intend to grant the benefit of the option to the trustee company

130 Lord Donovan dissenting: [1967] 2 A.C. 291 at 320. 131 [1967] 2 A.C. 291 at 314 (Lord Upjohn). 132 [1966] Ch. 261.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

Law Quarterly Review

[Vol. 124

beneficially nor did the trustee company intend to accept it for its own benefit.”133 However, an approach based on presumptions is flawed for, as we have seen, presumptions only operate where there is a gap in the evidence, whereas here, all the relevant facts were known. In truth, the Court of Appeal’s “presumption” was a rule of construction, not evidence, a fact clearly appreciated by the House of Lords. Looking at the evidence, Lord Reid and Lord Donovan held that the grant of the option was not on trust at all. No trust was declared in the grant to the trust company, and it was perfectly consistent with Vandervell’s plans that the company took the option outright or “beneficially”.134 The majority, however, held that the option had been granted to the trustee company on trust. Lord Upjohn, with whom Lord Pearce agreed, doubted whether the presumption of resulting trust could apply to options to purchase, but said that he need not decide that question because “the facts and circumstances are sufficient for this purpose without resort to this long stop presumption.”135 His Lordship found that the intention of Vandervell was that the option be held by the trust company upon such trusts as he or the trustee company should from time to time declare. Lord Wilberforce likewise rejected the use of presumptions to resolve the issue, saying there was “no need, or room . . . to invoke a presumption. The conclusion, on the facts found, is simply that the option was vested in the trustee company as a trustee on trusts, not defined at the time, possibly to be defined later.”136 A trust for objects to be defined later being void for want of objects, a failed trust resulting trust arose in Vandervell’s favour which, because it was not triggered by a fact proved by presumption, could not be rebutted by proof by evidence of any fact whatever. That the resulting trust in Vandervell had nothing to do with a presumption is also made clear in Megarry J.’s judgment in Re Vandervell’s Trusts (No.2),137 in which he coined the terms “presumed” and “automatic” to describe the operation of the two types of resulting trust. After a detailed analysis of the speeches of Lords Upjohn and Wilberforce, he said that the case showed that the failed trust resulting trust did “not depend on any intentions or presumptions,” but was “the automatic consequence of [the settlor’s] failure to dispose of what is vested in him.”138

133 [1966] Ch. 261 at 289 (Diplock L.J.) 134 [1967] 2 A.C. 291 at 308–309 (Lord Reid), 321–323 (Lord Donovan). 135 [1967] 2 A.C. 291 at 315. 136 [1967] 2 A.C. 291 at 329. 137 [1974] 1 Ch. 269. 138 [1974] 1 Ch. 269 at 294.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

The Vandervell litigation therefore shows that an explanation of the failed trust resulting trust in terms of a presumption of intent does not work. Despite Lord Browne-Wilkinson’s obiter comments in Westdeutsche, the failed trust resulting trust, both as a matter of logic and of the highest authority, arises for a reason having nothing to do with proof by presumption of anything, least of all an intent by the settlor to create a trust for himself. VIII. WHY DO FAILED TRUST RESULTING TRUSTS ARISE? If it is not an intent proved by presumption, then what is it which causes failed trust resulting trusts to arise? (a) The Vandervell explanation As we have seen, Megarry J. described the failed trust resulting trust as arising “automatically”.139 This label is misleading. “Automatic” means “self-acting; having the power of motion or action within itself.”140 However, there is nothing “automatic” about the law. The failed trust resulting trust does not arise of its own volition, but because courts say it does. Moreover, there need be no such response, for we know that where the failed trust is charitable, the funds are often applied cy pr`es. Nor is it an inevitable response here. If our only concern is to strip out the enrichment of the transferee of the failed trust, it could be done by giving the settlor a personal claim in unjust enrichment for the value received.141 However, by awarding the settlor the benefit of a trust, we take the rights out of the estate available for distribution to the unsecured creditors and give him a privileged position in relation to them. Megarry J. was of course only paraphrasing the speeches of Lord Upjohn and Lord Wilberforce in Vandervell v IRC. What for Lord Upjohn made the trust “automatic” was the fact that, “If the beneficial interest was in A and he fails to give it away effectively to another or others or on charitable trusts it must remain in him.”142 Lord Wilberforce spoke in similar terms: “But the equitable, or beneficial interest, cannot remain in the air: the consequence in law must be that it remains in the settlor.”143 However, a theory based on “retention” of an equitable interest does not work either, for the settlor generally has no equitable interest to retain. Suppose I convey my fee simple title to land to a friend to hold on trust for 139 Re Vandervell’s Trusts (No.2) [1974] Ch. 269 at 294. 140 Oxford English Dictionary, fn.80 above, Vol.I, p.805. 141 Especially since the bar on recovery of payments made under mistake of law was removed in Kleinwort Benson v Lincoln CC [1999] 2 A.C. 349. 142 [1967] 2 A.C. 291 at 313. 143 [1967] 2 A.C. 291 at 329.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

100

Law Quarterly Review

[Vol. 124

“such objects of benevolence and liberality as he in his absolute discretion should most approve of.” The trust will fail for want of objects and the friend will hold the title for me on resulting trust. But I have “retained” nothing. At the beginning of the story, I had a fee simple title to land. At the end of it, that title is held by my friend. What I now have is an interest under a trust, something I did not have before. Might it be said that prior to the transfer, I had both a legal and an equitable right, and that though I gave away the former, I retained the latter? This seems to be implicit in the reasoning of Lord Upjohn and Lord Wilberforce. But it is mistaken, for having an equitable right is the consequence of there being a trust, not the reason why one arises. As Lord Browne-Wilkinson explained in Westdeutsche: “A person solely entitled to the full beneficial ownership of money or property, both at law and in equity, does not enjoy an equitable interest in that property. The legal title carries with it all rights. Unless and until there is a separation of the legal and equitable estates, there is no separate equitable title. Therefore to talk about the [transferor] ‘retaining’ its equitable interest is meaningless. The only question is whether the circumstances under which the money was paid were such as, in equity, to impose a trust on the [recipient]. If so, an equitable interest arose for the first time under that trust.”144 Thus, although Lord Upjohn and Lord Wilberforce correctly identify the failed trust resulting trust as not resting on a presumption, the reason they go on to give as to why it arises is flawed. There is no retention of anything by the settler: a new interest arises in his favour and there is nothing “automatic” about it. (b) The Chambers explanation The view taken by Chambers is that the failed trust resulting trust does not rest on a presumption. But that does not mean that he agrees with the explanation in Vandervell. Instead, the three types of resulting trust are said to be one and the same, the only difference being that the fact “presumed” in the voluntary conveyance and purchase money resulting trusts is now the fact “proved” in the failed trust resulting trust.145 Since in the latter the facts show that the transferee was not meant to take outright 144 Westdeutsche Landesbank Girozentrale v Islington LBC [1996] A.C. 669 at 706. His Lordship’s talk of a “separation” of legal and equitable rights is unfortunate, for it implies their prior existence, like the separation of the yolk and white of an egg. We should instead talk in terms of the equitable interest being “engrafted upon”, not carved out of, the legal interest: Re Transphere Pty Ltd (1986) 5 N.S.W.L.R. 309 at 311 (McLelland J.). 145 “. . . the resulting trust arises because the provider did not intend to benefit the recipient. There is no need to resort to a presumption of the provider’s intention where it has been proved”: Chambers, fn.2 above (1997), p.43.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Explaining Resulting Trusts

101

but only as trustee, albeit for a trust which has failed, a transfer with “no intention to benefit” has been “proved” to have occurred. However, for reasons which echo those given when rejecting his explanation of the “presumed” resulting trust, this explanation does not work either. First, the contrast between “proof” and “presumption” is false. As we have seen, presumptions, like evidence, are methods of proof, so that “presumed” facts are just as much proved as facts proved by evidence.146 Secondly, it is impossible to say that the trust arises because of proof of a transfer with “no intention to benefit”, for such a transfer is, as we have seen,147 something unknown to English law. There are transfers which cause transferees to hold outright, on trust, or by way of security, but there is no fourth category, at least where that phrase describes something distinct from a transfer on trust or as security. Thirdly, even if there was such a thing, the phrase “transfer with no intention to benefit” does not describe a fact and is therefore not susceptible of proof, by evidence or presumption. At best, “transfer with no intention to benefit”, like “legitimacy” and “resulting trust”, is a legal conclusion from proved facts. Thus, while it makes sense to describe Morice v Bishop of Durham as involving a proved by evidence mistaken transfer, we cannot say that it involved a proved by evidence “transfer with no intent to benefit the transferee”, for what would now be “proved” would be a legal conclusion from facts, not an additional fact in itself. Finally, even if “transfer with no intention to benefit” was a fact susceptible of proof, there is no compelling reason why the law should respond to its “proof” by raising a trust. As we have seen, if our only desire is to strip out the transferee’s enrichment, a response in the nature of a personal claim for the value received serves that purpose perfectly well. Indeed, because it ensures equality of treatment between creditors in the event of the transferee’s insolvency, it might be thought to do it better. IX. CONCLUSION Before any progress can be made in the search for an explanation of resulting trusts, a secure understanding of presumptions is required. There are many false presumptions, but only one true one, the presumption of law, where proof by evidence of one fact generates proof of a second fact without the need for evidence. This is the type of presumption in play in presumed resulting trusts. The difficulty with the two explanations of resulting trusts examined in this article is that this does not seem to have been appreciated. 146 cf. Chambers’ distinction of “actual” and “presumed” facts: fn.2 above (1997), p.43. 147 Text to fnn.100–114 above.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

102

Law Quarterly Review

[Vol. 124

We saw that the voluntary conveyance and purchase money resulting trusts arose because of the operation of a true presumption, the fact proved by presumption being that the transferor declared a trust in his own favour. This was the position in the seventeenth century and nothing has changed since. The argument of Birks and Chambers, that the fact “presumed” in such circumstances is that the transferor did not intend to benefit the transferee, was shown to be based on a number of misunderstandings. First, gratuitous transfers outside the relationships of advancement are not “apparent gifts”, only ambiguous transfers. Secondly, suspicions are not the same things as presumptions, and in any case, equity is not “suspicious” of gifts. Thirdly, it is not possible for equity to “presume” that “apparent” gifts are not gifts, for “not-gift” is at best a legal conclusion from proved facts, not a fact in itself. Fourthly, a “presumption” of “not-gift” cannot be a “presumption” of “non-beneficial transfer” for the law does not recognise a notion of non-beneficial transfer distinct from transfers on declared trusts or as security. And fifthly, no satisfactory explanation was given as to why, assuming there is such a thing as a “nonbeneficial transfer”, the law should respond to its “proof” by the raising of a trust for the transferor. For these reasons, the argument that there should, by a logical extension of the traditional resulting trusts, be resulting trusts in the generality of cases of unjust enrichment is unsustainable. We then examined the resulting trust which arises on proof by evidence of a transfer of rights on trusts which fail. Though Lord Browne-Wilkinson said it arose because of the operation of a presumption, Lord Upjohn and Lord Wilberforce showed in Vandervell v IRC 148 that this was not so: there is no room for a presumption where all the facts have been proved by evidence, and in the case of a failed trust resulting trust there is no fact left outstanding. However, the explanation then put forward by their Lordships in Vandervell itself did not work, for the settlor retains nothing when making such a transfer. And Chambers’ explanation, framed in terms of “proof” of a “non-beneficial transfer”, failed for the very same reasons that a “presumption” of “non-beneficial transfer” could not be the explanation of the “presumed” resulting trust. Thus, though we have a convincing though anachronistic explanation for the “presumed” resulting trust, the “automatic” resulting trust still defies legal analysis. WILLIAM SWADLING.*

148 [1967] 2 A.C. 291 at 329.

* Fellow of Brasenose College, Oxford. I have benefited greatly from discussions with Alexandra Braun, Neil Jones, Mike Macnair, James Penner, Colin Tapper, and James Walmsley. Presumptions; Resulting trusts

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

CAUSATION AND RISK IN THE HIGHEST COURTS OF CANADA, ENGLAND, AND FRANCE

INTRODUCTION THE past two decades have witnessed a rise in judicial concern over problems in proving factual causal links, especially in the presence of scientific or factual uncertainty. Appeals for increased flexibility in causal assessments have become more numerous, as have judicial decisions responding to these calls. Liability based on negligently increased risk is one of the most controversial amongst the different avenues explored by the supreme courts of Canada and the United Kingdom to depart from the traditional causal assessments in cases involving uncertainty. Risk-based analysis has proved particularly useful in the context of uncertainty affecting alternative causation mechanisms resulting in indivisible injuries.1 In these cases, one of two or more mutually exclusive causes, including the defendant’s negligence, may have produced the whole of the damage, but uncertainty renders impossible the determination of which of the different causal candidates is culpable. As difficulties resulting from the existence of several alternative causes are characterised by the impossibility of determining each of the causes’ effective contribution to the injury, all that can be said is that the defendant increased the total risk of the outcome by adding one more possible cause. Risk enhancement has justified inferring that causation has been proven by the required standard of proof where the defendant has materially increased the plaintiff’s risk of injury. It has been relied on in different decisions as either permitting the judge to draw a discretionary factual inference of causation or material contribution to the injury, or as requiring a legal inference to be drawn absent evidence disproving the existence of causation. The notion of risk has also grounded an alternative test of causation whereby materially increasing the risk of injury is treated as the equivalent of materially contributing to the injury. Most recently, material increase of a risk has been considered a sufficient basis for liability without the need to prove causation. All of these variations allow the plaintiff to recover in full. However, risk-based analysis may also ground partial recovery, in proportion to the defendant’s risk contribution. This paper first argues that there is currently no general trend in the case law of the highest courts of England and Canada to use risk-based 1 Cumulative causal processes will therefore not be commented on. On the distinction, see J. Stapleton, “The Gist of Negligence. Part II: the Relationship Between ‘Damage’ and ‘Causation”’ (1988) 104 L.Q.R. 389 at 401.

103

104

Law Quarterly Review

[Vol. 124

reasoning to ground causal assessments in the presence of uncertainty. Apart from very circumscribed situations, the supreme courts of England and Canada have generally insisted on grounding causal analyses in traditional civil liability and evidence principles. The absence of a general trend can be observed with regard to risk-based analysis used to justify, despite the uncertainty of causation, full recovery of the plaintiff’s damage (I), as well as proportional recovery (II). Recently, however, the House of Lords has proven more expansive by relying on the defendant’s negligent increase of the risk of injury to allow for proportional recovery of the claimant’s damages. Therefore, this paper subsequently proceeds to argue that in light of the 40-year history of proportional liability reasoning in the French Cour de cassation, common law courts should give careful consideration to the French experience with the numerous conceptual problems that this mechanism presents (III). I. ABSENCE OF A TREND IN FAVOUR OF RISK-BASED FULL RECOVERY Despite the striking generosity of recent House of Lords decisions, there is no emerging trend in common law favouring risk-based liability to solve instances of causal uncertainty involving alternative mechanisms.2 The highest court of the United Kingdom, together with the Canadian Supreme Court, demonstrates a general preference for preserving the integrity of traditional causation and evidence principles and only exceptionally accords generosity in assessing causation based on the idea of risk. Historical developments In common law, the increase of risk justification for full recovery resulted from the necessity to adapt the Bonnington Castings v Wardlaw test of material contribution to the damage3 to cases involving uncertainty in alternative causal mechanisms. In such cases, it was impossible to assess whether or not the defendant had contributed to the resulting damage other than by simply increasing the risk that the plaintiff suffer injury by adding an additional potential triggering agent. To address this challenge, the House of Lords in McGhee v National Coal Board 4 granted full recovery by accepting that materially increasing the plaintiff’s risk of injury could, under the particular circumstances of this case, be treated as equivalent to materially contributing to the plaintiff’s injury. The predominant Canadian interpretation of McGhee is that it grounds a factual inference of causation on risk increase.5 Although the House of Lords also initially subscribed to 2 The trend in the lower courts is different, however. 3 Bonnington Castings Ltd v Wardlaw [1956] A.C. 613. 4 [1973] 1 W.L.R. 1, HL. 5 Snell v Farrell (1990) 72 D.L.R. (4th) 289; [1990] 2 S.C.R. 311.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

105

this interpretation,6 it later abandoned this position7 for an interpretation of McGhee’s increase of risk test as being either a new approach to causation necessitated by policy8 or, more recently, as grounding liability without proof of causation.9 McGhee was soon restricted in its application. The House of Lords refused to extend it to compensate a premature baby suffering from retrolental fibroplasias (RLF) in the 1988 case of Wilsher v Essex HA.10 Lord Bridge of Harwich explained that the only way the plaintiff’s claim could succeed would be to establish, according to the traditional rules of evidence, that the defendant’s negligence was at least a material contributor to the injury. McGhee, an industrial disease case, had involved a single risk agent, brick dust, originating from two sources, one innocent and the other negligent,11 both controlled by the same defendant, the claimant’s employer. Unlike McGhee, Wilsher involved more than one possible alternative cause, each linked to distinct risk agents: the defendant’s negligence in administering excessive oxygen to the baby or different medical conditions that can affect premature babies. In Wilsher, McGhee was presented as a case in which the majority, adopting a robust and pragmatic approach to undisputed primary facts, had concluded that it was a legitimate common sense inference of fact that the defender’s negligence had caused (materially contributed to) the pursuer’s injury.12 Thus, the McGhee test was restricted to cases where an element of the causal inquiry has been proven, as where the injury was caused by one well-determined risk agent controlled by the defendant in both its innocent and negligent forms. Both McGhee’s increase of risk reasoning and the restrictions imposed to it by Wilsher, influenced the subsequent development of Canadian and UK13 common law. Canadian common law Although discretionary judicial inferences of causation are admitted in Canada, the Supreme Court of Canada’s position on inferences based on risk-enhancement remained unclear for some time. After Wilsher, the Supreme Court approached full recovery based on increase of risk 6 Wilsher v Essex Area HA [1988] 1 A.C. 1074. 7 Fairchild v Glenhaven Funeral Services [2003] 1 A.C. 32 at [21]. 8 [2003] 1 A.C. 32 at [21] while acknowledging absence of proof of causation under the traditional

but-for or material contribution tests. However, see Lord Hutton who treats McGhee as introducing a legal inference of causation: at [108]. 9 Barker v Corus UK Plc [2006] UKHL 20; [2006] 2 A.C. 572. 10 Wilsher v Essex Area HA, fn.6 above. 11 In the form of prolonged exposure to brisk dust due to the absence of a shower on the work premises. 12 See fn.10 above, at 1090. 13 This text refers to “UK common law” to acknowledge that some of the main cases discussed here actually originate from Scotland.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

106

Law Quarterly Review

[Vol. 124

cautiously. In Snell v Farrell, Sopinka J. conceded the possibility of drawing “common sense” inferences of causation “although positive or scientific proof of causation has not been adduced”.14 The admission of such a possibility was crucial to the outcome in this case since none of the experts had been able to affirm the cause of the plaintiff’s injury—atrophy of the optic nerve leading to blindness in her left eye. This injury could have been the result of the defendant’s negligence in continuing—contrary to recognised medical practice—the surgical removal of a cataract despite the presence of a small retrobulbar bleed in the patient’s eye. However, the patient was also suffering from glaucoma, hypertension and diabetes, all of which are potential causes of optic nerve atrophy. In response to this difficulty, Sopinka J. stated that common sense inferences could be drawn where “circumstances, other than a positive medical opinion, permit”.15 However, he was ambiguous about whether or not risk increase could be considered such a “circumstance”.16 While not expressly rejecting the possibility to rest causal inferences, discretionary or not, on risk increase, Sopinka J. treated Wilsher v Essex Area HA17 as authoritative. He also discarded what he referred ambiguously to as McGhee’s two “alternatives”: simply proving that the defendant created a risk of injury or imposing on the defendant the burden of disproving causation.18 Whether he was willing to allow such reasoning or not, some Court of Appeal cases of various common law provinces subsequently interpreted Snell v Farrell as supporting the grounding of factual inferences on risk increase.19 However, 11 years after Snell v Farrell was decided, the Supreme Court of Canada put an end to this line of cases by rejecting this possibility when addressing a civil law appeal from the province of Qu´ebec.20 In St-Jean v Mercier, the court clearly rejected risk-based full recovery, giving priority to traditional methods of causal assessment. The plaintiff 14 Snell v Farrell, fn.5 above, at 330. 15 Snell v Farrell, fn.5 above, at 330 and 336. This was so because of the differences in legal and

scientific conceptions of “certainty” and “probability”: at 335–336. See also Laferri`ere v Lawson (1991) 78 D.L.R. (4th) 609; [1991] 1 S.C.R. 541 at 609. 16 Sopinka J. did infer causation, but the inference was grounded on the defendant’s particular knowledge of the facts regarding causation and his negligent undermining of the plaintiff’s ability to prove causation. 17 Snell v Farrell, fn.5 above, at 326. However, Sopinka J. distinguishes Snell from Wilsher on the ground that the experts’ evidence in Wilsher was seriously in conflict, while in Snell the experts were simply unable to reach any conclusion. 18 He allows an exception only “where defendants who have a substantial connection to the injury are escaping liability because plaintiffs cannot prove causation”: Snell v Farrell, fn.5 above, at 362. Sopinka J. subsequently discusses the option of reversing the burden of proof, which may lead to think that this rejection of McGhee was limited to Lord Wilberforce’s position. 19 e.g. Webster v Chapman (1997) 155 D.L.R. (4th) 82 at 93, Man CA; Suchorab v Urbanski (1997) 156 Sask. R. 46 at 55, QB a contrario; Taylor v Hogan (1994) 119 Nfld & P.E.I.R. 37, NfldCA; Robinson v Sydenham District Hospital Corp (2000) 130 O.A.C. 109 at 116, Ont CA. 20 Qu´ebec is the only Canadian province where civil law principles originating from France govern private law matters. The Supreme Court of Canada acts as the highest appellate court for both civil law and common law Canadian jurisdictions.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

107

was hit by a car at 90km per hour while hitchhiking. He was immediately operated on by the defendant orthopaedic surgeon who, although aware of a suspected fracture of the T7 vertebra, concluded that it was either stable or benign and would not impede his ability to operate. It was only two days later that the patient started showing signs of neurological damage. After he was discharged and subsequently consulted a neurosurgeon, a diagnosis of paraparesis was made. The plaintiff was unable to convince the Supreme Court that his paraparesis had been caused by the defendant’s negligent omission to order his immobilisation and to undertake neurological tests that would have allowed the identification of his vertebral fracture. The court preferred the opinions of the defendant’s expert witnesses who testified that the neurological impairment was complete by the time the patient was admitted to the hospital, but that its manifestation was delayed; the car accident being thus the sole cause of the injury. Gonthier J. nevertheless addressed, in obiter, the issue of causal inference since the plaintiff had argued that causation should be presumed where, as here, the doctor’s fault had created a risk within the ambit of which the injury fell. He refused to rest presumptions of causation on risk increase and insisted that the only judicial presumptions that can be drawn are those permitted by Art.2849 of the Civil Code of Québec, i.e. factual inferences (called “factual presumptions”) based on the judge’s discretionary evaluation of the whole of the case’s evidence. Gonthier J. rejected the adoption of a separate means of proof based on increase of risk which would provide the plaintiff with a less stringent standard of proof to satisfy. In his opinion, such reasoning constituted an unacceptable attempt “at circumventing the traditional rules of proof on the balance of probabilities”,21 the standard of proof that Québec civil law inherited from the common law.22 While lacking in clarity, these statements can be, and were, interpreted as rejecting risk-increase as the ground for a binding inference, which would force the judge to deduce causation as soon as the defendant is shown to have negligently increased the plaintiff’s risk of sustaining an injury. St-Jean v Mercier was thereafter accepted as authoritative by the lower common law courts which have accordingly rejected risk-based full recovery.23 However, St-Jean, although partly based on a reference to Snell v Farrell, was influenced by considerations particular to the Québec

21 St-Jean v Mercier (2002) 209 D.L.R. (4th) 513; [2002] S.C.R. 491 at [116]. 22 Y.M. Morissette, “L’influence du droit franc¸ais sur le droit de la preuve au Qu´ebec” in H.P. Glenn

(ed.), Droit québécois et droit fran¸cais: communauté, autonomie, concordance (1993), pp.424 et seq.; J.C. Royer, La preuve civile, 3rd edn (2003), paras 43, 48, 56–65. 23 Allen v University Hospitals Board (2002) 5 Alta. L.R. (4th) 98, 13 C.C.L.T. (3d) 95, Alta CA; Jackson v Kelowna General Hospital (2006) 148 A.C.W.S. (3d) 807 at [33], BCSC; McPherson v Bernstein (2005) 139 A.C.W.S. (3d) 780 at [205], Ont SCJ; O’Grady v Stokes (2005), 50 Alta. L.R. (4th) 98, 375 A.R. 109, [2006] 6 W.W.R. 144, AltaQB at [123].

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

108

Law Quarterly Review

[Vol. 124

civil law tradition. This militates against a too hasty treatment of this decision as authoritative for common law jurisdictions. The Supreme Court’s insistence on an orthodox application of traditional civil law principles may actually be explained by two elements: its desire to protect the integrity of Québec civil law from the influence of common law techniques, as well as its understanding of the civilian judiciary’s creative role. According to the Québec (as well as the French) civil code, only a legislative text can create a “legal” presumption, binding on the courts.24 Consequently, one author argues that the civilian judge’s power to draw factual presumptions does not include the drawing of presumptions deduced from general pre-determined postulates, a technique that raises no conceptual difficulties in the common law. Judicial presumptions must consequently be based solely on inductions from the individual factual circumstances of each particular case.25 It is conceivable that the decision in St-Jean v Mercier was motivated by a desire to protect the integrity of the civil code by refusing to grant those judicial powers characteristic of the common law. Interestingly, despite the common origin of their respective civil law principles, Québec orthodoxy in this area is in clear contrast with the expansive approach taken by French law. French doctrine and case law have long acknowledged the existence of judicially created legal presumptions of causation, including those based on risk increase.26 The Supreme Court’s conservatism is thus not shared by the jurisdiction that has inspired most of its civil law principles. In February 2007, the Supreme Court of Canada, in the problematic decision of Hanke v Resurfice Corp,27 gave some indication that causation could be found based on negligent risk creation where it is impossible to prove it under the but-for test due to factors outside of the plaintiff’s control, such as absence of scientific knowledge.28 It is thus now unclear what the position of the highest Canadian court is with regard to findings of causation based solely on negligently increased risk in the face of causal uncertainty. Meanwhile, although the predominant position in the United Kingdom has been to reject this approach, the House of Lords has allowed for some striking exceptions.

24 Art.1350 C.N.; Art.2847 C.C.Q. 25 R.P. Kouri, “From Presumptions of Fact to Presumptions of Causation: Reflections on the Perils of

Judge-made Rules in Quebec Medical Malpractice Law” (2001) 32 R.D.U.S. 213 at 237, 240. Against: L. Khoury, Uncertain Causation in Medical Liability (2006), pp.207–211. 26 Khoury, fn.25 above, at Ch.5. French courts rely on such reasoning less than on the loss of chance technique, however. 27 (2007) 278 D.L.R. (4th) 643. This decision, which came too late for full discussion here, is open to criticism since it confuses the concepts of material contribution to the injury and material increase of risk: at [25], as well as the notions of cumulative and alternative causation: at [23], [27]–[28]. 28 (2007) 278 D.L.R. (4th) 643 at [25].

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

109

UK common law Prior to the recent House of Lords decision in Barker v Corus (UK) Plc, all the cases in which the House of Lords relied on risk assessment to face causal uncertainty were instances of industrial diseases involving difficulties in identifying the culpable wrongdoer in a situation of alternative causation.29 Based on these decisions, increase of risk may ground a finding of liability if: (i) the risk agent (e.g. asbestos) is known to be in a causal relationship with the ultimate damage (e.g. lung cancer)30 ; (ii) the risk agent originates from different sources which are all under the control of one or several negligent defendants; and (iii) there are no other alternative causes aside from the defendants’ negligent actions. Beyond these situations, the authority of Wilsher prevails and English orthodoxy is maintained in cases where distinct risk agents operate alternatively to cause an injury.31 After Wilsher, the House of Lords indeed repeatedly reiterated its attachment to the traditional approach in alternative causation cases involving distinct risk agents. McGhee’s increase of risk reasoning was maintained only in well-defined and restricted circumstances involving industrial diseases litigation, and this essentially for policy reasons. In Fairchild v Glenhaven Funeral Services Ltd,32 the House of Lords restricted McGhee to cases in which the connection between the risk agent and the injury could be demonstrated. However, Fairchild did extend McGhee’s application to a greater variety of causal difficulties since it involved a very different set of facts. In McGhee, the risk agent, brick dust, originated from a single source, the employer. It was, however, unknown whether the employer’s wrongdoing—not providing showers on the work premises—participated at all in causing the disease, since the injury could still have resulted from contact with brick dust despite the availability of showers. Therefore, a single source, the employer, was responsible for both innocent (not wrongful) and guilty conditions which could equally have been the cause of the disease.33 By contrast, in Fairchild, the three claimants contracted mesothelioma after several years of service under different employers, during which they had inhaled asbestos fibres.34 Though mesothelioma is clearly caused by asbestos, the process by which inhaling a fibre can, but need not, result in mesothelioma 29 In Canada, compensation of industrial diseases is dealt with under Workers Compensation Schemes only. Moreover, problems of identification of the culpable wrongdoer may be dealt with by reversing the burden of proof of causation in accordance with Cook v Lewis [1951] 1 D.L.R. 1; [1951] S.C.R. 830. 30 A. Layard, “Toxic Torts—a Landmark Decision” (2002) 4.4 Environmental L. Rev. 241 at 241. 31 This limitation has particular importance in the area of medical malpractice. 32 See fn.7 above. 33 E. Peel, “‘Loss of a Chance’ Revisited: Gregg v Scott” (2003) 66 M.L.R. 623 at 628. 34 It seems that there was also environmental exposure in this case, but that was ignored by the court: J. Stapleton, “Lords a’Leaping Evidentiary Gaps” (2002) 10 Torts L.J. 276 at 281.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

110

Law Quarterly Review

[Vol. 124

is unknown.35 Therefore, medical evidence could not say whether any particular exposure to asbestos dust at any particular time caused the plaintiffs’ disease. Although one of the negligent defendants undoubtedly caused the plaintiff’s injuries, it was impossible to attribute the cause to any one of them. The House of Lords nevertheless extended McGhee’s reasoning and found all of the defendants liable for the whole of the plaintiffs’ damage, on the ground that each of them increased the plaintiffs’ risk of injury. McGhee’s material increase of risk test was interpreted as an adaptation of the orthodox test of causation to the demands of justice which required that the pursuer be provided with a remedy.36 McGhee was described alternatively as a variation to the ordinary approach to causation,37 as a different and less stringent test of causation,38 and as treating material increase of risk as sufficient to satisfy the causal requirement.39 In reaching their decision, the Law Lords’ were influenced by policy factors,40 as well as by the fact that it had been established that the plaintiffs’ disease was caused by exposure to asbestos, thereby satisfying at least some element of the causal connection.41 The Law Lords were wary of encouraging unwarranted extensions of risk-based causation in tort law.42 Lord Hoffmann and Lord Rodger of Earlsferry limited its application by imposing strict conditions. Lord Rodger’s list of conditions is particularly illuminating.43 It demonstrates that Fairchild is applicable to only two types of situations: (i) where damage is proven to originate from one single agent emanating from the misconduct of at least one identified defendant, and creating a risk from 35 J.J. Godleski, “Role of Asbestos in Etiology of Malignant Pleural Mesothelioma” (2004) 14 Thoracic Surgery Clinics 479 at 479 and 482; R. Bueno and G.J. Gordon, “Genetics of Malignant Pleural Mesothelioma: Molecular Markers and Biologic Targets” (2004) 14 Thoracic Surgery Clinics 461 at 461. Other suspected co-carcinogenic causes include radiation therapy, genetic predispositions, and the SV40-contaminated polio vaccines administered to millions of people in the late 1950s and early 1960s in the United States and Europe: K.V. Shah, “Causality of Mesothelioma: SV40 Question” (2004) 14 Thoracic Surgery Clinics 487 at 487; H.I. Pass, M. Bocchetta and M. Carbone, “Evidence of an Important Role for SV40 in Mesothelioma” (2004) 14 Thoracic Surgery Clinics 489 at 489; L. Zellos and D.C. Christiani, “Epidemiology, Biologic Behavior, and Natural History of Mesothelioma” (2004) 14 Thoracic Surgery Clinics 469 at 469 and 471. 36 Fairchild v Glenhaven, fn.7 above, at [21] (Lord Bingham of Cornhill). Against: Lord Bridge in Wilsher who treated it as grounding an inference: at [22] (Lord Bingham), [144] [150] (Lord Rodger). 37 Fairchild v Glenhaven, fn.7 above, at [35] (Lord Bingham). Lord Bingham admits that his approach or that of drawing legal or factual inferences do not result in much practical difference in this case. 38 At [45] (Lord Nicholls). 39 At [142], [144] (Lord Rodger). Most Lords rejected the factual inference interpretation in cases where medical experts’ opinion did not support them: at [35] (Lord Bingham), [150] (Lord Rodger). See also at [65] and [70] (Lord Hoffmann). Against: Lord Hutton, at [108], [100] and [102] and Snell v Farrell, fn.5 above. However, Lord Hutton ultimately defends the view that McGhee’s inference should be considered legal, not factual: at [108]. 40 At [32]–[33] (Lord Bingham), [36] (Lord Nicholls), and [56] (Lord Hoffmann). 41 At [62] (Lord Hoffmann). 42 At [73] (Lord Hoffmann), [34] (Lord Bingham), [43] (Lord Nicholls), [118] (Lord Hutton), and [169]–[170] (Lord Rodger). 43 At [169]–[170]. See also Lord Hoffmann: at [61].

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

111

which the damage materialises (Fairchild ); and (ii) where the different possible negligent and innocent causes of the damage involve an identical risk agent originating from only one defendant (McGhee). The Law Lords also confirmed Wilsher’s continued governance of cases not falling within the circumscribed exceptions, particularly those cases of alternative causation involving a non-tortious cause in which the exact agent of the damage is unidentifiable and might have flowed from disparate risk sources.44 Hence, causation findings based on risk increase would not be found in all alternative cause cases where the background risk originates from a cause foreign to the defendant, and not attributed to anyone’s negligence.45 This was confirmed in the 2004 House of Lords decision in Gregg v Scott.46 Fairchild is one exceptional case in which the House of Lords allowed a risk-based assessment to ground full recovery. Indeed, liability for the damages was not apportioned in Fairchild on the basis of the extent of the defendant’s contribution to the risk, but this was essentially because the defendants did not request it. Had they done so, such apportionment might have been problematic since it would amount to indirectly accepting a form of risk-based proportional assessment of damages which was yet to be admitted by the House of Lords.47 Indeed, since the defendants’ contribution in Fairchild was to the risk of injury, limiting their liability to the extent of that contribution would come close to treating the injury as the loss of a chance.48 Barker v Corus (UK) Plc was to break down many of the above limitations in May 2006. It extended McGhee and Fairchild beyond those cases in which the alternative causes were tortious acts by several defendants (Fairchild ) or the wrongful and non-wrongful acts of one defendant (McGhee), although still necessitating that all alternative causes be linked to the same risk agent. It also recognised that liability without proof of causation may exist for policy reasons. Finally, it departed from the traditional House of Lords wariness of proportional recovery by accepting the redefinition of the injury as an increased risk, thereby allowing recovery in proportion to the risk created by each defendant. This last aspect of the case is discussed in section III. 44 At [21] (Lord Bingham), and [118] (Lord Hutton). Medical cases are for the most part governed by such causal mechanism. 45 J. Stapleton, “Cause-in-Fact and the Scope of Liability for Consequences” (2003) 119 L.Q.R. 388 at 398, relying on Fairchild ’s approval of Wilsher; M. Hogg, “The Raising of Lazarus: the Resurrection of McGhee v National Coal Board ” (2003) 7 Edinburgh L. Rev. 80 at 81, 85. Against: M. Fordham, “Causation in the Tort of Negligence—a Dispensable Element?” [2003] Singapore J. Legal Stud. 285 at 298. 46 [2005] UKHL 2; [2005] 2 A.C. 176 at [78] (Lord Hoffmann). See also at [174] (Lord Phillips, dissenting on other grounds). 47 Khoury, fn.25 above. 48 See also Fordham, fn.45 above, at 299 and text between fnn.61–66 below.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

112

Law Quarterly Review

[Vol. 124

As in Fairchild, the claimant in Barker died of asbestos-related mesothelioma. Although its causal mechanism remains unclear, mesothelioma was treated as involving an alternative causal mechanism since it is thought that even a single fibre can trigger the disease. The causal agent (asbestos dust) originated from three different alternative sources, including the claimant’s employment under two negligent employers, one of which was bankrupt and without an identified insurer. However, Fairchild and Barker can be distinguished because in Barker the third source of the risk agent was independent of the defendant employers’ conduct and outside of their control. It was Barker’s contributory faulty exposure to asbestos dust while self-employed. Lord Hoffmann stated that the material increase of risk concept could extend to cases in which there is a non-actionable source of risk, a risk created by someone who is not a wrongdoer in relation to the injured person.49 Nonetheless, the authority of Wilsher remained intact since in Barker, as in McGhee and Fairchild, the agent of the risk, although originating from various sources, is always of the same nature (or, in the words of Lord Hoffmann, operates in the same way50 ). Thus, in those alternative cause cases where there are several risk agents emanating from distinct tortious and non-tortious causes, traditional principles still apply.51 Besides further extending risk-based liability, Lord Hoffmann, writing the main majority decision, also reinterpreted the mechanism through which liability is imposed. In his opinion, Fairchild did not proceed upon a fiction that a defendant who has created a material risk of mesothelioma is deemed to have caused (materially contributed to) the plaintiff’s disease.52 Rather, it established a way to impose liability without causation. Lord Scott of Foscote agreed, stating that liability was not found because the defendants’ breaches of duty had caused the mesothelioma—causation had actually not been proven—but because they had materially increased the risk that the employee would contract it.53 Barker’s imposition of liability without evidence of causation was not entirely new to the House of Lords; similar reasoning had just been applied in the informed consent case of Chester v Afshar (even in the absence of any evidence of increase in risk) although it could have been resolved under traditional principles.54

49 See fn.9 above, at [16]–[17] (Lord Hoffmann). 50 See fn.9 above, at [24] (Lord Hoffmann). 51 At [18]–[19] (Lord Hoffmann). 52 At [31] (Lord Hoffmann). As raised by Lord Rodger at [80], this interpretation is subject to criticism since most of the Lords agreed in Fairchild that McGhee’s increase of risk test constituted a less stringent test of causation: e.g. at [45] (Lord Nicholls), [77] (Lord Rodger). Only Lord Hoffmann expressly refused to proceed on that basis: Fairchild v Glenhaven, fn.7 above, at [65]. 53 At [53], [61]. See also Lord Walker’s more ambiguous position: at [104], [113]. 54 L. Khoury, “Chester v Afshar: Stepping Further Away from Causation?” [2005] Singapore J. Legal Stud. 246.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

113

Allowing liability to be imposed even in the absence of the traditional demonstration of the causal connection permitted the House of Lords in Barker to go one step further and grant proportional recovery. This was because the court was of the opinion that liability in the absence of causation can not fairly exist unless it results in proportional recovery of damages, such as according to each defendant’s risk contribution. II. ABSENCE OF A TREND IN FAVOUR OF RISK-BASED PROPORTIONAL RECOVERY Generalities Proportional recovery55 regards exposure to risk and lost chances as actual injuries worthy of redress. For instance, if it is established that the defendant increased the risk that the plaintiff would suffer an injury by 40 per cent, or made the plaintiff lose a 40 per cent chance of avoiding an injury, the plaintiff may recover compensation for 40 per cent of the final outcome. This amount reflects the risk created by the defendant and the damage suffered by the plaintiff as a result of that risk, or it represents the proportion of the chance to avoid suffering injury of which the defendant deprived the plaintiff. This reasoning has long been accepted in the “classic” loss of chance cases in which the defendant’s negligence crystallises the plaintiff’s situation, preventing him altogether from seizing a chance. Some examples are when a lawyer has caused the limitation period of his client’s claim to lapse56 or where a person has been prevented from participating in a beauty contest.57 In these cases, the language of chance is used to assess the resulting injury58 by calculating not the value of the anticipated end result, but the probability of its realisation.59 But these classic cases must be distinguished from the more controversial instances that interest us here. In these cases, the chance language is used to impose liability notwithstanding that causation between the defendant’s negligence and the plaintiff’s actual injury—for instance death, sickness or disability—is not shown on traditional evidence and causation principles. To overcome this impasse, causation may be assessed with regard to an alternative head of injury, which is now defined to be the lost chance of avoiding the 55 Also called probabilistic assessment of damages, or probabilistic increased-risk concept. 56 Kitchen v Royal Air Force Association [1958] 1 W.L.R. 563, CA. 57 Chaplin v Hicks [1911] 2 K.B. 786. 58 J. Bor´e, “L’indemnisation pour les chances perdues: une forme d’appr´eciation quantitative de la

causalité d’un fait dommageable” J.C.P. 1974.I.2620, para.29; J. Penneau, La responsabilité médicale (1977), para.104; S.M. Waddams, “Damages: Assessment of Uncertainties” (1998) 13 J.C.L. 1 at 5. 59 J. Flour and J.L. Aubert, Droit civil. Les obligations. 2) Le fait juridique (1997), para.134; Boré, fn.58 above, para.7; I. Vacarie, “La perte d’une chance” (1987) 3 Revue de la recherche juridique 903 at 924; V. Tacchini-Laforest, “Réflexions a` propos de la perte de chance” (1999) 142 Petites Affiches 7 at 8.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

114

Law Quarterly Review

[Vol. 124

ultimate damage, a chance of which the defendant negligently deprived the plaintiff. While the doctrine and case law has most often illustrated the use of proportional recovery in the context of compensation for lost chances, it may actually be grounded in various other concepts. For example, it may be the result of a court’s decision to indemnify the plaintiff for the negligently increased risk he is now facing.60 Risk-based liability (or liability for risk) sees risk, not lost chance, as an actual injury worthy of redress.61 It seeks to compensate for the proportion of the risk of damage that the defendant’s negligence increased. In many cases, whether the injury is expressed in terms of a lost chance or an increased risk may not make any difference: the two approaches may coincide in their results. However, this will not always be so. Depending on the selected reference point, chance or risk, the calculation of the plaintiff’s damages may differ. Nevertheless, loss of chance and risk-based proportional recovery are but two species of the same phenomenon: proportional recovery of damages. Numerous judges and authors have recognised this duality of the phenomenon of proportional recovery.62 Fleming has argued that loss of chance “has its mirror image in the proposition that exposure to risk should itself be compensable”.63 However, both Fleming and French author, Ren´e Savatier, distinguish the two concepts by stating that chance quantifies the probability of a positive outcome, while risk describes the probability of a negative outcome.64 Others have also conflated chance and risk by stating that, under the loss of chance argument, it is sufficient to prove that the wrongdoing increased the risk of damage and that compensation

60 Another approach views proportional recovery as replacing the all-or-nothing balance of probability standard with a technique that calculates damages on the basis of a judge’s assessment of the ex post probabilities of existence of causation between the injury and the fault: A. Porat, A. Stein and S.M. Cohen, “Indeterminate Causation of Damages: an Essay on Holtby, Allen, and Fairchild” (2003) 23 O.J.L.S. 667. See also M. Stauch, “Causation, Risk and Loss of Chance in Medical Negligence” (1997) 17 O.J.L.S. 205 at 223; Savatier, note under Cass civ 1re, 14 December 1965, J.C.P. 1966.G.II.14753 (note Savatier), also cited at: Bull. civ. 1965.I.541, [707]; D. 1966.jur.453; (1967) R.T.D. civ. 181 (note Durry). This approach is found in some French cases, e.g. Grenoble, 24 October 1962 (unpublished) commented on by A. Tunc, “Perte de chance de guérison par suite d’une erreur de diagnostic” (1963) R.T.D. civ. 334 at 335. 61 D. Gerecke, “Risk Exposure as Injury: Alleviating the Injustice of Tort Causation Rules” (1990) 35 McGill L.J. 797 at 801. This form of proportional recovery or apportionment is to be distinguished to that which takes place as between wrongdoers on the basis of their respective degree of fault as a corollary to the imposition of joint liability. 62 e.g. Lord Scott in Barker v Corus, fn.9 above, at [85] and Lord Mackay in Hotson v East Berkshire AHA [1987] 1 A.C. 750 at 786. 63 J.G. Fleming, Law of Torts, 9th edn (1998), p.229. See also Lord Walker in Barker v Corus, fn.9 above, at [114]; C.J. Lewis, “Negligence and the Chance of Injury” (1986) Professional Negligence 119; Stauch, fn.60 above, at 223; G. Mémeteau, “Perte de chance en droit médical français” (1986) 32 McGill L.J. 125 at 149; Peel, fn.33 above, at 626. Against: N. Jansen, “The Idea of a Lost Chance” (1999) 19 O.J.L.S. 271 at 295. 64 Fleming, fn.63 above, at p.229; R. Savatier, “Le droit des chances et des risques dans les assurances, la responsabilité civile dans la médecine et sa synthèse dans l’assurance de responsabilité médicale” (1973) 44 Revue Générale des Assurances Terrestres 457 at 458.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

115

should be in proportion to this increase of risk.65 By increasing the risk that the plaintiff would suffer the illness, the defendant has in effect reduced the latter’s chances of avoiding the outcome that the risk portends.66 French courts have also recognised this parallel implicitly by sometimes compensating lost chances of cure or survival on the basis that the defendant doctor had, by his negligence, increased the patient’s risks of suffering the injury which has occurred.67 There is thus enough commonality between the two types of reasoning to treat proportional recovery based on chance as the flip side of proportional recovery based on risk. Therefore, the approaches to loss of chance provide interesting points of reflection in addressing the validity of risk-based proportional liability in the common law. There is no trend in UK and Canadian common law favouring the use of proportional recovery, either grounded on risk or chance, as a technique to by-pass causal difficulties. In fact, the Supreme Court of Canada has expressly rejected proportional recovery grounded in loss of chance language in alternative cause cases shrouded with causal uncertainty. Although the House of Lords has not taken such a firm stand, this court has failed to apply proportional recovery based on loss of chance in cases where it was clearly possible to do so. Yet, the House of Lords has recently taken position in favour of proportional recovery based on the increase of risk line of reasoning, a stand antithetical to the Canadian Supreme Court’s conservative position. In doing so, the House of Lords clearly introduced into UK common law very similar reasoning to the French Cour de cassation’s justification of proportional recovery in the context of lost chances. Canadian common law As in the case of full recovery based on risk, the Canadian common law position on proportional recovery in light of causal uncertainty has been greatly influenced by a civil law decision rendered by the 65 D. Jutras, “Expertise médicale et causalité” in Congrès du Barreau du Québec (1992), p.897 at p.905. 66 This leads Robertson to argue that McGhee was concerned with loss of chance: G. Robertson,

“Overcoming the Causation Hurdle in Informed Consent Cases: the Principle in McGhee v N.C.B.” (1984) 22 U.W.O.L. Rev. 75 at 90. See also M. Hogg, “Lost Chances in Contract and Delict” (1997) S.L.T. 71 at 73. 67 e.g. Cass crim, 20 March 1996, Resp. civ. et ass. 1996.comm.283, commented on by P. Jourdain, “La perte d’une chance, dommage par ricochet” (1996) R.T.D. civ. 912 (In France, the victim of a penal infraction who has sustained an injury may become a civil party to a penal trial); Cass civ 1re, 17 November 1982, D. 1984.jur.305, 308 (note Dorsner-Dolivet), Bull. Civ. 1982.I.n333; D. 1983.380; J.C.P. 1983.G.IV.41; J.C.P. 1983.G.II.20056 (note Saluden)) commented on by G. Durry, “Sur un probable revirement de jurisprudence en matière de réparation de la perte d’une chance de guérison ou de survie” (1983) R.T.D. civ. 547 at 547; Toulouse ch.1, July 6, 1998, JurisData no.1998-043577; Chambéry, 5 February 2002, quashed by the Cour de cassation on another question: Cass civ 1re, 29 June 2004, no.02-13416, Juris-Data no.2004-024471. Against: Cass civ 1re, 10 January 1990 and 5 February Bull. Civ. 1990.I.n10; D. 1991.somm.358: the increase of risk was too small to apply loss of chance.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

116

Law Quarterly Review

[Vol. 124

Supreme Court of Canada. In the 1991 case of Laferrière v Lawson, the Supreme Court rejected the use of loss of chance in medical liability cases involving absent or uncertain causation. Laferrière involved a breast cancer specialist who failed to inform his patient of a malignant lump in her breast and to arrange follow-up treatment. Probabilities were to the effect that the deceased patient would not have survived, even had appropriate treatment been administered in time. The Supreme Court refused to compensate the plaintiff for the loss of a chance of getting treatment at an earlier stage.68 It not only rejected the argument that loss of chance constitutes a simple redefinition of the damage, but also held that, when used in the medical context, it artificially bypasses the causation requirement. Speaking for the majority, Gonthier J. considered compensation for lost chances admissible only in exceptional cases where the damage could be understood solely in probabilistic and statistical terms, and where it is impossible to evaluate whether or how the chance would have been realised in the plaintiff’s particular case. As an illustration, he referred to the classic example of the individual negligently deprived of a chance to participate in a lottery. Gonthier J. also considered the loss of chance analysis less objectionable when used to evaluate hypothetical or future damages where no other competing causal factor aside from the defendant’s negligence can be identified. However, he approached loss of chance with extreme caution in cases where there are serious doubts as to the defendant’s causal role because of other identifiable causal factors. He was also concerned with the fact that usually, in the medical context, the damage has already occurred, manifesting itself in sickness or in death. Thus, the chance is not suspended, with the future outcome remaining forever hypothetical, as in the classic cases; it has already been realised.69 Thus, Gonthier J.’s decision epitomises a general rule that denies proportional liability where (i) other causal factors can be identified (in an alternative causal process); (ii) there are serious doubts as to the defendant’s causal role in the face of these other causal factors; and, (iii) the chance has been realised and the damage has occurred.70 Canadian common law courts thereafter treated this judgment 68 Laferrière v Lawson, fn.15 above, at 609–612 (LaForest J. dissenting). It granted damages for anguish and frustration, however. 69 Laferrière v Lawson, fn.15 above, at 605. 70 See also Athey v Leonati (1996) 140 D.L.R. (4th) 235; [1996] 3 S.C.R. 458 at 478. Major J. observes that if the trial judge had found the existence of a realistic chance that injury would have occurred at some point in the future without the accident, a reduction of the damages would have been considered. He did not reduce damages on the basis of the defendant’s contribution to the injury, assessed at 25 per cent by the trial judge. This decision involved a cumulative causation mechanism, however. On this case: J. Cassels and C. Jones, “Rethinking Ends and Means in Mass Tort: Probabilistic Causation and Risk-Based Mass Tort Claims after Fairchild v Glenhaven Funeral Services” (2003) 82 Can. Bar Rev. 597 at 621 and 626–627 and D. Cheifetz, “Materially Increasing the Risk of Injury as Factual Cause of Injury: Fairchild v Glenhaven Funeral Services Ltd. in Canada” (2004) 29 Advocates’ Q. 253.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

117

as authoritative despite the fact that it was analysed according to the rules of Qu´ebec civil law and was predominantly based on the critiques of the French doctrine. This incorporation may be justified, however, since the objections cited in the French doctrine regularly find their source in many concerns relevant to the common law, some of which are detailed in section III. Meanwhile, the House of Lords adopted a similar position, although it has demonstrated interest in proportional recovery on some occasions. UK Common Law The House of Lords’ attitude towards proportional recovery as the remedy for causal uncertainty is less categorical than its Canadian counterpart’s. Never squarely rejecting the concept, the House of Lords has, however, refused to resort to proportional recovery based on loss of chance in order to overcome causal difficulties in medical malpractice litigation, restricting it to its traditional application to the classic cases. However, in the recent industrial disease case of Barker, the Law Lords clearly allowed such recovery in the absence of causation, referring to risk justification rather than lost chance. But they simultaneously maintained a firm reluctance to extend its application to medical malpractice cases. Although Hotson v East Berkshire Area Health Authority 71 is often cited as the authority for the rejection of loss of chance, the House of Lords actually avoided directly addressing the issue in that case. A negligent delay in treatment allowed for the development of avascular necrosis in the claimant’s left hip. The medical evidence showed that it was likely the disability would have resulted anyway but that there was a 25 per cent chance that the claimant would have recovered had he been promptly treated. The Law Lords avoided addressing the contention that a substantial, as opposed to a speculative, chance of benefit or advantage is an asset, the loss of which could be compensated in tort if foreseeable and proven on the balance of probabilities to have been caused by the defendant’s negligence. They preferred to focus on whether the claimant had demonstrated a causal connection on the balance of probabilities between the avascular necrosis and the defendant’s negligence. Lord Ackner and Lord Bridge restricted the loss of chance concept to the valuation of damages, which arises only once such a link is established on the balance of probabilities.72 Lord Mackay of Clashfern, Lord Ackner, and Lord Bridge distinguished the determination of contingent future events from the determination of past events, stressing that the latter must be decided on the balance of probabilities.73 However, Lord Bridge 71 See fn.62 above, at 750. 72 At 782 (Lord Bridge), 793 (Lord Ackner). 73 At 792–793 (Lord Ackner), 785 (Lord Mackay).

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

118

Law Quarterly Review

[Vol. 124

admitted that sometimes, particularly in instances of medical negligence, causation may be so shrouded in mystery that the court can only measure statistical chances, although that was not the present case.74 As for Lord Mackay, he drew a parallel between materially increasing the risk of injury and materially decreasing the chance of escaping injury. He was thus not ready to hold that loss of chance could never support a successful claim in damages until the House of Lords overruled McGhee. However, in Hotson the plaintiff had no chance of avoiding the necrosis, since, by the time he first presented himself at the hospital, he was “doomed”.75 Seventeen years later, Gregg v Scott confirmed the House of Lords’ reluctance to adopt loss of chance reasoning, although several of the judgments comment favourably on the concept, with Lords Nicholls of Birkenhead and Hope of Craighead even accepting it. In Gregg, in contrast with Hotson, the final outcome was still prospective at the moment of the trial. The defendant physician’s negligent misdiagnosis of a cancerous lump, which led to a nine-month delay in the claimant’s treatment, reduced the latter’s 10-year prospects of disease-free survival from 42 per cent to 25 per cent.76 Referring to the Qu´ebec civil law decision of Laferri`ere v Lawson,77 Lord Hoffmann stated that the loss of chance principle has no application unless the ultimate damage can be clearly attributed to the defendant’s wrongful act. That was not the case here since the claimant’s prospects of being cured were less than 50 per cent at the moment of the doctor’s breach of duty.78 He considered the adoption of this possible, rather than probable, causal link as the criterion of liability, so radical a change to the law as to amount to a legislative act better left to Parliament.79 Lord Phillips of Worth Matravers was willing to concede to some application of the loss of chance doctrine, where the adverse outcome has occurred, but not where it was still prospective.80 He also worried about the possible blow to the coherence of the common law “if special tests of causation are developed piecemeal to deal with perceived injustices in particular factual situations”.81 Baroness Hale of Richmond distinguished Gregg from Hotson,82 but she was also concerned with the consequences of adopting the loss of chance argument since, inter alia, almost any claim for loss of an outcome could be reformulated as a claim

74 At 782. 75 At 786 (Lord Mackay). 76 Gregg v Scott, fn.46 above. 77 Gregg v Scott, at [81]. 78 At [67]–[68], [71]. 79 At [90] (Lord Hoffmann). Lord Phillips and Baroness Hale agreed. 80 At [188], [190]–[191]. 81 At [172]. He also raises practical difficulties in identifying in terms of percentage the effect that clinical negligence has on the chances of a favourable outcome: at [170]. 82 At [211]–[212].

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

119

for loss of a chance of that outcome.83 As for the dissenting judges, Lords Hope and Nicholls, they would have granted damages for loss of chance despite also distinguishing Hotson.84 Among other arguments, Lord Nicholls contended that whether the prospect of recovery was above or below 50 per cent, the patient had lost something of value, and that denying him recovery in the second case would void the substance of the doctor’s duty to his patient.85 He was ready to extend loss of chance to cases where the patient’s reasonable prospects of recovery from a preexisting illness or injury were fraught with a significant degree of medical uncertainty, albeit under some restrictions.86 In May 2006, the House of Lords finally granted proportional recovery, although it was for a negligently increased risk, rather than a lost chance. The question left unresolved by Fairchild, namely the issue of whether damages should be apportioned based on each defendant’s risk contribution, was resolved positively in Barker v Corus (UK) Plc.87 Refusing to impose joint liability (solidary liability) on the defendants, the House of Lords held that the necessary and fair corollary to the Fairchild decision which imposed liability based on negligent risk increase without actual proof of a causal link to the injury, was to apportion damages based on the extent of each defendant’s contribution to the risk.88 Interestingly, Lord Hoffmann, writing the majority reasons supporting this conclusion, justified it using reasoning long known to the French Cour de cassation and extensively applied in its loss of chance case law. He insisted that if the basis of liability is the wrongful creation of a risk or chance of causing the disease, then “the damage which the defendant should be regarded as having caused is the creation of such risk or chance”. This approach involves having to quantify the likelihood that the ultimate injury, which has materialised, was caused by the particular defendant,

83 She was also concerned that nothing would prevent this approach from being extended to medical negligence: at [225] and that the coexistence of loss of chance and the all-or nothing rule as alternatives would have substantial implications: at [224]. 84 At [38] (Lord Nicholls) and [109]–[110] (Lord Hope). 85 At [3]–[4]. cf. Chester v Afshar [2004] UKHL 41; [2005] 1 A.C. 134 (breach of a medical duty to inform). Part of Lord Nicholls’ reasoning is also based on the distinction between past events, future possibilities, and hypothetical events: see [9]–[10] citing Lord Diplock in Mallett v McMonagle [1970] A.C. 166 at 176, although he is critical of it, especially when applied to hypothetical events: at [13]–[14], [17]. 86 For these, see Gregg v Scott, at [46] and [48]–[53] (Lord Nicholls) and [115]–[116] (Lord Hope). cf. Chester v Afshar, fn.85 above. 87 See fn.9 above. 88 At [61] (Lord Scott): Fairchild is perhaps an anomalous exception to causation principles necessitating consequential adjustments to other principles of tortious liability. Recovery in proportion to each defendant’s contribution to the injury had previously been granted by the Court of Appeal in Holtby v Brigham & Cowan (Hull) Ltd [2000] 3 All E.R. 421, CA and Allen v British Rail Engineering Ltd [2001] I.C.R. 942, CA which are not cited in Barker. However, these cases involved cumulative causes, where the degree of exposure was considered to impact on the gravity of the injury. On these cases see Porat and others, fn.60 above.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

120

Law Quarterly Review

[Vol. 124

as is done in classic loss of chance cases.89 Lord Walker of Gestingthorpe agreed that the damage must be apportioned according to each defendant’s contribution to the risk,90 to be calculated by taking into account the duration of the exposure,91 as well as its intensity.92 Baroness Hale disagreed with the approach that describes the gist of the claim as the risk of contracting mesothelioma. However, she arrived at the same conclusion as did the majority in relation to apportionment based on risk contribution on the ground that there should be a logical connection between the law’s approach to “causation” and its approach to the extent of liability.93 Lord Rodger also dissented on the characterisation of the injury, noting that in Fairchild, the defendant was liable for causing mesothelioma, not for causing the risk of mesothelioma. In other words, the damage was mesothelioma, not the risk of causing it.94 Preoccupied with maintaining the consistency of the Fairchild approach with the main body of law on personal injuries, he also dissented on the question of apportionment and objected to treating an increase of risk as the injury.95 For Lord Hoffmann, the decision was taken for policy reasons, in order to “smooth the roughness of the justice which a rule of joint and several liability creates”.96 Lord Walker also insisted on the need to reduce the unfairness to an employer caused by burdening him with liability for an injury which may not in fact have been caused by the exposure for which he was responsible.97 Regardless of their different approaches, the Law Lords all insisted on the exceptional nature of the liability found in this case. Lord Hoffmann was careful to protect the Gregg v Scott precedent and refused to extend the Barker approach to medical malpractice cases. Lord Walker agreed on the basis that “such an extension would lead to a great uncertainty in a large number of clinical negligence cases”.98 Lord Scott argued that while Fairchild did not establish an overarching principle in the law of tort,99 there was no conceptual objection to treating the diminution of the chance of a favourable outcome or the increase of the risk of an unfavourable outcome as actionable damage. He nevertheless expressed concerns about the possible extension of Fairchild to all cases of medical negligence, 89 See fn.9 above, at [35]–[36]. 90 At [48]. 91 At [62] (Lord Scott), [109] (Lord Walker). 92 These factors demonstrate that proportional recovery based on risk may lead to results different from

that based on lost chances. 93 At [120], [124]. This reference to causation appears erroneous if one considers that Baroness Hale agrees with Lord Hoffmann’s reasons. 94 At [68]. 95 At [84]–[85]. 96 At [43]. 97 At [109]. See also [125], [127] (Baroness Hale). 98 At [114]. 99 At [57].

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

121

if not beyond, which would create inconsistencies with Wilsher, a case which did not resort to the loss of chance doctrine even though the defendant did in fact deprive the patient of the chance to avoid injury.100 He consequently restricted the application of proportional recovery to cases in which only a single injurious agent is involved101 ; for example, in cases where it is known that asbestos caused the injury, but the source of the culpable fibre is unknown.102 Thus, limitations on the extent of Barker’s application are based on the fear of creating uncertainty in the law, the reluctance to establish new generalised tort principles, and the desire to maintain the authority of Wilsher. As demonstrated above, the House of Lords judicial acceptance of proportional recovery is so narrowly restricted that one cannot identify a general inclination in its favour. The UK statute that reversed Barker reinforces this conclusion.103 This reversal was not intended to act as an opposing stance to proportional recovery as such, rather, it was based on Parliament’s desire to maintain full recovery for victims of mesothelioma. Therefore, despite this recent legislative measure, the approach adopted in Barker still opens the door to a larger debate on the admissibility of proportional risk-based liability. Moreover, since the reasoning in Barker can be paralleled with the French Cour de cassation’s treatment of proportional liability over the past 40 years, the extensive French experience can operate as a remarkable guide in assessing both the validity of the concept, as adopted in Barker, and its related policy implications. It simultaneously serves a cautionary tale against the use of such reasoning. III. THE FRENCH EXPERIENCE WITH PROPORTIONAL RECOVERY Comparisons between Barker and the French case law demonstrate that the French Cour de cassation has previously and systematically accomplished much of what the House of Lords attempted in Barker. In France, proportional damage assessment is premised on the reasoning that a lost chance may constitute a valid head of damage independent of the ultimate injury suffered by the plaintiff, even where such injury has been realised. However, the French approach to proportional recovery appears to be conceptually superior to that of Lord Hoffmann in Barker. Despite a similar redefinition of the injury, French law has managed to 100 At [39]. 101 At [64]. 102 At [64]. 103 Compensation Act 2006. The relevant sections of this Act came into force on December 1, 2006.

The Act allows a return to the Fairchild principle with the result that in the McGhee-Fairchild-Barker mesothelioma situations, defendants will be held jointly and severally liable for any contribution to the claimant’s risk, allowing the latter to recover 100 per cent of his injury against any defendant which is in front of the court (s.3(2)). The defendant(s) providing payment may thereafter seek a finding of contributory negligence or contribution from other responsible persons calculated in proportion to the length of the periods of exposure or on some other basis (s.3(3), (4)).

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

122

Law Quarterly Review

[Vol. 124

maintain a traditional assessment of causation. The Cour de cassation does so by insisting that causation between the defendant’s negligence and the plaintiff’s lost chance of avoiding an injury—or increased risk of suffering an injury—that eventuated be demonstrated. However, as the House of Lords should have considered, critics of the French approach have been emphasising for several decades that affording compensation for this alternative head of damage introduces several conceptual difficulties. The French judicial acceptance of proportional recovery Proportional recovery based on the loss of chance concept is the French Cour de cassation’s preferred tool for dealing with causal difficulties which arise often due to the French law’s stringent standard of proof of requiring of certainty in the causal link.104 Perhaps it is in reaction to the severity of its evidential rules that French law often surmounts causal difficulties by (i) redefining the injury as the loss of a chance; (ii) maintaining the orthodox requirement of causation by insisting that the plaintiff prove with certainty that the defendant’s fault caused the injury, redefined as the lost chance; (iii) restricting the extension of this reasoning by requiring that the ultimate injury has manifested itself and that the chance lost be “real and serious”; and (iv) granting “partial” recovery calculated in proportion to the chance lost. As a result of this redefinition of the plaintiff’s damage as the lost chance of avoiding injury, French law shows no concern for distinguishing between past facts and future eventualities and easily extends this reasoning to cases involving uncertain alternative mechanisms. France’s highest court has applied this reasoning extensively in instances of medical liability and beyond, as a means of addressing difficulties with which plaintiffs contend when proving causation in the face of uncertainty. It has also used the loss of chance technique in cases where the defendant’s fault is characterised either as having destroyed a chance previously held by the plaintiff, as well as, less frequently, when it has increased the risk of injury to which the plaintiff was already subject before the negligent act occurred.105 The Cour de cassation decisions characterise loss of chance as an independent type of damage which results from the loss of an opportunity either to realise a benefit or to avoid an injury. This loss can be compensated if the opportunity’s destruction is found to have been caused by the defendant’s fault.106 The Cour de cassation has applied this 104 G. Viney and P. Jourdain (J. Ghestin, dir.), Traité de droit civil. Les conditions de la responsabilité, 2nd edn (1998), para.362; X. Lagarde, “Finalités et principes du droit de la preuve” J.C.P. 2005.I. p.771 at 777; K.M. Clermont and E. Sherwin, “A Comparative View of Standards of Proof” (2002) 50 Am. J. Comp. L. 243 at 250. 105 See fn.67 above. 106 Couturier, note under Cass civ 1re, 7 June 1989, D. 1991.Jur.158 at 159 (also reported at D. 1991.somm.323, J.C.P. 1989.G.IV.294, D 1991.somm.323). See also, e.g. Cass civ 1re, 18 March 1969

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

123

reasoning both to cases where the injury was of an economic nature and to those where it consisted of bodily injury. It granted recovery for loss of chance in a factual context very similar to that of Hotson as early as December 14, 1965.107 In this case, the lower court’s experts expressed doubts as to whether the plaintiff child could have avoided the injury had the defendant doctor correctly, and in a timely fashion, diagnosed his dislocated elbow. But it was only after drawing a factual presumption of causation between the injury suffered by the child and the defendant’s misdiagnosis that the Cour de cassation used the loss of chance language to characterise the injury suffered. Subsequently, many decisions of the highest court compensated for lost chances, even doing so without addressing, through presumptive reasoning or otherwise, the uncertainty of the causal link between the fault and the ultimate injury suffered in the form of sickness, disability, or death.108 Losing a chance of recovery is characterised as a form of damage in its own right, independent of the final outcome suffered. Although the compensation is a fraction of what would be awarded for the final outcome suffered,109 the redefinition of the damage as a lost chance allows granting full recovery.110 However, the Cour de cassation has insisted that loss of chance can only be used to qualify the injury and rules it out where the evidence demonstrate the absence of a causal link between the fault and the final injury,111 or the certainty of its existence,112 on an application of the ordinary certainty standard.113 Proven absence of causation between the actual injury and the fault is said to automatically exclude causation between and 27 January 1970, J.C.P. 1970.G.II.16422; Cass civ 1re, 25 May 1971 (1re et 2e esp`eces), Bull. Civ. 1971.I.n170, D. 1972.534 (note Penneau), J.C.P. 1971.G.II.16859; Cass civ 1re, 21 November 1978, J.C.P. 1979.G.II.19033 (note Savatier); Cass civ 1re, 2 May 1978, Gaz. Pal. 1978.2.somm.251, J.C.P. 1978.G.II.18966 (note Savatier); Cass civ 1re, 24 March 1981, D. 1981.jur.545 (note Penneau), J.C.P. 1981.G.IV.212, Bull. Civ. 1981.I.n98; Cass civ 1re, 17 November 1982, n 67 above; Cass civ 1re, 8 January 1985, D. 1986.jur.39 (note Penneau); Cass civ 1re, 12 November 1985, Bull. Civ. 1985.I.n299 (Germain v Estragnat) and Bull. Civ. 1985.I.n298, J.C.P. 1986.G.IV.42 (Konstantinow v Manrique); Cass civ 1re, 10 January 1990, fn.67 above. 107 See fn.60 above. See also Grenoble, 24 October 1962, fn.60 above. 108 e.g. Cass civ 1re, 18 March 1969 and 27 January 1970, n 106 above; Cass civ 1re, 25 May 1971 (1re et 2e esp`eces), fn.106 above; Cass civ 1re, 17 November 1970, D. 1971.46; Cass civ 1re, 9 May 1973, J.C.P. 1974.G.II.17643 (note Savatier), Gaz. Pal. 1973.Jur.631 (note Doll), D. 1973.116; Cass civ 1re, 27 March 1973, J.C.P. 1974.G.II.17643 (note Savatier), D. 1973.Jur.595 (note Penneau); Cass civ 1re, 2 May 1978, fn.106 above; Cass civ 1re, 24 March 1981, fn.106 above; Cass civ 1re, 8 January 1985, fn.106 above; Cass civ 1re, 12 November 1985, fn.106 above; Cass civ 1re, 18 January 1989, D. 1989.somm.315; Cass civ 1re, 7 June 1989, fn.106 above; Cass civ 1re, 10 January 1990, fn.67 above); Cass civ 1re, 16 July 1991, J.C.P. 1992.G.II.21947 (note Dornser-Dolivet) (breach of the duty to inform); Cass crim, 20 March 1996, fn.67 above (note Jourdain); Cass civ 1re, 4 November 2003, D. 2004.n9. p601. 109 e.g. Cass civ 1re, 27 March 1973, fn.108 above; Cass civ 1re, 9 May 1973, fn.108 above; Cass civ 1re, 18 January 1989 D. 1989.somm.315. 110 e.g. Cass civ 1re, 1 June 1976, J.C.P. 1976.II.18483 (note Savatier). 111 A. Dorsner-Dolivet, note under Cass civ 1re, 16 July 1991, fn.108 above, at [4] and the following, e.g. Cass civ 1re, 11 October 1988, fn.114 below; Cass civ 1re, 25 May 1971, fn.106 above; Cass civ 1re, 15 June 2004, Juris-Data no.024201. 112 Couturier, note under Cass civ 1re, 7 June 1989, fn.106 above, at 160. 113 See also Bernfeld, fn.146 above.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

124

Law Quarterly Review

[Vol. 124

the fault and the lost chance.114 These decisions can be interpreted as confining the function of loss of chance to its classical role of defining the damage115 and as underscoring that the concept should not be used to bypass the clear absence of causation between the final injury and the defendant’s fault.116 However, distinguishing loss of chance as a head of injury from its use as a means to facilitate the plaintiff’s proof causation is not always straightforward; in fact, redefining head of injury as the lost chance necessarily has the effect of lightening the plaintiff’s burden of proof in demonstrating causation. The French courts have extended the loss of chance reasoning to several areas. For instance, it is used in informed consent cases to afford maximum protection to the patients’ autonomy and right to choose by allowing the patient to be compensated for his lost chances to make a different therapeutic choice or to refuse a procedure.117 This approach allows the French courts to ignore the causal uncertainty about what exactly the plaintiff would have done had the risk been revealed. It is also used more generally to respond to all cases where a professional has breached his obligation to inform.118 Risk-based proportional recovery and the French doctrinal critique By ensuring that the award of damages is calculated in proportion to the increased risk, the chance lost, or the possibility that the defendant actually caused the plaintiff’s injury, the proportional recovery technique overcomes the unfairness of the traditional evidential and causal principles. Victims can obtain damages despite causal uncertainty, and the defendant, held liable only to the extent that he is shown to have contributed to the risk of injury or to have destroyed the plaintiff’s chances, is not forced to shoulder an unmerited burden.119 In the common law, proportional recovery also eliminates the conceptual tension between accepting full recovery based on risk increase and simultaneously resisting the loss of 114 e.g. Cass civ 1re, 2 May 1978, fn.106 above; Cass civ 1re, 17 November 1982, fn.67 above; Cass civ 1re, 12 November 1985, fn.106 above; Cass civ 1re, 11 October 1988, Bull. Civ. 1988.I.n281; Cass civ 1re, 10 January 1990, fn.67 above. 115 Mémeteau, fn.63 above, at 141; G. Mémeteau, “Perte de chance et responsabilité médicale” [1997] Gaz. Pal. 1367 at 1370; Durry (1983), fn.67 above, at 548; C. Müller, La perte d’une chance (2002), para.108. 116 Durry (1983), fn.67 above, at 547; Dorsner-Dolivet and Saluden’s notes under Cass civ 1re, 17 November 1982, fn.67 above. For a more detailed account of the French approach to loss of chance: Khoury, fn.25 above, at Ch.4. 117 e.g. Cass civ 1re, 7 February 1990, Bull.I.n39 at 30; Cass civ 1re, 16 July 1991, fn.108 above; Cass civ 1re, 7 December 2004, Bull.I.n302, p.253; Cass civ 1re, 14 June 2005, Juris-Data no.028962. See also, Müller, fn.115 above, at [429] et seq., particularly cases cited at fn.1740. Cf. Chester v Afshar, fn.85 above. 118 Müller, fn.115 above, para.429. 119 R. Goldberg, “The Role of Scientific Evidence in the Assessment of Causation in Medicinal Product Liability Litigation” in H. Reece (ed.), Law and Science (1998), p.55 at p.59.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

125

chance approach.120 This is particularly hard to reconcile since rejecting compensation for a lost chance arguably constitutes a refusal to grant compensation for the fact that the defendant’s negligence has merely increased the plaintiff’s risk of injury.121 Barker resolves this difficulty by treating proportional recovery as a corollary to liability based on risk increase in cases involving causal uncertainty.122 Besides for these positive impacts, the strongest argument in favour of proportional recovery is the contention that loss of chance or increased risk are valid types of damage, independent of the ultimate loss, i.e. the benefit missed or the injury that was not avoided.123 As Lord Hoffmann noted in Barker, there is a fine line between accepting that increase of risk reasoning may ground a more generous approach to causation, and assimilating increase of risk to a compensable head of damage. Although Barker never developed this argument, defining increase of risk as a type of injury may allow the liability analysis inquiry to fall within the traditional framework, since causation must still be proven between the lost chance or increased risk and the wrongdoing, whether on the balance of probabilities or with certainty.124 Moreover, once this causal link is demonstrated, the loss calculated based on the statistical chance of its realisation is payable in full and there is no reduction for the possibility that the chance might have been lost through the operation of other causes.125 This approach, which has been adopted predominantly in French law, is prima facie compelling because it allows for apparent conceptual validity of the notion and conformity with the rules of civil liability. However, Barker’s characterisation of increased risk as a valid head of injury is not without conceptual problems. It deserves to be assessed in the light of the French civil law’s 40-year-old loss of chance debate. Despite its wide judicial application, the approach that redefines the injury as the lost chance has been extensively critiqued by French commentators.126 Their criticism is predominantly based on the fact that this alternative head of injury is not truly distinct, as its calculation is always contingent on applying the percentage of the lost chance to the 120 Porat and others, fn.60 above. 121 See text accompanying fnn.62–66. 122 This was noticed by Lord Mackay in Hotson v East Berkshire AHA, fn.62 above, at 786. 123 Stapleton, fn.1 above, at 392–393, 396–397; Jansen, fn.63 above, at 274 and 282; Gerecke, fn.61

above, at 802; G. Durry, “Faute médicale et perte de chances de survie” (1972) 70 R.T.D. civ. 408 at 410; A Tunc, “Perte d’une chance de guérison par suite d’une erreur de diagnostic” (1963) 61 R.T.D. civ. 334 at 335; Boré, fn.58 above, at para.38. 124 F. Descorps Declère, “La cohérence de la jurisprudence de la Cour de cassation sur la perte de chance consécutive a` une faute du médecin”, D. 2005.p.742 at 746. 125 Stapleton, fn.1 above, at 396–397. 126 e.g. F. Bouvier, “La responsabilité médicale” Gaz. Pal. 1984.doctr.284 at 288; Penneau, note under Cass civ 1re, 27 March 1973, fn.108 above, at 597; Dorsner-Dolivet, note under Cass civ 1re, 17 November 1982, fn.67 above, at 306; Mémeteau (1986), fn.63 above, at 145.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

126

Law Quarterly Review

[Vol. 124

quantum of damage relating to the final injury.127 It is argued that the final injury, when it occurs, “absorbs” the intermediate head of damage (pr´ejudice interm´ediaire) that loss of chance represents, so that when damages for a lost chance are granted, they constitute, in effect, partial compensation for the actual injury.128 If loss of chance were a truly independent type of injury, defendants would be forced to compensate the plaintiff even for the lost chance that results in no actual injury. Thus, proportional risk- or chance-based recovery can also been criticised for theoretically leading to liability for exposure to risk even in cases in which the risk has not yet eventuated and damage remains speculative.129 This consequence can in practice be avoided by requiring that the final outcome have occurred or that it will occur in the near future.130 However, while such a restriction ensures the rejection of speculative claims, it simultaneously reinforces the arguments of those civil law authors who insist that this head of damage lacks independence. The identification of the exact nature of the interest which the defendant’s negligence has destroyed is also described as problematic. Opponents—as well as some supporters131 —denounce the arbitrary, inaccurate and subjective nature of its assessment.132 Savatier, the most vehement French adversary to accepting loss of chance, has observed that, in the medical context, even the statistical calculation of lost chances is impossible given the “infinite diversity” of medical faults, individual biological states, and medical pathologies.133 Even common law adherents to the loss of chance doctrine admit that it may be impossible, because of uncertainty, to disentangle the background risk to which the plaintiff is already exposed from the additional risk due to the defendant’s wrongdoing.134 French critiques also insist that statistical chances of a disease’s progress and an individual’s distinct chances of survival or

127 P. Jourdain, “Sur la perte d’une chance” [1992] R.T.D. civ. 109 at 114; Penneau, note under Cass civ 1re, 27 March 1973, fn.108 above, at 597. 128 Mémeteau (1986), fn.63 above, at 145; Descorps Declère, fn.124 above, at 747; Penneau, note under Cass civ 1re, 24 March 1981, fn.106 above, at 547. 129 In Québec: S. Philips-Nootens, “La perte de chance: détournement du lien de causalité ou dommage distinct?” (1990) 50 Can. Bar Rev. 611 at 620. 130 As imposed by Cass civ 2e, 3 May 2006, no.05-12.069 (unpublished) and Cass civ 1re, 16 June 1998, D.1998.IR.180. 131 Viney and Jourdain, fn.104 above, at [371] who admit the evaluation is arbitrary. 132 R. Savatier, “La responsabilité médicale en France (Aspects de droit privé)” [1976] R.I.D.C. 493 at 502; R. Savatier, “Une faute peut-elle engendrer la responsabilité d’un dommage sans l’avoir causé?” D. 1970.chr.123 at 123; Savatier (1973), fn.64 above, at 473; Viney and Jourdain, fn.104 above, at para.371; Penneau (1977), fn.58 above, at paras 105–106. In common law, Stapleton also admits that such apportionment will necessarily be artificial in cases where the defendant’s contribution to the total risk is uncertain: Stapleton (2002), fn.34 above, at 284 et seq., 298 et seq. 133 Savatier (1973), fn.64 above, at 473. 134 Stapleton, fn.1 above, at 407.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

127

cure must be distinguished from one another.135 Perhaps wary of this shortcoming, French courts are often vague as to the basis of their evaluation of the lost chance and simply state the quantum of damages without identifying their method of calculation.136 However, some courts require that the experts situate the chance lost on a spectrum ranging from “weak” to “very important”.137 Sometimes, they have simply granted a lump sum, a tendency that has since been condemned by the Cour de cassation in numerous decisions.138 Most importantly, the French experience is a warning of the slippery slope upon which even isolated acceptance of proportional liability lies. French law demonstrates that the approach that redefines the injury as the increased risk of suffering the final outcome, if applied consistently, can potentially lead to a complete revolution in the traditional rules of evidence and civil liability, as has been the case in the French medical malpractice case law and beyond.139 Recognising this problem, some French advocates of the loss of chance approach have attempted to restrict its application to cases involving true uncertainty,140 without, however, defining which cases might qualify as such. One could respond to this argument by stressing that the expansive application of the loss of chance technique was necessary in France because of the greater hurdle French courts face in assessing causation.141 Their demanding standard of proof has had an obvious effect on the judicial responses to causal uncertainty. Because of this certainty requirement, which lacks the flexibility of the more generous common law standard of balance of probabilities,142 the French courts must ground their generosity on something other than their evaluation of the evidence such as on a redefinition of the traditional substantive conditions for liability.143 However, in practice, French judges are often eager to relax 135 Penneau (1977), fn.58 above, para.110; Savatier, note under Cass civ 1re, 2 May 1978, fn.106 above. Against: Saluden, note under Cass civ 17 November 1982, fn.67 above. Accepted in common law in Gregg v Scott, fn.46 above, at [28] (Lord Nicholls), [111] (Lord Hope). 136 S. Porchy, “Lien causal, préjudices réparables et non-respect de la volonté du patient” D. 1998.chr.379. 137 Bernfeld, fn.146 above. 138 Bernfeld, fn.146 above and Cass. civ. 1re, 18 July 2000, Gaz. Pal. 1998.panor.p34; Cass civ 1re, 29 June 2004, Juris-Data no.024471; Cass civ 1re, 28 September 2004, Juris-Data no.025004; Cass civ 2e, 21 December 2006, no.03-20.421 (unpublished). 139 Savatier, fn.132 above, at 124–125; Savatier (1973), fn.64 above, at 476; Dorsner-Dolivet, note under 17 November 1982, fn.67 above, at 307. Also in the Québec doctrine: S. Gaudet, “L’application de la notion de perte de chances en droit médical” (paper presented at the C.I.A.J.’s Conference: “Responsabilité médicale et hospitalière: Aspects juridiques et e´ thiques”, Montréal October 29, 1990), p.7; Nootens, fn.129 above, at 618; I. Parizeau, “Le lien de causalité . . . au bord du gouffre” (1989) 49 Can. Bar Rev. 514 at 514, 517. 140 Descorps Declère, fn.124 above, at 743. 141 Laferrière v Lawson, fn.15 above at, 601–603. 142 Clermont and Sherwin, fn.104 above, at 266. 143 Viney and Jourdain, fn.104 above, para.371. This aim is admitted by G. Durry, “La faute du médecin diminuant les ‘chances de guérison’ du malade” (1969) 67 R.T.D. civ. 797 at 798. See also Gonthier J.’s comments in Laferrière v Lawson, fn.15 above, at 602.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

128

Law Quarterly Review

[Vol. 124

this strict standard of proof and to find that certainty in causation may be presumed on the finding of a “sufficient degree” of probability.144 Causal presumptions based on sufficient probabilities have cohabited with the frequent resort to the loss of chance doctrine. One could also argue that there is no danger of an expansive application of risk-based proportional recovery in UK common law since the option to apply it beyond the McGhee-Fairchild-Barker exceptions is arguably foreclosed by decisions such as Gregg v Scott and Fairchild. Even Barker attempts to restrict its application to cases in which all potential sources, including those not under the defendants’ control, are related to the same risk agent. It is true that what characterises the McGhee-Fairchild-Barker situations is the fact that an element of causation is proven, since the risk agent that brought about the loss is well identified. However, the dividing line between those cases, and the cases in which the alternative causes relate to distinct sources of risk, is extremely thin. Aside from the fear of overextending the doctrine, there is no compelling conceptual justification to strictly maintain this boundary between alternative cause cases where the risk agents operate in substantially the same manner and alternative cause cases in which risk agents are distinct. French courts have, in fact, mainly applied the loss of chance reasoning in cases involving alternative causal mechanisms related to distinct sources of risk which are not all under the control of negligent defendants—frequently in cases involving the negligent misdiagnosis, or negligent delays in the diagnosis, of an underlying disease or injury.145 Several authors recognise that the usefulness of the loss of chance doctrine is in responding to cases where it is impossible to identify whether the cause of the final injury is the doctor’s fault or the underlying pathology from which the patient already suffered and which initially led him to consult the defendant.146 However, Chabas requires some connection between the patient’s pre-existing risk of injury and the risk introduced by the defendant’s negligence. He excludes from the ambit of loss of chance cases in which the defendant’s negligence has created an entirely new source of risk unrelated to the patient underlying condition. To his opinion, where the patient is not already engaged in a negative process in relation to the injury that occurred at the moment the fault was committed, one cannot say that his situation is reduced to only chances of avoiding the injury.147 144 Viney and Jourdain, fn.104 above, paras 347, 362–369; R. Legeais, Les règles de preuve en droit civil (1955), p.174. 145 e.g. in the recent case law: Cass civ 1re, 29 November 2005, Bull.I.n455, p382; Cass civ 1re, 7 February 2006, Juris-Data no.032118. For an application to indeterminate defendants situations: Cass civ 1re, 10 January 1990, Bull. Civ. 1990.I.n10; D. 1991.somm.358 and Savatier (1966), fn.60 above. 146 Y. Lambert-Faivre, Droit du dommage corporel, 5th edn (2004), para.582-2; A. Dorsner-Dolivet, La responsabilité du médecin (2006), p.178; C. Bernfeld, “L’indemnisation de la ‘perte de chance’ dans le domaine des accidents médicaux: une situation dramatique pour les victimes” Gaz. Pal. 2006.doctr.33. 147 F. Chabas, “La perte de chance en droit français” in O. Guillod (ed.), Développements récents du droit de la responsabilité civile (1991). However, critiques observe that the distinction between these two

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

129

Policy, however, may justify the above dichotomy. In the first class of cases where there is only one type of risk agent, the causal link between the injury and the risk agent is proven, thereby demonstrating elements partially fulfilling the factual causation requirement. The uncertainty lies only in assigning the culpable risk agent to one of the defendants. Such partial demonstration of the causal connection in cases involving causal difficulties has inspired Canadian and French law to adopt pro-plaintiff solutions in other circumstances.148 Finally, one may contend that although the strict requirement of proof of causation appears to be maintained, in reality, the qualification of the injury as the lost chance conceals that liability is found despite the uncertainty of a causal link between the defendant’s breach of the standard of care and the actual damage.149 Viney and Jourdain have argued that the determination of loss of a chance has sometimes deviated (détournée) from its usual meaning (as a compensable head of injury) and instead been used to facilitate the plaintiff’s task of proving causation, particularly in medical malpractice cases.150 Using the loss of chance doctrine in this alternate way, the judge can grant partial recovery despite doubts remaining as to the causal link between the negligence and the final and concrete damage.151 Indeed, the case law shows that loss of chance essentially finds application where the causal link between the fault and the actual injury remains uncertain.152 For many French authors, loss of chance, whether defined “artificially”153 as a head of injury or not, has the effect of allowing partial recovery where the existence of causation is uncertain and thus the traditional standard of proof is not met.154 This approach of allowing liability in proportion to the causal role that fault has played in producing the injury had long been rejected in France.155 Savatier has even gone further, arguing that what judges identify as an objective chance is in fact the extent of their subjective uncertainty as to categories of cases may not always be easily drawn and may lead to arbitrary results: Jourdain, fn.127 above, at 111. For those cases that Chabas excludes from the ambit of loss of chance, Viney and Jourdain suggest that full recovery based on the negligent creation of a risk should be granted!: fn.104 above, paras 371 et seq. 148 e.g. the reversal of the burden of proof of causation in the Supreme Court of Canada decision in Cook v Lewis, fn.29 above; presumptions of causation in the French HIV-HCV litigation and other product liability cases: Khoury, fn.25 above, at Ch.5. 149 Penneau, note under Cass civ 1re, 24 March 1981, fn.106 above, at 547. 150 Viney and Jourdain, fn.104 above, para.370. See also Bernfeld, fn.146 above. 151 This is even recognised by its supporters: Stapleton, fn.1 above, at 392; Jansen, fn.63 above, at 283. 152 e.g. Cass civ 1re, 18 March 1969 and 27 January 1970, fn.106 above; Cass civ 1re, 15 June 2004, no.02-11.527 (unpublished); Couturier, note under Cass civ 1re, 7 June 1989, fn.106 above, at 159. 153 This qualifier is from Viney and Jourdain who are not troubled by this fact, however: fn.104 above, paras 370–371. 154 See account of Viney and Jourdain, fn.104 above. 155 B. Starck, “La pluralité des causes de dommage et la responsabilité civile. La vie brève d’une fausse e´ quation : causalité partielle = responsabilité partielle”, J.C.P. 1970.I.2339. There are examples in the French case law where this approach has been admitted, however: fn.60 above.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

130

Law Quarterly Review

[Vol. 124

the cause of the injury.156 The amount of compensation granted can be seen as an expression of the extent of the judge’s uncertainty regarding causation.157 Both the Canadian and French doctrine has widely acknowledged that loss of chance is often used to by-pass the absence of causation; this outcome of the doctrine was also determinant in the Supreme Court of Canada’s rejection of the loss of chance doctrine. Meanwhile, the Cour de cassation has largely continued to ignore it. In that sense, although subject to criticism, Lord Hoffmann’s explicit recognition in Barker, that proportional recovery based on risk is really granted in the absence of causation, deserves some praise. It has the advantage of acknowledging that, if merely increasing the risk of harm serves to satisfy the test for factual causation, or ground liability altogether, “causation has been effectively eliminated as a distinct element of the cause of action in negligence” since “breaching a standard of care consists in subjecting someone to an increased risk of harm”.158 Thus, the House of Lords has arguably been more transparent and honest than the French Cour de cassation by recognising that proportional liability is, in such cases, imposed without proof of causation between the negligence and the actual loss, due, essentially, not to logical conceptual reasons, but to the existence of overriding policy objectives. CONCLUSION Overall, the highest courts of the United Kingdom and Canada have maintained a strong attachment to traditional theory in the area of causation. The result is a general reluctance by the Supreme Court of Canada and the House of Lords to apply a more generous approach to causation or liability, using concepts such as risk increase or loss of chance. Until Resurfice, Canadian law adopted the most rigid attitude, categorically rejecting both risk and chance-based reasoning to support full or proportional recovery of damages. Interestingly, the Supreme Court of Canada has stated these positions most clearly in the context of civil law cases, which have thereafter been treated as authoritative by the lower common law courts. The House of Lords shows more flexibility by accepting both full and proportional recovery for negligently enhanced risk, despite causal uncertainty. However, these instances are squarely 156 Savatier (1973), fn.64 above, at 475. 157 Savatier (1973), fn.64 above, at 473; Tacchini-Laforest, fn.59 above, at 10. This objection is widely

admitted in France and Québec: Viney and Jourdain, fn.104 above, paras 370–371; Boré, fn.58 above, para.2; Jourdain (1992), fn.127 above, at 110; Dorsner-Dolivet, note under Cass civ 1re, 17 November 1982, fn.67 above, at 307; P. le Tourneau and L. Cadiet, Droit de la responsabilité civile (1998), para.876; Against: G. Boyer-Chammard and P. Monzein, La responsabilité médicale (1974), p.101. 158 D. Cheifetz and V. Black, “Material Contribution and Quantum Uncertainty: Hanke v Resurfice Corp” (2006) 43 C.B.L.J. 155 at 160.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Causation and Risk

131

limited to industrial disease cases where a single causal agent is clearly related to the injury and the uncertainty lies only in attributing culpability to one or more of the negligent defendants. Although proportional recovery grounded on negligent risk increase is confined to one House of Lords decision, recently reversed by statute, any attempts to develop this approach as a means to by-pass causal uncertainty should be informed by the extensive French experience and critiques. These critiques have entertained, for several decades, a sophisticated debate of proportional recovery that emphasises several issues relevant to the common law analysis. Criticisms particularly worthy of attention are those that raise the artificially independent nature of the alternative head of damage that lost chances or increased risks constitute; the difficulties in assessing the value of these redefined injuries; and the potential for an application much broader than that illustrated by the McGhee-Fairchild-Barker line of cases. Ultimately, however, the House of Lords also has a lesson for French law in Lord Hoffmann’s recognition of the true result of a wholesale adoption of proportional recovery: the abandonment of the traditional causation requirement. LARA KHOURY.*

* LL.B. (Sherb.), B.C.L. (Oxon.). D.Phil. (Oxon.). Assistant Professor, Faculty of Law and Institute of Comparative Law, McGill University. This text is partly based on Uncertain Causation in Medical Liability (2006). It was presented in June 2006 at the international conference “Emerging Issues in Tort Law” held at the University of Western Ontario Law School. I would like to thank David Cheifetz for his helpful comments on an earlier version of this text, as well as Chana Edelstein and Ava Chisling for their editing assistance. Canada; Causation; France; Proportionality; Risk assessment

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

DECEPTION, MISTAKE AND VITIATION OF THE VICTIM’S CONSENT

I. MISTAKE AND AUTONOMY APPARENTLY autonomous consent to a particular action or transaction in criminal law will be undermined by the discovery that it was given in circumstances of pressure or under a misapprehension. Cases of pressure have, since Olugboja,1 turned on whether the consent amounted to reluctant acquiescence (valid consent) or mere submission (non-consent), on the assumption that it is possible to find such a line on a spectrum of pressure. Cases of misapprehension,2 on the other hand, have always been dealt with by dividing mistakes into categories, with only some categories of mistake being thought to affect consent.3 As a result there is a mismatch between the treatment of pressure cases and the treatment of misapprehension or mistake on the part of the victim, and yet, as Joan McGregor argues, “deception or fraud, similarly to coercion, affects the voluntariness of an agent’s action” and thus “undermin[es] the legitimacy of consent”.4 Not only have cases of mistake been treated differently from cases of pressure, it is also difficult to see why certain categories of mistake have been chosen over others (if indeed this has always been a conscious decision) and where the line is to be drawn in any given case. The resulting category approach as a whole thus tends to give the impression of being based on unpredictable, unarticulated and possibly even unjustifiable assumptions. However, it should not be thought that the correct way to deal with mistakes by the victim is easy or obvious. On the one hand, it is at least arguable that, as a matter of fact, any misapprehension undermines autonomy if consent would not have been given had the consenter known the truth, because the individual’s right to determine their course of action is impaired when they lack this knowledge.5 As Ripstein argues, the giving of consent confirms a right on the part of the consenter because they have exercised a choice between granting and refusing consent6 and 1 [1982] Q.B. 320. 2 The precise meaning of “misapprehension” or “mistake” and their relationship with fraud and

deception will be discussed further below. 3 This point is made by Simon Gardner in “Appreciating Olugboja” [1996] L.S. 275 at 286–291. 4 J. McGregor, Is it Rape? On Acquaintance Rape and Taking Women’s Consent Seriously (2005) at p.181. 5 For a convincing argument along these lines, see J. Herring, “Mistaken Sex” [2005] Crim. L.R. 511. 6 A. Ripstein, Equality, Responsibility and the Law (1999), p.209.

132

JANUARY 2008]

Vitiation of the Victim’s Consent

133

it is inherent in this right that the choice should be an informed one. On the other hand, the practical consequences of allowing all “but-for” factual mistakes to have legal import may not be wholly desirable in policy terms. It is therefore necessary to examine which mistakes, if any, ought to be considered legally operative and how such mistakes should be distinguished from those, if any, which are not. Of course, criminal law is not the only area of law in which such issues arise,7 but the binary guilty/not-guilty nature of criminal law presents the problem in a stark form, with no real possibility of loss-sharing. Equally, there are several reasons why comparisons should not be drawn even between different areas of criminal law, let alone between criminal law and other fields.8 Yet while the policy considerations in each area are different and require nuanced treatment, it is also arguable that at the core of each of these areas is an investigation into the presence of consent. If it is true that any “but-for” mistake undermines consent as a matter of fact, then consideration of mistakes should at least begin in the same way in each area. If other policy considerations or rules are to play a part later they can then do so openly with full explanations of why apparently similar mistakes are being treated differently.9 II. THE EXISTING CRIMINAL LAW It is necessary to begin with an understanding of the current situation and, as noted above, in contrast to the rules dealing with consent under pressure the law has adopted a category-based approach to the rules dealing with mistaken consent. These current rules are largely based on unsupported assertions that mistakes as to certain matters vitiate the victim’s consent 7 Other examples obviously include the impact of mistake on contracts and on informed consent in tort law. 8 See, for example, R. v Cort [2003] EWCA Crim 2149; [2004] Q.B.388, noted by Williams [2004] C.L.J. 271, where Buxton L.J., giving the judgment of the Court of Appeal had “severe reservations about the use of civil law implications of fraud and mistake to assess the effect of the mistakes about the circumstances of sexual intercourse” (at [18]). Or Whittaker v Campbell [1984] 1 Q.B. 318, where Robert Goff L.J. (as he then was) not only warned that “there is. . . danger in assuming that the law adopts a uniform definition of the word ‘consent’ in all its branches” (at 326), he was very careful to ensure that his definition of “consent” should not even be carried across to other areas of the criminal law, stating that instead the word “must be construed in its own particular context” (at 330). 9 And it should equally be noted that there have been several calls for precisely such a comparison to be made. For example, the Law Commission expressed concern that “it would make things extremely difficult for those who have to enforce the law if two quite separate regimes for consent existed in relation to sexual and non-sexual offences against the person”: Law Commission, Consent in Sex Offences: a Report to the Home Office Sex Offences Review, available online at http://www.lawcom.gov.uk/231.htm#lcr271 , at paras 5.16 and 6.10, citing support from the S.P.T.L. as it then was (now the S.L.S.). See also Virgo, “When is Consent not Consent? (When it is Vitiated by Mistake)” (1996) 6 Archbold News 5/7/95. Similarly, in R. v Cuerrier [1998] 2 S.C.R. 371; [1998] 162 D.L.R. (4th) 513, L’Heureux-Dub´e J. specifically disagreed with Cory J. on the basis that the relevant principles should be ‘consistent across the different assault contexts’ and Cory J.’s approach would not achieve that result (para.13), but on the other hand, Cory J. did himself draw an analogy with fraud in property offences (para.135). In Linekar [1995] Q.B. 250 the Court of Appeal drew the more far-reaching analogy between the context of sex offences and “the principles governing mistake vitiating apparent manifestations of will in other chapters of the law.”

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

134

Law Quarterly Review

[Vol. 124

while others do not. The results produced by this technique are arbitrary and indeed are accepted by some to be so.10 This is unsatisfactory in itself; the rule of law requires us to treat like cases alike unless we can distinguish sensibly between them. Arbitrariness would be marginally less problematic if the law were at least able to fulfil the other requirement of the rule of law: legal certainty. But this has hardly been achieved either. The law of consent arises in various contexts within criminal law, yet in each area it has been treated differently. 1. Sex offences In brief, the law on mistaken consent for the purposes of sex offences is now governed by the Sexual Offences Act (“SOA”) 2003, under which a victim consents if “he agrees by choice, and has the freedom and capacity to make that choice” (s.74). Section 75 of the Act contains a series of situations which will raise an evidential presumption against the defendant that the victim did not consent, but all of these concern factors such as force or physical incapacity. For our purposes, the most interesting section is s.76 which contains conclusive presumptions about consent: “(1) If in proceedings for an offence to which this section applies it is proved that the defendant did the relevant act and that any of the circumstances specified in subsection (2) existed, it is to be conclusively presumed— a) that the complainant did not consent to the relevant act, and b) that the defendant did not believe that the complainant consented to the relevant act. (2) The circumstances are that— a) The defendant intentionally deceived the complainant as to the nature or purpose of the relevant act; b) The defendant intentionally induced the complainant to consent to the relevant act by impersonating a person known personally to the complainant.” Thus, if the complainant is deceived by the defendant about the very fact that they are engaging in sexual activity of any kind, as in Flattery 11 and Williams,12 under s.76(2)(a) the complainant will not be consenting. Similarly, the old rules on partner impersonation from s.1(3) of the Sexual Offences Act (SOA) 1956 and Elbekkay 13 remain. This has always 10 G Williams, “Mistake as to Party in the Law of Contract” (1945) 4 Can. Bar Rev. 271 at 282. See also Gardner, above, fn.3, who contrasts the victim-centred approach taken to consent in cases of pressure with the categorised approach used for consent in cases of misapprehension, and Herring, above, fn.5. 11 (1877) L.R. 2 Q.B.D. 410 12 [1923] 1 K.B. 340. 13 [1995] Crim. L.R. 163.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

135

been an odd line to draw. Unlike a misunderstanding about the very nature of the activity, in some cases the identity of the other party may not even be a necessary condition of factual consent to sexual activity in the first place. Yet when it is a necessary condition it will also automatically become legally sufficient to render consent inoperative as a matter of law. However, if the defendant is, unknown to the complainant, HIV positive, this will not vitiate the complainant’s consent to sex under the SOA 2003,14 though it will vitiate their consent to run the risk of infection, as was established in the landmark case of Dica.15 This suggests that the court in that case did not feel the existence of the SOA 2003 gave it the opportunity to widen the rules, but rather crystallised the existing common law rules, or even a narrower version of them. Unusually, given that as a whole the Act has widened criminal liability in the sexual context, s.76 itself does appear to have reduced liability for sexual activity with mistaken consent. One example of this is that s.3 of the SOA 1956 (“procuring a woman by false pretences”) has no direct counterpart in the new Act. In Harvinder Singh Jheeta 16 the victim was caused to believe that she was being stalked and that the police officers involved in her case recommended she have sex with the defendant in order to prevent the defendant from committing suicide. As a result she consented to sex with the defendant when otherwise she would not have done so. In fact the “police officers”, who only ever contacted her by text message, were the defendant himself, as was the supposed stalker, and none of the defendant’s alleged suicide attempts had in fact taken place. The defendant was convicted of two counts of the s.3 offence in relation to his sexual encounters with the victim which took place before April 30, 2004, but in relation to the episodes which took place after the SOA 2003 came into force on May 1, 2004 it was held not to be possible to use the conclusive presumptions in s.76. Ultimately the Court of Appeal held that nevertheless the defendant had been safely convicted of rape on the basis that the victim had not given consent by free choice as required by the general provision of s.74 of the Act. Its reasons for doing so in terms of the general moral culpability of the defendant are obvious, and counsel for the Crown had argued that “a statute which brought together all the offences of a sexual nature cannot have been intended to decriminalise deliberate conduct designed to deceive a woman into having sexual intercourse.”17 14 R. v B [2006] EWCA Crim 2945; [2007] 1 W.L.R. 1567. 15 [2004] Q.B. 1257. 16 [2007] EWCA Crim 1699; [2007] 2 Cr. App. R. 34. 17 At [20].

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

136

Law Quarterly Review

[Vol. 124

Nevertheless, in doing so the Court of Appeal was forced to create a difficult precedent for application in future cases. Unless the element of deception in Jheeta is singled out as distinguishing this case from the traditional “casting-couch” problem,18 the Court of Appeal may have inadvertently set a surprisingly wide limit to the scope of the criminal law. And if the element of deception is indeed singled out in future cases as distinguishing the two situations, this must mean instead that the law now has two sets of deceptions to contend with, those under s.76 and those under s.74 more generally. This in turn sits ill with the Court of Appeal’s refusal to examine the issue of consent to sex in Dica 19 and B.20 Another example of the more restrictive approach of the 1956 Act is that although facts such as those in Flattery, Williams, and Elbekkay now give rise to conclusive presumptions under s.76, this only happens where the defendant has deceived the complainant intentionally. Where the victim was spontaneously mistaken it is now unclear whether the cases can be included in s.76 when the defendant takes advantage of the mistake, which seems unlikely; whether (as seems most likely) such mistakes should be considered as part of a general investigation into consent under s.74; or whether implicitly such cases no longer give rise to criminal liability because they are not explicitly included in the Act, which would be an unnecessary and undesirable conclusion. Conversely, in his article in the Criminal Law Review, Jonathan Herring has argued that the use of the word “purpose” in s.76(2)(a) expands the scope of s.76 beyond the old rules so that it would now include deceptions that the sex was for the purpose of consummating a marriage, as in Papadimitropolos 21 or even for the purpose of making money for the victim, as in Linekar,22 though the Court of Appeal did not seem persuaded that sex for the purpose of stopping the defendant committing suicide was the result of such a deception in Jheeta.23 2. Non-sexual offences against the person As well as dealing differently with cases of HIV transmission, the rules for sexual and non-sexual offences against the person have also diverged in 18 If a victim is threatened with prosecution unless she has sex with the defendant or is offered a benefit (such as an acting job) if she agrees to have sex with the defendant when she would not otherwise have done so, does she consent within the parameters of Olugboja (above, fn.1) or now s.74 or not? For further discussion of this conundrum see, e.g. Simester and Sullivan, Criminal Law Theory and Doctrine, 3rd edn (2007), at pp.428, 434 and 689–691 and Smith and Hogan, Criminal Law, edited by D. Ormerod, 11th edn (2005), at p.600. 19 Above, fn.15. 20 Above, fn.14. 21 (1957) 98 C.L.R. 249. 22 Above, fn.9. 23 Above, fn.16.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

137

their treatment of the facts in Tabassum 24 and Richardson.25 In Tabassum, a woman believed the defendant to be medically qualified when he was not, and his examination of her breasts was therefore held to be indecent assault. Although the case was decided under the old law on sex offences there is no reason to suppose that its result would be different today. Tabassum’s victims could still be considered to lack consent to the nature of his actions under s.74, if not s.76 of the SOA 2003. In contrast, in Richardson, the fact that a dentist had been suspended from practice was held not to vitiate her patients’ consent to treatment for the purposes of non-sexual offences against the person. There are three slight differences between Tabassum and Richardson. First, there is the difference in the type of offence (sexual as opposed to non-sexual), and the influence of this difference was perhaps enhanced by the unconvincing nature of Tabassum’s denial of sexual motive. However, the court in Richardson made it plain that the type of offence at issue should not change the rules in principle,26 and in Dica, although the case concerned a nonsexual offence against the person, the court referred to Tabassum, not to Richardson. Secondly, there is the more important fact that Richardson had at least been trained and had practised as a dentist, while Tabassum had never received sufficient medical training, a fact which could have been used to distinguish the cases but was not. Finally, there is the ground actually used by the Tabassum court to distinguish the cases, namely that Tabassum’s lack of qualifications affected the nature of his act, while the court in Richardson had examined whether her lack of qualifications affected her identity and concluded that it did not, since qualifications were only a matter of attribute. This third supposed difference cannot really be made out, since, as Sir John Smith commented, Tabassum “did not seem distinguishable on its facts from Richardson”.27 Either case could have been decided on the basis of an identity mistake or a mistake as to the nature of the act, so the use of one basis for one case and the other for the other seems a matter of random chance, rather than a principled choice. A further non-sexual offence against the person in which this issue has arisen is that of kidnapping. In Cort,28 the defendant would stop at bus stops and offer lone women lifts to their destinations on the pretence that the bus for which they were waiting had broken down. On two occasions the women accepted his offer. Although handcuffs, condoms, string, a 24 [2000] Cr. App. R. 328. 25 [1999] Q.B. 444. 26 “There is no basis for the proposition that the rules which determine the circumstances in which consent is vitiated can be different according to whether the case is one of sexual assault or one where the assault is non-sexual. The common element in both cases is that they involve an assault, and the question is whether consent has been negatived. It is nowhere suggested that the common law draws such a distinction”, per Otton L.J. above, fn.25 at 450. 27 [2000] Crim. L.R. 686 at 689. 28 Above, fn.8.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

138

Law Quarterly Review

[Vol. 124

knife and tape were recovered from his car, he pleaded that these were there for innocent purposes and the Court of Appeal observed, perhaps surprisingly, that there was no reason to suggest that this was “anything other than truthful and accurate.”29 Kidnapping is a common law offence, defined by Lord Brandon of Oakbrook in R. v D 30 : “. . . the offence contains four ingredients as follows: (1) the taking or carrying away of one person by another; (2) by force or by fraud; (3) without the consent of the person so taken or carried away; and (4) without lawful excuse.”31 In Cort, the appellant argued that since his victims were under no mistake in relation to the act of riding in the car to an intended destination they were thus not mistaken as to the nature of the act or the identity of the defendant, and their consent was therefore valid. Buxton L.J., giving the judgment of the court, held that the law applicable to assault and rape32 need not be applied to the offence of kidnapping. There was thus no need for the prosecution to show that the victims had been mistaken about either the nature of the act or the identity of the defendant. The misapprehension under which they had operated was sufficient in itself and the defendant’s conviction was thus upheld. It therefore appears that in Cort the existence of a deception on the part of the defendant was permitted to expand the categories of mistake by the victim that would be held to invalidate consent as a matter of law. A similar oddity used to occur in the relationship between the old s.3 of the SOA 1956 (“procuring a woman by false pretences”) and the offence of rape under that Act, in the sense that it was possible for a defendant to be convicted of the s.3 “false pretences” crime even if the victim had not been mistaken about the defendant’s identity or the nature and quality of the act, as would have been required for a conviction of rape. As noted above in relation Jheeta,33 the precise scope of s.74 is now less clear on this point. 3. Property offences The issue of victim mistake is also relevant to property offences, and here again the rules have varied. There are several property offences in which this issue has arisen or could do so. For example, in Gilks 34 the defendant’s conviction for theft was upheld when he received money paid to him by a bookmaker, knowing that the 29 At [4]. 30 [1984] A.C. 778. 31 At 800. 32 Even assuming that they are similar to each other, and as noted above this is not necessarily the case. 33 Above, fn.16. 34 [1972] 1 W.L.R. 1341.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

139

bookmaker mistakenly believed that the defendant had backed the winning horse. Victim consent is also relevant to burglary contrary to s.9 of the Theft Act (TA) 1968 in determining whether or not the defendant is a trespasser. One example of this arose in Gallagher,35 where the defendants gained entry to a 94-year-old’s residence by claiming to be an employee of the water board,36 though interestingly there does not seem to have been any discussion in that case of the kinds of mistake that would or would not vitiate the victim’s consent. As a further example, s.12 of the TA 1968 prohibits the taking of a conveyance without the owner’s consent. In Whittaker v Campbell 37 the defendants represented to the director of a vehicle-hire firm that the first defendant was someone who held a full driving licence, which he produced. Acting upon this misrepresentation, the director agreed to their hiring a van. He would not have done so had he known that neither defendant held a full driving licence and would not therefore be insured to drive the van. Nevertheless, Robert Goff L.J., giving the judgment of the court, held that the s.12 offence was not made out38 and his reasoning on this point is essentially the corollary of that used by Buxton L.J. in Cort and inherent in the old distinction between rape and s.3 “procuring a woman by false pretences’. In Whittaker, Robert Goff L.J. took the view that he was not dealing with an offence of deception and thus the consent given, albeit reached through deception, was legally valid. Conversely, in Cort and in s.3 of the SOA 1956, the view taken was that where the offence was one involving fraud, consent would automatically be invalid. Thus although they approach the question from different directions, all three cases share the belief that there is a fundamental distinction between offences of fraud and offences concerning a mere lack of consent. In all three, consent is held to be automatically vitiated when an offence of fraud or deception is made out, regardless of the precise content of the mistake, whereas if lack of consent is simply one of the elements the prosecution has to prove and it tries to do so by adducing evidence of the victim’s mistake, only certain types of mistake will be successful. This conclusion fits with a further widening of the operative categories of mistake in another series of cases concerning deception. In Lambie,39 the defendant used a credit card in a shop after she had exceeded her credit limit and agreed to return the card to the bank. She was convicted 35 [2005] EWCA Crim 389. 36 Virgo had previously suggested that this could be a possibility in a case such as Collins [1973] 1

Q.B. 100 (above, fn.9, at 8), giving the example of a defendant who gains entry to a house by pretending to be a gasman who has come to read the meter and then steals property. 37 Above, fn.8. 38 However, he did point out that he expressed no opinion on the question whether the defendants could instead have been accused of dishonestly obtaining services by deception contrary to s.1(1) of the TA 1978: at 330. 39 [1982] A.C. 449.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

140

Law Quarterly Review

[Vol. 124

of obtaining a pecuniary advantage by deception, contrary to s.16(1) of the TA 1968,40 on the basis that she had evaded a debt by deception through falsely representing that she was authorised to use the credit card to obtain the goods. The previous case of Charles 41 had established that when the drawer of a cheque uses a cheque card, he represents to the payee that he has the actual authority of the bank to enter on its behalf into the contract expressed and that the bank will honour the cheque when it is presented. The problem with applying that case in Lambie was that the shop seemed not to have relied on the defendant’s apparent authority to use the card at all. All that really mattered to the shop was that it should receive payment, and it was bound to do so in any case as a result of its agreement with the bank on the basis of the card. It was argued on behalf of the appellant that the case could therefore be distinguished from Charles because no representation made by the defendant to the shop staff in this case had induced them to complete the transaction. The House of Lords nevertheless held that if the shop assistant had known that the defendant was acting dishonestly and had no authority from the bank to use the card in that way, she would not have completed the transaction. The offence was thus made out, not because the shop assistant had positively relied on her mistaken belief, but because although she had positively relied on the prospect of the shop receiving payment from the bank, negatively she would still not have agreed had she known of the countervailing lack of authority. Sir John Smith in his comment on the later case of Kassim 42 thought this resulted in “an artificially wide meaning of deception”.43 The deception offences contained in ss.15–16 and 20(2) of the TA 1968 have since been repealed by the Fraud Act 2006. This describes three forms of fraud, any one of which will constitute the commission of an offence under s.1 of the Act. Section 2 describes “fraud by false representation”, s.3 describes “fraud by failing to disclose information” and s.4 describes “fraud by abuse of position”. Each form of the offence is made out if the defendant undertakes the relevant form of prohibited behaviour dishonestly, intending through the behaviour to make a gain for himself or another, or to cause a loss to another or to expose another to a risk of loss. None of the three forms of the offence, therefore, actually requires the defendant to cause a loss or make a gain; the actus reus appears to consist solely of the prohibited behaviour in each case.44 40 Now repealed: see Fraud Act 2006 Sch.3. 41 [1977] A.C. 177. 42 [1988] Crim. L.R. 372. 43 [1988] Crim. L.R. at 375, though he argued that as a matter of law those decisions would now

equally apply to the offence of procuring the execution of a valuable security by deception contrary to s.20(2) of the TA 1968, the offence at issue in Kassim itself. 44 It will be argued below, however, that in fact the s.1 fraud offence is an inchoate result crime aimed at preventing the possibility that people will in fact be deceived. Otherwise it is difficult to see what harm is captured by the offence.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

141

Section 11 of the Act, which creates the offence of “obtaining services dishonestly”, is different. A person is only guilty of this offence if he actually “obtains services for himself or another” by a dishonest act. In addition, the services must be made available on the basis that payment has been or will be made in respect of them, the defendant must obtain them without making full payment for them, he must know that they are or might be made available on this basis and he must intend not to pay for them in full. Unlike the s.1 offence, the requirement of a dishonest act under s.11 means that the offence cannot be committed by omission. The offence can cover situations of deception (the Explanatory Notes to the Act give the example of someone who uses false credit card details to obtain a service),45 but it need not do so, since the Explanatory Notes also give the example of someone who climbs over a wall to watch a football match without paying, and someone who attaches a decoder to a television to watch channels without paying for them.46 In this, s.11 is no different from the other offences mentioned above, such as burglary, which equally can be committed by deception but need not be. However, the other changes made by s.1 of the Fraud Act, turning the offences into inchoate result crimes or crimes of behaviour only do sever many of the links between property offences and vitiation of the victim’s consent, since for those offences as noted above it is not necessary that the prosecution show any causal impact on the victim’s consent at all. III. WHAT IS A MISTAKE AND WHAT DOES IT DO TO CONSENT? It is clear, therefore, that the effect of different kinds of mistake on victims’ consent in criminal law varies even within single types of crime, let alone between different areas such as sexual as opposed to non-sexual offences against the person. It is also apparent that in order to make these rules more coherent and consistent it will be necessary to separate out some of the strands contributing to the present tangle. 1. What is a mistake? The first question, though obvious, is also one of the most difficult to answer; what is a mistake? And in particular, how sure must one be that one is right about a fact before one can be considered to be mistaken about it? This issue has arisen more frequently in private law,47 but nevertheless, 45 Fraud Act 2006 Explanatory Notes para.35. 46 Fraud Act 2006 Explanatory Notes paras 35–36. 47 Obviously this is closely linked to the question of the causative operation of the mistake. From the

context of unjust enrichment, in Barclays Bank Ltd v WJ Simms Son & Cooke (Southern) Ltd [1980] Q.B. 677 Robert Goff J. held that the claim may fail where “the payer intends that the payee shall have the money at all events, whether the fact be true or false, or is deemed in law so to intend” and in Maskell v Horner [1915] 3 K.B. 106 the claimant could not recover because he had paid to avoid litigation despite

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

142

Law Quarterly Review

[Vol. 124

doubt or uncertainty about a particular state of affairs can arise equally in criminal law. The case of Konzani 48 concerned a defendant who was HIV positive and failed to reveal that fact to his sexual partners. The focus of the case was on the defendant’s mens rea, but the court also discussed the nature of the victim’s mistake and held that: “there may be circumstances in which it would be open to the jury to infer that, notwithstanding that the defendant . . . concealed his condition from the complainant, she may nevertheless have given an informed consent to the risk of contracting the H.I.V. virus. By way of example, an individual with H.I.V. may develop a sexual relationship with someone who knew him while he was in hospital, receiving treatment for the condition. If so, her informed consent, if it were indeed informed, would remain a defence, to be disproved by the prosecution”49 From this passage it appears that some partial knowledge on the part of the victim about the fact of infection may be compatible with a state of mistake, since knowing of a person’s treatment for HIV apparently does not automatically mean that consent to sex with that person was informed. The point was left open for decision in other cases and is not crucial to what follows, but may nevertheless have to be tackled in the future. Arguably, once the victim is aware enough of the facts to be accepting some of the risk it becomes difficult to suggest that he or she is still labouring under a mistake. 2. The effect of mistake on consent Secondly, there is disagreement over whether the effect of a mistake by the victim is indeed to “vitiate” or to “negate” the victim’s consent, or

being in doubt as to his liability to pay. In Kleinwort Benson Ltd v Lincoln CC [1999] 2 A.C. 349, Lord Hope of Craighead held that: “A state of doubt is different from that of mistake. A person who pays when in doubt takes the risk that he may be wrong”, dicta which have more recently been applied in Brennan v Bolt Burdon [2004] EWCA Civ 1017; [2005] Q.B. 303. Burrows believes that the claim should only fail if the claimant thinks the facts are probably as they are in truth (Law of Restitution, 2nd edn (2002), p.140), whereas Virgo would exclude recovery whenever the claimant was aware that there was a possibility that he or she was mistaken (Principles of the Law of Restitution, 2nd edn (2006), p.163). See also the first instance decision in Deutsche Morgan Grenfell Group Plc v IRC [2003] EWHC 1779 (Ch); [2003] 4 All E.R. 645, which may have suggested that a level of uncertainty was not incompatible with mistake, though in the Court of Appeal [2005] EWCA Civ 78; [2006] Ch 243 Rix and Buxton L.JJ. preferred Lord Hope’s remarks in Lincoln to the effect that a state of doubt is different from that of mistake. The majority of the House of Lords approved the decision at first instance essentially on the facts, finding that Park J. had accepted evidence that the claimant was not in doubt whether or not a state of doubt could in theory be compatible with making a mistake: [2006] UKHL 49; [2007] 1 A.C. 558, esp. Lord Hoffmann at [29]; but Lord Brown of Eaton-under-Heywood preferred the approach of Rix and Buxton L.JJ. in the Court of Appeal. 48 [2005] EWCA Crim 706; [2005] 2 Cr. App. R. 14 49 At [44], emphasis added.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

143

whether the consent remains but is qualified in some way.50 It was argued above that in principle all “but-for” mistakes should be regarded as having a negative impact on factual consent because all such mistakes deny the victim the opportunity to give autonomous, informed consent. What we are now trying to establish is whether all such factual mistakes should also invalidate consent as a matter of law. Arguably it therefore does not matter whether we perceive this invalidation as “vitiation”, “negation” or “absence” of consent. The point is simply to ask whether the consent apparently given in fact was legally operative or not. 3. Mistake or deception? Similarly, there is also disagreement over whether it is a mistake by the victim that gives rise to criminal liability, or rather the inducement of that mistake by the defendant. As noted above, on three occasions (Cort, Whittaker v Campbell and the old distinction between procuring a woman by false pretences and rape) the law seems to have suggested that cases of deception belonged in a different category from cases of simple vitiation of consent through mistake, so that while only certain mistakes would remove consent in the latter category, all mistakes caused by the defendant would remove consent where the offence was one of deception. In a similar vein, the conclusive presumptions under s.76 of the SOA 2003 only arise where the defendant has “intentionally deceived” or “intentionally induced the complainant to consent”. It is submitted that this contrast between deception and mistake is wrong. It is difficult to see why the law wishes to discourage deception unless it is because deception of any kind reduces the victim’s autonomy and ability to consent, whether that consent is to a lift in a car, to a financial transaction or to some kind of physical contact. This is what Buxton L.J. must have meant when he argued that there was “no room” for “consent” “when one is dealing with a case of fraud”.51 Deception cannot therefore be contrasted with or seen as an alternative to lack of consent, because that lack of consent is one of its three constituent elements, the other two being some level of causal contribution to the mistake by the defendant and some form of mens rea concerning that causation. In the light of the Fraud Act 2006, on the other hand, as noted above, the same may no longer be true of “fraud”. Indeed, it might initially be thought that the new offence created by s.1 of that Act runs counter to the argument here, because the defendant can be guilty of that offence if he makes a false representation, fails to disclose information or abuses his 50 For example, in Flattery, above, fn.11 and Linekar, above, fn.9, the courts referred to a need for an “absence of consent” by the victim to the defendant’s actions; but in Tabassum, above, fn.24, the court held that there was “no true consent” (at 337, emphasis added) and in Richardson, above fn.25, Otton L.J. referred to “vitiation” of consent (at 450). 51 Above, fn.8, at [22], emphasis added.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

144

Law Quarterly Review

[Vol. 124

position intending to produce a gain for himself or a loss to another, even if no such gain or loss is actually made. In other words, the offence is made out even if the victim’s consent is not vitiated at all, and the victim hands over the relevant property voluntarily. Vitiation of consent cannot therefore be one of the constituent elements of fraud, unlike deception. If that is the case, the new Fraud Act offences must either be inchoate result offences52 or behaviour offences. In either case it is difficult to see how false representations, failure to disclose information and abuse of position could lead to criminal liability without any impact on the victim’s consent unless the concern is that there could have been such an impact and the defendant thus posed a potential threat. In that sense there is no difference in principle between the harm captured by deception and that captured by fraud, and the argument here remains intact. Both concepts are concerned to protect the freedom to make full, informed, autonomous choices, but while “deception” under the previous Theft Act only allowed the law to intervene when there had actually been a causative impact on the victim’s consent,53 “Fraud” is, like other inchoate offences, concerned to prevent even the possibility of this causative impact. It is true that fraud is wider than an offence of attempted deception (though it would certainly cover cases previously charged as such) but it is arguable that all its forms are concerned in one way or another to allow the law to intervene to prevent people from dealing with property without the owner’s fully informed consent.54 If all this is correct we can now reject the reasoning behind Cort, Whittaker v Campbell and s.3 of the old SOA 1956. The question is whether we must also reject their results. Can the existence of those two other elements; causation of the mistake by the defendant with some 52 The Explanatory Notes to the Act essentially support this conclusion when, in contrast to their comments on the s.1 offence they provide that “the [s.11] offence is not inchoate: it requires the actual obtaining of the service” (above, fn.45, para.35). See further, D. Ormerod, “The Fraud Act 2006—Criminalising Lying?” [2007] Crim. L.R. 193. 53 Charles, Lambie, above, fnn.41 and 39, respectively. Interestingly, deception under s.76 of the SOA 2003 does not require this impact in fact, but rather the existence of deception concerning one of the two key issues on the part of the defendant raises a conclusive presumption that the victim was not consenting and that the defendant lacked a reasonable belief in consent. This seems to fall somewhere between the new meaning of “fraud” and the old meaning of “deception”. 54 For example, one of the purposes of the reforms was to capture two difficult fact situations; “phishing” (falsely disseminating an email to large groups of people to elicit credit card and bank account numbers which are then used by the defendant to make or obtain money transfers) and machines. Machines could not be deceived for the purposes of the previous offences (Davies v Flackett [1973] R.T.R. 8, DC: see J. Chapman, “Can a Computer be Deceived?” (2000) 64 J. Crim. Law 89) but it is possible to “defraud” them under the new offence. This raises the difficult question of the relationship between mistake and ignorance which the law of unjust enrichment and restitution has long found to be controversial (see, e.g. Goff and Jones, Law of Restitution, 6th edn (2002), p180; J. Swadling, “A Claim in Restitution?” [1996] L.M.C.L.Q. 63, and compare Burrows, Law of Restitution, above, fn.47, Ch.4; Birks, Introduction to the Law of Restitution, rev. edn (1989), pp 140–146; and Virgo, Principles of the Law of Restitution, above, fn.47, at pp.139–140 and 131). This is not the place to revisit that discussion, but it is not impossible that a situation in which a defendant “tricks” a machine by inserting a false coin so that he obtains goods from the machine without their owner’s knowledge, is on the same spectrum as a situation in which he tricks the owner himself with the same false coin.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

145

form of mens rea, widen the categories of mistake that the law recognises as invalidating consent? Deception by the defendant certainly seems to provide a policy reason for giving legal effect to a factual “but-for” mistake, and yet if we are concerned with the kinds or types of mistake that should be legally relevant, then it is policy factors relating to that issue that should be considered. We should not confuse matters by introducing policy factors relating to the source of the mistake. This has the added advantage of ensuring that we treat like victims alike. If all “but-for” mistakes vitiate consent as a matter of fact, it becomes difficult to argue that a mistake made by victim A vitiated her consent as a matter of law because it was induced by the defendant, whereas the same mistake made by victim B did not because it was induced by a third party. The defendant’s added culpability in cases of deception might thus be better dealt with instead by the rules of mens rea or by exacerbating their liability in some way. For this reason the approach of s.76 of the SOA 2003 is not quite as problematic as Cort, Whittaker v Campbell or s.3 of the SOA 1956. It is true that s.76 also focuses on the defendant’s deception or inducement of consent, but it does so only to create a conclusive presumption in certain instances, not to determine whether those instances can be legally relevant at all. If a victim has simply been mistaken about the nature and purpose of the act or mistakenly believes that the defendant is someone known personally to her, this fact could still be relevant in determining whether she has “agreed by choice with the freedom and capacity to make that choice” as required by s.74 of the SOA 2003 and nothing in s.76 prevents this.55 The only difference made by the lack of deception is that the prosecution will have to make out the victim’s mistake and the defendant’s mens rea on the evidence, rather than being able to rely on the conclusive presumptions of s.76. In this sense s.76 is only a means of exacerbating liability, not of determining it altogether. Under Cort, Whittaker v Campbell and s.3 of the 1956 on the other hand, mistakes were considered irrelevant altogether unless they were produced by the defendant’s deception of the victim. Nevertheless, the initial similarity between s.76 and these three other instances does create the risk that the same confusion may occur under the 2003 Act and this would be undesirable. 4. “Active” and “passive” mistakes; “reasons” and “conditions”; positive and negative mistakes A fourth preliminary consideration concerns the distinction sometimes drawn between what Graham Virgo refers to as “active” and “passive” 55 Indeed it is possible that the approach taken in Jheeta, see above, fn.16, strengthens this conclusion.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

146

Law Quarterly Review

[Vol. 124

mistakes.56 A passive mistake occurs where someone acts in ignorance of certain facts or has forgotten them, whereas an active mistake occurs where someone relies consciously on a fact at the time of action. However, given that the line between acts and omissions is notoriously difficult to draw, it is submitted that a better distinction (if indeed a distinction must be made at all) is instead between positive and negative mistakes, or what Honor´e calls “reasons” and “conditions”.57 If the victim makes a positive mistake, or acts for a “reason”, they are acting solely because a particular fact exists. For example, V might consent to a particular physical activity because V believes it to be a necessary part of medical treatment. In fact “reasons” are not the only kinds of positive mistakes; motives and purposes could also be included. The key point is that they can all be grouped together as factors that positively motivate the decision. Conversely, a negative mistake occurs where the victim acts for one or more positive reasons, but subject to a crucial negative caveat. For example, V may find X physically attractive and may thus be positively motivated to consent to sex, yet had V known that the person in question was HIV positive this condition would have acted negatively to prevent V from consenting. There is no question that in such clear-cut cases the line is easy to draw; if asked why she had sex with X, V would not reply “because he was HIV negative”. The most she could say would be that she would not have had sex with him had she known he was HIV positive. However, given other sets of facts, the line may be more difficult to draw. When I consent to medical treatment is my reason, my positive belief, that the person treating me is a doctor? Or is my positive belief simply that the treatment will improve my physical wellbeing, in which case my requirement that the practitioner be qualified only acts as a negative caveat to my consent? If this question is difficult to answer it does not mean that the distinction does not exist in principle, but it does mean that it may be difficult to draw on the facts of any given case. In addition to this, it is hard to argue that positive mistakes universally and inherently have any greater effect on the victim’s autonomy than negative, or vice versa. This in turn suggests that there may be good reasons not to accord the distinction any significance in law. Having thus separated out these different strands, the next stage is to re-synthesise them in a way that produces more coherent and predictable results than can be found in the current law.

56 Virgo, above, fn.47, at pp.145–146. 57 Honor´e, “Necessary and Sufficient Conditions in Tort Law”, Ch.5 of Responsibility and Fault (1999),

at pp.118–119, first published in Philosophical Foundations of Tort Law, edited by Owen (1995), at pp.363–385.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

147

IV. ABANDON THE CONCERN WITH MISTAKES AND VITIATION OF CONSENT ALTOGETHER One option would be to adopt the approach of the Fraud Act 2006 which, as noted above, allows the s.1 offence be made out even if there is in fact no causative impact on the victim’s consent. Nevertheless, it was seen that the new offence of fraud under s.1 of the Act has not abandoned its focus on consent, but has just increased protection for consent by taking the form of an inchoate or behavioural crime. Inchoate liability is, without doubt, necessary in some forms. However, it is still the exception rather than the rule of criminal liability and thus it does require strong justification. Creation of a risk of harm is not the same as creation of the harm itself and perhaps we should, therefore, hesitate before imposing such a wide form of liability too readily. Indeed, if the aim of the new Act was to capture “phishing” offences and “deception” of machines58 there may have been other ways to do this which would not have involved merging the deception offences with other forms of liability or with their own inchoate forms, as the s.1 offence now does. And even if we are happy with this inchoate or behavioural approach in the property offence context we might, for example, be more uneasy about an offence of “dishonestly making a false representation intending to cause a person to consent to sex”. Even the old s.3 offence of “procuring a woman by false pretences” did at least require that the woman in question be “procured’. On the other hand the Fraud Act is not the only statute to have adopted this approach.59 The question whether the criminal law should in general focus on the prohibition of harm or on the prosecution of certain kinds of behaviour is an important but much bigger one, beyond the scope of the current inquiry, and it is on that much bigger issue that the choice of approach must ultimately turn. Nevertheless, it is in the meantime possible to conclude that we should not necessarily abandon concern with the vitiation of consent through mistake altogether. If this is the case, we must have a means of deciding which “but-for” factual mistakes should be legally operative. V. THE SPECTRUM OF ALTERNATIVE LINES The answer is that there is a series of options that can be placed a long a spectrum which ranges from being pro-defendant at one end (because very few mistakes will be found to have a legal effect on consent) to 58 See above, fn.54. 59 See, for example, the offence of forgery under s.1 of the Forgery and Counterfeiting Act 1981: “A

person is guilty of forgery if he makes a false instrument, with the intention that he or another shall use it to induce somebody to accept it as genuine, and by reason of so accepting it to do or not to do some act to his own or any other person’s prejudice.”

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

148

Law Quarterly Review 1. All “But-for” causative mistakes

2. “But-for” positive mistakes only

3. “physical difference” mistakes

[Vol. 124

4. “Non est factum” mistakes only

5. No mistakes have a legal effect on consent

Pro-victim

Pro-defendant Improved policy balance

Greater rationality

FIGURE 1:

Greater rationality

The spectrum of alternative tests

being pro-victim at the other (because a large number of mistakes will have this effect) (see fig.1 above). Towards either end of this spectrum the lines drawn are better able to fulfil the rule of law in the sense that the reasoning behind the tests are readily understood and once the facts of a particular case have been determined it is relatively easy to decide whether the mistake made passes the test in question or not. This makes it easier to treat like cases alike and different cases differently, and the results of any particular test are more coherent and predictable. However, a reduction of arbitrariness is not the only important consideration. The chosen test must also produce desirable results in policy terms, and tests that are too pro-defendant or pro-victim in their application may not be best placed to do this. It is therefore necessary to determine the extent to which each potential test can fulfil these two requirements. 1. “But-for” causative mistakes Beginning at the most pro-victim end of the spectrum, consent could be legally absent wherever the prosecution can prove that the victim would not have consented to or performed the act in question without relying on something which turned out not to be true, whatever the source of that erroneous belief (deception or mistake) and regardless of whether the mistake is positive or negative. There is much to be said for this alternative to the existing rules.60 As Simon Gardner has noted,61 there are no a priori limitations on the types of pressure that will negate victim consent and a “but-for” test for mistakes would thus align them with pressure cases. In both, the concern would be to protect the victim’s autonomy and it is relatively straightforward to argue that there has been an interference with this autonomy whenever the prosecution can show that had there been no misapprehension the victim would have acted differently. No assumptions about admissible subjectmatter are required and it is not necessary to draw any fine distinctions between different types of mistake. 60 See, e.g. Herring, above, fn.5. 61 Above, fn.3.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

149

However, this test is not entirely without drawbacks. In return for rule of law-fulfilling rationality it could result in a wide expansion of liability,62 which may not be desirable for policy reasons. Several of these drawbacks are outlined and answered by Herring,63 including the fact that under this rule a failure to inform a new partner of a past change of gender could lead to a charge of rape. Herring accepts that this is “an invasion of their private life”, but goes on to suggest that “that right must be subservient to the right to sexual integrity of their partner.”64 It is certainly important that, as he states, the mistake should not have to be “to an issue which would be regarded as material to the reasonable person, if it was a prerequisite to consent for the particular victim.”65 To introduce such a rule would be to reintroduce arbitrary assumptions of the kind we already have. Nevertheless, it is possible to argue that in his transgender example the relevant fact does at least relate directly to the physicality of intercourse. But if Herring’s model were to be adopted it would incorporate mistakes which did not. In Cuerrier the Supreme Court of Canada deliberately chose not to set the rules so widely that they would encompass someone who “lied . . . as to [his] position of responsibility . . . or the level of his salary; or the degree of his wealth; or that he would never look at or consider another sexual partner; or as to the extent of his affection for the other party; or as to his sexual prowess.”66 Such lies might be “immoral and reprehensible”, but the court concluded that they should not result in criminal conviction. It is not difficult to see why. We might well feel less comfortable about striking Herring’s balance between privacy and sexual integrity where, for example, D practises a particular religion and fails to disclose this fact to her partner. In a case like that, where the condition placed on the act by the victim is wholly unrelated to the physical act per se and potentially socially undesirable, it is submitted that we should not readily give it legal effect.67 Connected to this problem is the role of mens rea. If “but-for” causative mistake negates consent so easily, a lot of weight will be placed on the defendant’s mens rea as the deciding factor for liability. Under the SOA 2003 this will now be fulfilled by a lack of reasonable belief in consent, so we are left with a situation in which the treatment of the victim 62 Subject to the presence of mens rea, which is discussed in more detail below. 63 Above, fn.5 at 520-523. 64 Above, fn.5, at 523. 65 Above, fn.5, at 517. 66 Above, fn.9, at 563. 67 Of course there are really two examples contained in this one; the presumably more socially acceptable situation where the victim’s own religious beliefs require him to have relationships only with people of the same faith and the presumably more socially unacceptable situation where the victim has no particular religious beliefs of his own but is, for example, anti-Semitic.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

150

Law Quarterly Review

[Vol. 124

of sex offences is wholly subjective while treatment of the defendant is objective. Moral decisions may therefore have to be made by juries about whether defendants ought to have known of their victims’ beliefs or prejudices. In probably the majority of pressure cases, the existence of pressure and its effect on the victim will be externally observable by the reasonable person and it will be relatively straightforward for the jury to assess the defendant’s failure to reach this standard. The assessment of reasonableness in mistake cases, on the other hand, will be much harder. Here the defendant must ask himself what factors the victim is likely to consider important to the particular situation and must establish whether the victim is in possession of all the relevant details about those factors. It will then be up to the jury to assess both whether the defendant’s mental list of relevant factors was reasonable and whether he should have taken further steps to clarify those factors with the victim. And it is not immediately apparent, for example, whether this reasonableness should be based on the public at large—is it ever reasonable to proceed without informing one’s sexual partners of one’s religious beliefs? Or whether it follows the more usual victim-centred approach—given that D knows X is anti-Semitic should D inform X that D is Jewish before having sex with X? At present, the “reasonable belief” mens rea requirement only applies under the SOA and possibly to non-sexual offences against the person in the light of Konzani,68 but in principle these arguments are not offencedependent. If reasonableness were to become relevant to the mens rea of property offences we would encounter the same problems there. In the light of these considerations it thus appears that while the “but-for” line fulfils the requirement of being coherent and predictable in principle, it may be difficult to apply in practice and it may entail policy consequences of an undesirable kind. 2. All “but-for” positive mistakes should have a legal effect on consent, but not negative mistakes about conditions This option would begin in the same way as the option above but would draw a line based on Honor´e’s distinction between “reasons” and “conditions’, or between “positive” and “negative” mistakes. Under this rule Lambie 69 could not have been decided as it was. If the victim’s reason for accepting a credit card in payment was solely that they believed payment would follow, then as long as that payment does follow the transaction is valid and consent is not inoperative simply because the defendant was not entitled to use the card. This entitlement is only a 68 Above, fn.48. See, e.g. M. Weait, “Knowledge, Autonomy and Consent: R. v Konzani” [2005] Crim. L.R. 763, esp. at 770. 69 Above, fn.39.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

151

condition placed by the victim on the transaction. If Herring is right about the potential for expansion of relevant mistakes using the word “purpose” in s.7 of the SOA 2003 the line that is drawn by the law in future might move towards this, since “purpose” mistakes are positive mistakes but probably not negative “conditions”.70 Certainly this would fit with the result in Dica,71 since it would be very difficult to regard the act in that case as “sex for the purpose of not transmitting HIV”. However, while a line of this kind can sensibly be drawn, it is more difficult than the straightforward “but-for” line. As noted above, it may be very difficult in some cases to decide which kind of mistake, positive or negative, has been made. Taking Herring’s example of sex “for the purpose of consummating a marriage”, it is equally possible to characterise such an act as “sex because the parties found each other attractive, on condition, from the victim’s point of view, that they were already married”. It is also an undesirable line to draw in policy terms. For example, it is difficult to see why sex “for the purpose of consummating a marriage” should really be separated from all later acts of sex between the same parties. The victim’s consent cannot really be said to be any less undermined on later occasions where the existence of a marriage is “only” a condition. As noted above, it is not immediately obvious that the infringement of autonomy present in Dica is any less serious than the infringement of autonomy caused by a positive mistake about the purpose or reason for the action. It thus seems difficult to support a distinction of this kind between positive and negative mistakes either on grounds of practical predictability and ease of distinction or on policy grounds. 3. The line (nearly) drawn by the current case law on offences against the person Moving further along the spectrum it is submitted that the next line, drawn by almost all the criminal law on offences against the person (sexual and non-sexual), seeks to establish whether anything physical about the case would have been different if the act had been as the victim thought it was. This could, for example, explain the line between Tabassum 72 and Richardson,73 since Richardson was presumably treating her patients the way she had always done before she was suspended from practice, whereas evidence of a specialist nurse in Tabassum suggested that several of his actions had been unnecessary and would not have been performed by a qualified practitioner seeking to compile a database of the kind he 70 Though as noted earlier, there may be other kinds of positive mistakes such as motives or reasons that do not necessarily constitute purposes. 71 Above, fn.15. 72 Above, fn.24. 73 Above, fn.25.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

152

Law Quarterly Review

[Vol. 124

described. This rule would also provide a means of prosecuting cases like Dica 74 while exempting cases such as Linekar,75 since in Linekar nothing would have been different about the act itself had the prostitute received payment afterwards, whereas sex resulting in the transmission of HIV could be said to be physically different from sex with an HIV negative person. Of course this was not the line used in Dica, since the very brief consideration given by the Court of Appeal to this matter concluded that consent to the sex had not been vitiated in that case, only consent to the transmission of HIV. This could be for one of two reasons. The first is that suggested above, that in the light of the recent enactment of the SOA 2003 without any provision for disease transmission cases, the Court of Appeal felt that it would not be possible to develop the law itself, especially when Dica had not been charged under the SOA.76 The second could be the fact that the Court of Appeal saw transmission of HIV as a physical act separate from the intercourse itself. As a matter of fact, the victim’s sexual autonomy is undermined just as much as their physical autonomy by the failure to disclose HIV positive status in any circumstance where they make a “but-for” mistake about the defendant’s HIV status. As a matter of law, on the other hand, by separating the sex from the transmission the Court of Appeal seems to draw a different line, to be discussed in the next section. The “physical difference” line, if correctly drawn by the law in all future cases, falls fairly centrally in the spectrum and thus provides a reasonable balance of rationality and desirable results. However, it cannot meet the challenge of “qualification cases”. From the property sphere this includes cases such as Lambie 77 where a person is not entitled to use a credit card yet does so. From the sphere of offences against the person it includes cases where the defendant lacks a particular qualification but (unlike Tabassum) acts in exactly the same way as if that qualification had been present. One example of this is Papadimitropolos,78 where the victim had sex with the defendant believing herself to be validly married to him. Another is the Canadian case of Bolduc and Bird 79 in which a physician allowed his friend to observe an internal examination for his own gratification when the patient had only agreed to this on the basis of a false belief that the friend was a medical intern. At no time did 74 Above, fn.15. 75 Above, fn.9. 76 Of course the Court of Appeal in Dica, above, fn.15, was developing the law on offences against the person without positive encouragement from Parliament (as indeed the House of Lords had earlier in Ireland and Burstow [1998] A.C. 147), but since there had been no implicit exclusion of this issue from recent legislation in the area of non-sexual offences against the person the Court of Appeal may have felt more secure in developing the law in this direction itself. 77 Above, fn.39. 78 Above, fn.21. 79 [1967] S.C.R. 677; (1967) 63 D.L.R. (2d) 82.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

153

the friend touch the patient. A third example could be Richardson,80 but this case is slightly different again, since although Richardson lacked the qualifications at the time she acted and thus falls into the category, she had at least held them in the past, unlike Papadimitropolos and Bird. On the one hand, nothing in these qualification cases is physically different from the way it would have been had the victim’s belief been correct. The sex, treatments and payments are unchanged and thus would not be caught by this test, yet on the other hand it is still arguable that the victim’s autonomy has been undermined through misapprehension. The “physical difference” test, therefore fulfils the requirement of being coherent and predictable, but it may not strike the correct balance in policy terms if we are indeed concerned to prosecute the “qualification cases”. 4. “Non est factum” mistakes only The fourth category along the spectrum borrows terminology from the contractual context, where it applies to “those who are permanently or temporarily unable through no fault of their own to have without explanation any real understanding of the purport of a particular document, whether that be from defective education, illness or innate incapacity.”81 Two criminal equivalents of the contract cases are Flattery 82 and Williams 83 where the victims were not even aware that they were engaging in a sexual activity at all. To give examples based on other kinds of crime, this rule would also catch a case in which V consents to D taking a box, not knowing that D has placed a valuable item of V’s in the box, or where V consents to being hit with a polystyrene stage weapon, not knowing that it is in fact a real, heavy one. Conversely, it was because there was no such mistake in the factum that Dica was thought not to be guilty of rape.84 His victims “consented to sexual intercourse”85 even if they did not thereby consent to the transmission of HIV. Wertheimer has argued that the distinction based on mistakes in the factum cannot actually be drawn since “everything turns on the way in which a case is described” and “it is not even clear how the factum/inducement distinction is meant to work”.86 Thus if A lies to B 80 Above, fn.25. 81 per Lord Reid in Gallie v Lee [1971] A.C. 1004 at 1016. 82 Above, fn.11. 83 Above, fn.12. 84 The point was not at issue in Dica itself, but was confirmed in B, above, fn.14. 85 Above, fn.15, per Judge L.J., at 39. 86 A. Wertheimer, Consent to Sexual Relations (2003), at p.206. See also J. McGregor, above, fn.4, at 185: “one way to defeat the in factum/in the inducement distinction is to build more into the description of what is consented to, thereby making all in the inducement cases into in factum.”

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

154

Law Quarterly Review

[Vol. 124

about the mileage of a car, Wertheimer argues that B could be said either to be mistaken in the inducement/attributes (B knows he is buying a car but is mistaken about the mileage) or the factum (buying a car with a particular mileage). But in response it should be noted that in both cases, unlike the victims in Flattery and Williams, B does at least know he is buying a car. B’s belief about the mileage of the car may be a “but-for” cause of his consent to the deal as a whole; it may even constitute a “physical difference” of the kind described above, but B is not mistaken as to the whole kind of activity he is undertaking. Indeed, in the context of sex offences, Posner has conjectured that if a woman is not “averse to having sex with a particular man, the wrong, if any, is in the lies . . . rather than in an invasion of her bodily integrity”.87 This is to go too far in the opposite direction. It is not possible to separate the two issues in this way; the woman’s bodily integrity is, as a matter of fact, undermined by the lies even if they do not lead to a non est factum mistake. Nevertheless, Posner is right to identify the difference between knowing that one is undertaking an activity at all and knowing the details of that activity. There are some cases in which mileage of cars or identity of partners will be a matter of concern to the potential victim, but there will also be cases in which they will not. Mistakes about mileage or identity cannot therefore be separated rationally from mistakes about any other feature of the activity. However, it can be said that no victim can consent to an activity if they do not even know it is taking place. It would therefore be possible to draw a coherent line that allowed only mistakes of this kind to prevent consent from operating, and had the law thus confined itself to Flattery and Williams it would have been defensible in terms of rationality. Where Wertheimer’s argument does carry force, however, is in pointing out that this increased rationality has a high price in policy terms. It is not so much that “everything turns on the way in which a case is described”, but rather the second half of this sentence that is crucial: “there is no reason to think that the intercourse/non-intercourse distinction is the only plausible factual basis on which to distinguish one case from another . . . Why assume that fraud in the inducement is not a serious matter?”88 Is it really desirable to draw the line so tightly, given that many “but-for” factual mistakes would not be legally recognised under such a rule and thus many cases in which the victims did not consent as a matter of fact would have no legal consequences for the defendants?

87 R.A. Posner, Sex and Reason (1992), at p.392. 88 Above, fn.86.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

155

It is worth noting that the law has in practice answered “no” to this question, allowing in cases where people know that they are having sex but are mistaken about the identity of the person with whom they are having it (Elbekkay 89 ). The resulting law may be incoherent and unpredictable but this is not because the courts have chosen to draw a wider boundary in the first place. Rather it is because having decided to give legal effect to mistakes which do not go to the factum the courts have not then found a satisfactory way to distinguish cases like Elbekkay from those like Dica. 5. Consent cannot be prevented from operating by any kind of mistake This most pro-defendant option is clearly rational but utterly undesirable in policy terms. Autonomy can, as argued at the outset, be undermined by misapprehension as much as by pressure. Obviously this choice would be slightly improved if there were to be an equivalent of the old s.3 “procurement by false pretences” offence, but this would still leave a gap where the victim’s autonomy had been undermined and their consent negated but not through the use of false pretences on the part of the defendant. Furthermore, as explained earlier, since a lack of consent is one of the three constituent elements of deception it would not make sense to recognise some mistakes as having legal effects for the purpose of a s.3 style offence when they would not equally have effect for the purposes of a pure lack of consent offence. Conclusion on the spectrum of alternatives The conclusion to be drawn from this spectrum of options must therefore be that even if a change to a universal test could be effected, it is very difficult to draw a line that will both operate rationally and produce desirable results in policy terms. Towards either end of the spectrum the lines drawn increase in coherence and predictability but decrease in desirability, being either too pro-victim or too pro-defendant. In the centre of the spectrum more desirable results are achieved in terms of policy, but the “physical difference” line almost drawn by the current law fails to provide a satisfactory answer to “qualification cases”. 6. A return to the “categories” approach? Perhaps, then, instead of trying to find a single test capable of providing all the answers, we should instead return to the “categories” approach to mistake cases with which we began. This existing approach has been found to have two major problems; first that it is out of line with the wholly 89 Above, fn.13; and see also the old definition of rape under the SOA 1956 as amended by the Criminal Justice and Public Order Act 1991 s.1(3): “A man also commits rape if he induces a married woman to have sexual intercourse with him by impersonating her husband.” See now s.76 of the SOA 2003.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

156

Law Quarterly Review

[Vol. 124

subjective approach used in cases of pressure90 and secondly that it is currently based on unarticulated and unjustifiable assumptions. However, it is submitted that neither of these problems is insoluble. First, it is true that Olugboja 91 seeks to establish a spectrum of pressure, on which the key line is to be drawn between “reluctant acquiescence” (consent) and “mere submission” (non-consent), but that line is notoriously difficult to draw.92 The interesting point is that in order to help juries to find the line, s.75 of the SOA 2003 has itself adopted a “categories” approach, listing certain circumstances in which evidential presumptions will be raised against the defendant. Whether or not the particular categories listed there are desirable in their own terms, their existence in the pressure context means that if they were to be introduced to cope with instances of mistake the difference between the two areas would no longer be so great. If this is accepted, then the first objection falls away and we are left with the second. The real problem here, however, is not the use of categories per se, but the use of finite, rigid, conclusive categories based on unarticulated and unjustifiable assumptions. Unlike the existing categories of mistake, any new categories must therefore fulfil three requirements. First, they must be rational. It must be possible to describe the categories in such a way that the reason for their existence can be easily understood. Thus, for example, we might include “non est factum” mistakes as one category on the basis that, as explained above, a person cannot be said to be consenting to an activity if they do not even know that it is taking place. Secondly, they must be predictable in the sense that it must be possible to ascertain in advance whether a particular case will fall into a given category or not. Thus at present one cannot, in the light of Tabassum 93 and Richardson 94 predict whether qualifications will be seen as going to “the nature of the act” or “the identity of the defendant’. However, it is possible to predict that if V knew she was having sex with D but did not know he was HIV positive she would not fall into the non est factum category (though she may fall into another, such as the “physical difference” category). Thirdly, the lines drawn must produce acceptable results in policy terms. Thus we may reject a category of all “but-for” mistakes on the basis that it produces undesirable results in the case of anti-Semitic victims, or we may create a category of “qualification mistakes” on the basis that we wish to protect the institution of marriage or the requirements of professional training. Of course, there will be differences of opinion about whether or not the 90 See Gardner, above, fn.3. 91 Above, fn.1. 92 See, for example, Simester and Sullivan’s scenario involving an illegal immigrant who is forced to work in the sex industry (above, fn.18, at p.413) 93 Above, fn.24. 94 Above, fn.25.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victim’s Consent

157

categories chosen do strike the right balance in practice, but the key point is that this discussion must take place openly and not be hidden behind apparently self-applying concepts such as “nature and quality of the act”. In case this conclusion is thought to be too open-ended, my opening bid is that we might wish to regard consent as inoperative where any of the following three kinds of mistakes have been made: (a) Non est factum mistakes While it may very well be undesirable to confine vitiating mistakes to this category alone, that does not mean that its identification as a category is useless. It can simply form one category of relevant mistakes that is readily identifiable, distinguishable and justifiable. It is likely that mistakes of this kind will only be negative, “condition” mistakes in the sense that if the victims were to be asked why they were consenting to the supposed “operations” they would not reply “because it is not sexual intercourse”, rather they would never have consented if they had known that it was. But as explained earlier, since both kinds of mistakes are capable of undermining autonomous consent there is no reason in principle to distinguish between them. (b) Physical difference mistakes Again, while this test cannot do all the work alone it is useful as far as it goes and mistakes of this kind should prevent consent from operating. This would include cases of disease transmission and also “whole identity” cases, where the victim mistakes the defendant for another person altogether, real or fictional, known personally or not. It would not, however, include qualification cases, or cases where the victim mistakenly believes the defendant has one particular attribute which would not make a physical difference to the action or transaction itself.95 Again, there is no reason in this category to distinguish between positive “reason” mistakes and negative “condition” mistakes; either kind should remove consent. (c) Legal qualification cases This may be an instance in which a particular categorical line does have to be drawn so that only cases of legal characteristics (such as marriage, entitlements to use credit cards or to practise) are included. It would not then matter whether the qualification in question was the defining characteristic or positive reason for the interchange (I do not care who my dentist is, or about any of his other characteristics, as long as he is a trained dentist) or simply a negative condition placed upon it (I go to my dentist because I have been pleased with his treatment of me in the past 95 For example, where V has sex with D believing that he owns a yacht.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

158

Law Quarterly Review

[Vol. 124

but would stop going were I to discover that he was not qualified to treat me). In policy terms, as mentioned above, such a category would support the protection given to such institutions elsewhere in law.96 VI. CONCLUSION It is almost impossible in this area to set out a completely watertight set of rules which will be certain, coherent and consistent in their operation and produce universally acceptable practical results. Certainly the existing law is a long way from achieving these goals. Like cases are not treated alike, even within the sphere of criminal law; many of the rules governing the present operation of mistakes are based on rigid, unarticulated assumptions and thus whether the victim’s autonomy is protected can be both a matter of chance and relatively unpredictable, as the law grapples with the meaning of words such as “nature’, “quality” and “attribute”. It has been suggested in the past that a test or spectrum approach is the correct way of dealing with this problem, since that would avoid the establishment of arbitrary categories and allows cases of mistake to be aligned with those in which consent is removed by pressure. However, it has been shown here that it is not possible to devise a test to replace the current rules because no single test can provide an acceptable balance between rationality and policy outcome. A range of tests exists, but towards the pro-defendant and pro-victim ends of the spectrum the tests, although predictable and coherent, are either over or under-inclusive in outcome. In the centre of the spectrum a better policy outcome can be achieved but there is no consistent or coherent means of dealing with the “qualification cases”. It has therefore been suggested here that we should instead revert to a category approach. It has not been that approach itself that has been at fault in the past, but rather the inflexibility of its operation and a failure to analyse and articulate the reasoning behind the choice of particular categories. In order to improve on this situation the law should therefore proceed on a case-by-case basis, articulating the similarities and distinctions between cases and the reasons for allowing certain categories of mistake to operate. Any category chosen must be readily understood; the reasoning behind it must be coherent; once the facts are found it must be relatively easy to establish whether or not a given case falls within it or not, so that its results are predictable, and it must not cause undesirable results in policy terms. It may well be that as an individual category it is then under-inclusive of the cases we think it desirable to catch, but that is not a problem as long as, in combination, the set of categories taken 96 For example, the offence of bigamy under s.57 of the Offences Against the Person Act 1861; the offence of impersonating a Police Officer under s.90 of the Police Act 1996, or pretending to be registered [as a medical practitioner] under s.49 of the Medical Act 1983.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Vitiation of the Victimâ&#x20AC;&#x2122;s Consent

159

as a whole are thought to cover the right ground. One such combination has been suggested above, comprising non est factum, physical difference and legal qualification cases, but there may be others who would take a different view. The most important point is that the correct criteria be used to choose the categories, since that will itself lead to a more predictable, coherent and desirable set of practical results than has been achieved so far. REBECCA WILLIAMS.*

* Fellow of Pembroke College, Oxford. This article is based on a paper given at the S.L.S. conference at the University of Strathclyde in 2005. I am very grateful for the feedback I received there and also for the helpful comments of Andrew Ashworth, Paul Roberts, Graham Virgo, Jo Miles, Matthew Weait, Jonathan Herring and John Armour, but of course any remaining errors are entirely my own. Consent; Deception; Mistake; Sexual offences; Victims

(2008) 124 L.Q.R., JANUARY ď&#x203A;&#x2122; SWEET & MAXWELL

AND

CONTRIBUTORS

REVIEWS AND NOTICES The Paradox of Constitutionalism: Constituent Power and Constitutional Form. Edited By MARTIN LOUGHLIN and NEIL WALKER. [Oxford: Oxford University Press. 2007. viii + 375 pp. Hardback: £50]. PARADOXES are rare creatures. A paradox consists of two or more apparently true propositions which generate a conclusion that appears to be false. Faced with a paradox, we have three options: we can dispute the premises, dispute the deduction, or embrace the conclusion. In their introduction to this collection, Martin Loughlin and Neil Walker assert that a paradox lurks within constitutionalism. First, they claim, constitutionalism asserts that governmental power is generated from the consent of the people. Secondly, they claim, constitutionalism asserts that governmental power will only be effective and sustained if it is split between institutions and constrained by rules. These propositions lead to the paradoxical conclusion that whilst the people are sovereign, they cannot exercise their sovereignty: they can never, or can no longer, exercise that power on which the existence and validity of the constitution is thought to turn. This supposed paradox provides a theme which runs through the book, with an array of writers responding to the challenges it presents in different ways. The first limb of the paradox makes an assertion about the source of governmental power. In the first part of the book the contributors employ the tools of intellectual history to tackle the origins and significance of this claim, scrutinising the debates that have surrounded the creation of various constitutional orders. John P. McCormick delves farthest into history, supplying a captivating discussion of Machiavelli’s Florence which he contrasts with contemporary civic republicanism. In a similar vein, Martin Loughlin provides an account of the constitutional debates in England during the seventeenth century, ambitiously contending that some of our modern constitutional confusions are the result of the obfuscation of the founding principles of this period. Lucien Jaume and Christoph M¨ollers consider French and German constitutional history respectively, whilst Stephen Griffin meditates broadly on constitutional change in America. All of these authors provide useful and stimulating papers, but it is hard not to feel that the tight focus on intellectual history leaves aspects of our paradox unexplored. One way of reading the first limb of the paradox is as a factual assertion: that governmental power actually does come from, in some sense, the people. It would have been interesting to see how the intellectual arguments interacted with the actual mechanisms devised to link the people with their constitution. The processes by which contemporary constitutions, such as the South African 160

JANUARY 2008]

Reviews and Notices

161

constitution and the ill-fated European Constitution, were drafted might have illuminated this aspect of the conundrum. Ulrich Preuss’ paper on the exercise of constituent power in Central and Eastern Europe in the 1990s looks as if it might tackle this task, but quickly turns to more abstract matters. Intellectual history to one side, consideration of these modern constitutional processes might have clarified the possibilities for, and limitations of, mass involvement in constitutional drafting. A number of other papers move away from intellectual history and directly address the implicit normative assertion embodied in the first limb of the paradox. Ulrich Preuss and Neil Walker, amongst others, draw a tight connection between constituent power and popular sovereignty. Preuss spends sometime wrestling with the mysteries of popular sovereignty: how to define the group that wields it, and what to do about those within the group who dissent from the majority. One worry with popular sovereignty is that its test of legitimacy may be so exacting that no actual political institution could ever meet its demands. The strain it creates becomes obvious when writers attempt to square international organisations with its requirements. Neil Walker’s powerful paper on the European Union provides a careful reflection on these issues. Walker argues that challenges presented by constituent power are now faced by the Union as much as by the Member States, and that the Union’s legitimacy depends on its connection with the people of Europe. More ambitiously still, Bardo Fassbender argues that the UN Charter “must” be referred to as the “constitution of the international community” (p.281). Fassbender goes on to argue that the requirement for popular participation is satisfied indirectly, through the democratic processes of the nation states. Fassbender’s paper provides a good example of the possible trap set by popular sovereignty. If we assume that the UN is, to an extent at least, authoritative and worthy of our support, and if we further assume that the first limb of the paradox is correct, we are led to reason backwards, and conclude that the power of the UN must be grounded in the consent of the people. Perhaps the resulting gymnastics suggest that we should reconsider our support of the first limb? One notable omission from The Paradox of Constitutionalism is a defence of the assertion that constitutional legitimacy is grounded in the consent of the people. If there are other ways in which a constitution might be legitimated, perhaps, for instance, through the benefits it confers on those subject to it, the force of the paradox will be reduced. The second limb claimed that constituent power is shut out of operating constitutions, excluded to ensure their continuing effectiveness and stability. Many of the contributors to the volume challenge this assertion, arguing that constituent power provides a resource for functioning constitutions, enabling them to change and respond to their citizens. Walker, amongst others, contends that the idea of constituent power (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

162

The Law Quarterly Review

[Vol. 124

demands an on-going connection between people and constitutions: it is not merely a claim about how a constitution should be founded, it also speaks to how the constitution should develop and operate. In an interesting division of opinion, some contributors split constituent power from popular sovereignty, and instead present it as the raw political power to make constitutional changes without constitutional authority. For Damian Chalmers, constituent power is exerted when the formal institutions of the constitution act outside of the constitutional powers allocated to them in response to the wishes of the political community. Similarly, Stephen Tierney points to extra-constitutional referendums and assemblies which can facilitate constitutional reform when the formal mechanisms for change prove insufficiently flexible. Here, constituent power helps preserve the institutions and structures of the constitution by allowing it to meet new challenges. James Tully, in gloomy contrast, reflects on the West’s imposition of constitutional forms on non-Western peoples. Tully identifies three forms of constituent power—political, economic and military—through which imperialism is accomplished, and a particular model of constitution enforced. Emilios Christodoulidis and David Dyzenhaus focus on the paradox’s conclusion, but draw different lessons from it. For Christodoulidis, the paradox is something to rejoice in, providing the continuing possibility of a challenge to capitalist forms of constitutionalism. In a stimulating paper, Dyzenhaus argues that the paradox demonstrates that normative legal theory requires us to deny the existence of constituent power. He argues that this brand of theory demands that authority within the constitution be constituted by law. There is no space left, therefore, for the extraconstitutional power which constituent power purports to be. Dyzenhaus’ case is at its strongest if we suppose that normative legal theory is exclusively concerned with judicial decision-making. The law cannot permit unlawful action—because such permission would render the act lawful—and so, perhaps, it also cannot acknowledge the possibility of unconstitutional constitutional change. At least two difficulties face Dyzenhaus. First, it is just not obvious why legal theory—normative or otherwise—should be limited in the way he prescribes. One of the many tasks of legal theory is to help us understand how legal systems change. An account that denied the possibility of unconstitutional change would be forced to ignore an important feature of the practice it purported to study. Secondly, even if we are narrowly concerned with judicial decision-making, there may still be a need for recognition of unconstitutional change. Whilst such actions can not be prospectively authorised by the constitution, they may be retrospectively endorsed or excused by the judges. Hidden within the volume we can also identify a second, and perhaps even more intriguing, apparent paradox. This relates to the temporal division implicit in the idea of constituent power: constituent power (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

163

is exercised when a constitution is created—or changed through extraconstitutional means—and yet we can only identify an exertion of this power by reference to its product. So, constituent power only exists once it has been successfully exercised, and can only be identified retrospectively; it can, by definition, never be shown to be present in the current constitution. This may help explain the difficulties which some of the contributors have encountered in identifying the holders of constituent power, and determining the conditions and manner of its use. This conundrum is the subject of Hans Lindahl’s paper. His paper is as demanding as it is insightful. Lindahl considers the interplay between the attribution of acts to social groups and their identity. He presents constituent power as a continuing redefinition of the people in the light of their constitution; a constant re-creation of the founding of the constitution. Loughlin and Walker have edited a fine collection of essays. The topic of constituent power has been, until now, under-explored. The essays are of a generally high quality, and will spur further reflection on the issues they raise. N.W. BARBER.* The Refugee in International Law. By GUY GOODWIN-GILL and JANE MCADAM. [3rd Edition. Oxford: Oxford University Press. 2007. xii + 840pp. Hardback: £80]. THE very title of this account of refugee law makes the obvious but often overlooked point that asylum law and practice is mediated through the normative acts of sovereign states but is, in origin and evolution, part of public international law. It is a commonplace within that academic discipline that the International Court of Justice does not enjoy an extensive jurisdiction. It has had so little an impact on the development of refugee law that Art.38 of the Refugee Convention, which assigns a role to the court in the settlement of disputes between the states party to the Convention, that the authors need not have included, and did not, include a reference to this provision in the table of international instruments. The practitioner seeking judicial opinion on the meaning of, for example, the key terms of “persecution” and non-refoulement must refer to decisions of national courts although in the future the European Court of Justice (ECJ) will be able to interpret recent Directives of the European Council in what may be called the reworking of the Refugee Convention in its application to Member States of the EU. The book last appeared in its second edition a decade ago. The new edition is divided into three parts which reflect the authors wish to locate the concept, indeed the persons and individuality, of refugees as the centre * Trinity College, Oxford.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

164

The Law Quarterly Review

[Vol. 124

piece of this work. First, there is a discussion of the relationship between the legal status of the refugee in international law and the political reality of states wishing to retain their obligations of non-refoulement but, at the same time, taking measures to prevent refugees from arriving on their frontiers. The chapter moves on to consider the responsibility of the UN Security Council in maintaining international peace and security when the acts of states may have created a refugee crisis—examples are given of the invasion of Kuwait by Iraq in 1991 and the more internalised conflict in Kosovo in 1998. But the authors make their view clear that the interpretation by national courts of the Refugee Convention as containing a principle of surrogacy of protection is inimical to the true interests of the refugee. A number of recent decisions in the House of Lords are criticised (pp.10–11) for adding a requirement of proof of the inability or unwillingness of a state to offer protection against the risk of serious harm to the risk of such harm occurring whether from the state or from non-state agents. But it is difficult to see how a person can be classified as a refugee if the state can and does (always a difficult issue of evidence) offer a sufficiency of protection against the acts of non-state agents. The book then addresses the central question of who is a refugee. The writers trace the concept as a term of international law from instruments initially arising from the League of Nations and later from the United Nations. But the book does not constantly look backwards. It introduces the EU Qualifications Directive at an early juncture. Internally displaced persons, those in a related category fleeing from civil war or disturbance and those seeking protection from natural disasters are all examined with copious reference to the sad reality that organised hostilities have continued for most of the last century and they show no sign of abating today. Indeed, the category of terrorism and the dangers it poses to human rights receives separate treatment within a longer section on the loss and denial of refugee status. This is the most appropriate moment to refer to the very interesting discussion of the relationship between the Refugee Convention and the European Convention on Human Rights (ECHR). The Refugee Convention contains well-developed exclusion provisions in Art.F. Yet when interpreting the ECHR, almost an exact contemporary of the Refugee Convention, the Strasbourg Court has consistently held that the prohibition in Art.3 of torture and inhuman and degrading treatment is absolute. It applies as much to expulsion cases as to claims against a signatory state in relation to conduct on the territory of that Member State. This prohibition also generates protection for individuals irrespective of their personal conduct. The leading authorities are Soering v United Kingdom (1989) 11 E.H.R.R. 439 and Chahal v United Kingdom (1996) 23 E.H.R.R. 413. Litigation continues before the court about the principles set out in these two cases. The authors express the view that the court is unlikely to resile (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

165

from its principled stance. There is continuing debate within the United Kingdom about withdrawal from the ECHR and its replacement by a national measure that neither contains nor adopts the absolute form of Art.3. Professor Goodwin-Gill has played a significant role in developing submissions by interveners in the continuing European litigation. Although clearly supportive of the maintenance of the absolute nature of protection under Art.3 both authors do not adopt a partisan approach when explaining the contrary position adopted by the United Kingdom and some other Member States of the ECHR. They set out the existing law, express their view that the court is unlikely to modify its existing stance and interweave their treatment of the European Union Qualification Directive (Council Dir.2004/83). This part of the book is amongst the best treatment of a provision that has only been required to have been implemented by October 2006. Instead of merely expounding the text of the law (a common habit of legal writers at both national and international level) the authors grapple head on with the consequences of the instrument having been negotiated at a political level and the difficulties this may create in future references to the ECJ from within Member States. Another feature of modern state practice concerns the policies of, notably, the United States and Australia in intercepting maritime traffic in order to prevent the arrival of potential claimants for asylum. The central paradox of the Refugee Convention is exposed. The principle of non-refoulement does not imply, still less expressly require, an obligation to admit an asylum seeker in order to pursue that very claim. When discussing the Roma Rights decision of the House of Lords ([2005] 2 A.C. 1), the authors skillfully set out how far customary international law contains a principle of admission at least for temporary purposes of asylum seekers in order that their needs for protection can be assessed. The book goes on to consider the nature of the obligation, if any, to provide a procedure for the adequate determination of refugee status. Here the discussion ranges over the omission in the Refugee Convention of any right to a particular method of the assessment of claims to the procedure specified as a minimum level of protection in the European Procedures Directive (EU Council Dir.2005/85) and how each of these strands of developing international law may impose increasing obligations on states. The lacunae in the Refugee Convention are gradually being made good. The purpose of sound procedures is, of course, to ensure that the provisions in the Refugee Convention are fairly and accurately applied. This leads to a discussion of the various strands of persecution envisaged by the Refugee Convention and their development in the case law of Member States. The most interesting ground of persecution and the one that is most likely to expand is that of a “social group” (Art.A2). There is detailed reference to UK jurisprudence and to that of other common law (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

166

The Law Quarterly Review

[Vol. 124

jurisdictions. After Fornah v Secretary of State for the Home Department [2006] UKHL 46 this area of protection appears likely to lead to more claims. The text is supported by a fulsome set of international instruments and a thorough bibliography, very many footnoted references to other books, articles in learned journals and the case law of many countries. It should be at the reach of any practitioner in asylum and human rights. It ought also to inform decisions by the Home Office on initial asylum claims. It is also essential for the specialised immigration and asylum judiciary and for those assembling an appellate case in the Court of Appeal or the House of Lords. Human rights law is an increasing component of law degree courses in the United Kingdom. Academic lawyers will find the book invaluable. It is perhaps as well that the authors lay out in the early part of their work their conceptual approach and their general sympathy for those who challenge adverse state decisions. An exposition of the law which contains no hint of the authorial perspective would leave the reader with a sense of unrequited curiosity. CHARLES BLAKE.* Foundations of Private Law: Property, Tort, Contract, Unjust Enrichment. By JAMES GORDLEY. [Oxford: Oxford University Press. 2006. x + 481pp. Hardback: £60]. THE foundations of the philosophy of modern private law can be traced to the Greeks, particularly Aristotle (Nichomachean Ethics, especially book V). For centuries, however, famous writers in the Aristotelian tradition, such as Thomas Aquinas, were concerned with problems of moral philosophy rather than legal philosophy. But in the sixteenth and seventeenth centuries well known Dutch, Spanish and German writers such as Grotius and Vinnius, Lessius, and Pufendorf, began focussing on legal philosophy and particularly natural law. Although some modern North American writers such as Coleman, Perry, Weinrib and Wright now follow this tradition, a powerful counter-revolution exists in North America, particularly the United States. This is the school of law and economics. In contrast with the philosophy of private law, the foundations of private law doctrine can be traced to the Romans, particularly Gaius (Institutes) whose writing was a powerful influence upon Justinian’s Institutes and Digest. Although common law and civilian systems diverged in the Middle Ages, as English scholars and judges broke free from the constraints of the forms of action they borrowed heavily from Civilian thinking, rooted in * Solicitor and Immigration Judge, London.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

167

Roman law. Perhaps due to increasing recognition of this shared history, modern doctrinal scholarship in England (although less so in the United States) has an increasing comparative focus on Civilian law. In a monumental work of scholarship, Gordley sets for himself the task of placing modern private law doctrine in the context of both its philosophical and doctrinal roots and defending an Aristotelian account of the modern law, particularly from incursions by law and economics scholars who conceive the underlying matrix of the law as economic principle divorced from ethics. Gordley’s task requires him to consider several millennia of philosophical and doctrinal legal history. His study encompasses property, tort, contract and unjust enrichment. It spans from the principles of Roman law to principles of English, French, German and US law with occasional glances over to modern divergences in Australian or Canadian law. In short, he sets himself the task of a lifetime’s work; that is, the lifetimes of at least a dozen philosophers and doctrinal legal scholars combined. Whilst Gordley’s book is a phenomenal work of scholarship, the nature of his task constrains the book in two important ways. Both constraints arise because of the book’s Roman structure. Gaius wrote that the law could be divided into persons, things and actions (G I.8). Things in this broad sense, were later subdivided into property (“things” in a narrow sense) and obligations (primarily contract and delict: G III.88). The adoption of this Roman structure accords with Civilian law and the usual categories in which common law is taught. But it conceals the fact that property rights can arise as a result of the same event that generates personal obligations. An obvious example is the rights to servitudes considered in detail by Gordley. These are often the creations of contract. A more controversial example is the legal category of unjust enrichment which is considered in Chs 19–21. Suppose, labouring under a serious mistake, a claimant makes a gift of his prized racehorse to the defendant. Title to the horse passes to the defendant at common law. But the defendant has been unjustly enriched at the claimant’s expense. He must make restitution of its value. Why should not there be a proprietary response to this unjust enrichment? Gordley relies heavily on the German principle of unjust enrichment in §812(1) of the BGB that one who has received something at another’s expense “is obligated to give it back”. The Aristotelian conception of corrective justice comfortably underlies such statements. Yet, in such a case, surely the best measure of restitution, most consonant with corrective justice, is the return of the racehorse rather than merely its value? For a Romanesque structure which separates the law of property from the law of obligations, such a question was concealed. Although later Roman law described different types of ownership (“. . . But afterwards ownership was made divisible”: G II.40) this was really just a way of implicitly recognising relative title in a system which (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

168

The Law Quarterly Review

[Vol. 124

claimed that ownership was absolute and property was completely distinct from obligations. English law is not so constrained. Cases have suggested different possible answers: a proprietary power to re-vest title by the act of the claimant or order of the court or beneficial rights under a trust. The book’s second constraint is related to the first. The absence of discussion by Gordley of the trust is a gap in his consideration of the historical development of English and North American private law. Indeed, the story of the development of the trust and its effect on private law is an enormously important addendum to the detailed discussion of rights that Gordley presents in other areas. The genius of the trust was that it was an institution which gave a third party (the sui juris, fully entitled beneficiary) not merely personal rights against a trustee, and not merely a proprietary power to vest legal rights held by the trustee, but also a type of proprietary subsidiary right, that is a right in relation to the trustee’s rights themselves, exigible against third parties. In the context of the enormous scope of the book these constraints are miniscule. Further, although the essence of the book is what the English would describe as common law rather than equity, many of the legal issues in equity are identical. For example, the discussion of torts focuses upon wrongs such as negligence, trespass (to person and property) and nuisance rather than the wrongs developed in the courts of Chancery like breach of trust or fiduciary duty (the equitable wrong of breach of confidence is discussed in the context of a developing common law tort of infringement of privacy). But many problems of theory are the same at common law and in equity, just as they are common to common law and Civilian jurisdictions. Gordley attacks those problems head on. For instance, in relation to wrongdoing, he observes that the structure of the common law of torts owes much to the treatise writers of the eighteenth and nineteenth centuries who borrowed from Civilian ideas. German law combined two different ideas. First, a defendant must infringe a legally protected right of the claimant’s. Secondly, that infringement must occur either negligently or intentionally (see §823(1) BGB). An exception is the residual provision, §826, which covers losses intentionally caused contra bonos mores (which includes many of the torts known in English law as “economic torts” as well as torts such as malicious falsehood and deceit). In contrast, after 1932, the common law treated the two ideas as creating concurrent actions. The tort of “negligence” became a pervasive action despite the fact that it did not, even obliquely, protect a particular right. The same story could be told, to a lesser degree, of the development of equitable wrongs. And the same problems arise, such as the limited recovery of “pure economic loss”, both at common law and in equity. How is such a restriction to be justified? Even if there is no particular right to economic profit, why is there no general right not to be subjected to economic losses as a result of the fault of another? Why do jurisdictions such as Germany (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

169

and England have general rules excluding liability for pure economic loss but not France, Italy and the Netherlands? (The United States allows recovery generally where the economic loss is caused wilfully but not negligently.) In a chapter devoted to this problem, Gordley argues that the rule became entrenched for reasons which commended themselves to nineteenth-century conceptualists, but which no longer exist. However, he suggests that for reasons of commutative justice (notions that one person should not “subsidise another’s vulnerability”) the rule can be defended, subject to exceptions. In the instance of torts described above, as with property, contract and unjust enrichment, Gordley’s explanations are based on a mix of rightsbased considerations and economic, utilitarian theory. For philosophical purists who consider only one or the other to be legitimate justificatory theories this will be an anathema. But Gordley’s opening chapters provide a powerful defence for this mix which, as he observes, William James described derisively as “common sense made pedantic” (p.8). This short review cannot do justice either to the careful explanation of his philosophical methodology or to the succession of powerful and jolting insights the book contains. Few will agree with all of Gordley’s conclusions but none will leave the book unchallenged. His carefully presented arguments come thick and fast: for instance, an attack on the theory (for which Pollock is the most famous) that ownership differs only from other possessory rights in the number of people against whom the possession is exigible; a defence of a theory that motive mistakes count in contract; a suggestion that gratuitous, donative promises might be enforceable based upon a requirement of deliberation, not formality or reliance; or arguments that servitudes should only be enforced to ensure that successive landowners did not need to remake them. There are many, many more. Occasionally, there is a diversion such as Ch.6 which uses the instance of water rights to illustrate his approach to rights which are annexed to the use of property and to advocate a return to a principle which focuses upon distribution of water rights based upon need, and only in the case of imperfect knowledge, first user. Although a diversion, the short chapter is the perfect philosophical companion for any student of water law to Getzler’s A History of Water Rights at Common Law (2006). No one interested in the history and theory of private law can sensibly avoid reading Foundations of Private Law. JAMES EDELMAN.*

* Keble College, Oxford.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

170

The Law Quarterly Review

[Vol. 124

Financial Markets and Exchanges Law. Edited by MICHAEL BLAIR Q.C. and GEORGE WALKER. [Oxford: Oxford University Press. 2007. liv + 615pp. Hardback: £160]. Financial Markets and Exchanges Law (FMEL) is the companion volume to Financial Services Law (FSL), also published by Oxford University Press under the editorship of Michael Blair and George Walker in 2006. As with FSL, the contributors include academics, lawyers in private practice and regulators. The editors identify the purpose of FMEL as the examination of the legal and regulatory measures that presently govern the structure of financial markets and exchanges. They contrast FMEL with FSL: whereas the focus of FSL was financial activities, financial institutions and financial sectors, FMEL is designed to consider more specifically the financial markets and exchanges which service these activities, institutions and sectors. The short title of the Financial Services and Markets Act 2000 (FSMA), the primary piece of governing legislation, covers financial services and financial markets: the division follows this lead. The editors stress that, by “market”, they mean any identifiable location, system or other set of formalised locations through which any commodity or product may be bought and sold, with a “financial market” being any organised process or procedure through which financial contracts or claims may be issued and traded. Stock exchanges are identified as one type of financial market as are other electronic, internet or digital systems, including alternative trading systems (ATSs). With these definitions in mind, FMEL follows its opening general chapter on financial markets, with chapters in Pt II on the general regime under the FSMA and Financial Services Authority (FSA) Rules governing “Recognised Investment Exchanges” (RIEs) and “Recognised Clearing Houses” (RCHs—the world of financial regulation is rich in acronyms, making it unsurprising that FMEL’s list of acronyms and abbreviations runs to seven pages) and follows with a chapter on ATSs. In Pt III, FMEL’s focus moves from the general examination of RIEs and RCHs to particular RIEs and RCHs with discussion of: the London Stock Exchange and Alternative Investment Market; virt-x and other electronic UK equity markets; UK financial derivatives and commodities markets; and settlement in the United Kingdom. Part IV, while remaining in the United Kingdom, shifts to a different type of market, the wholesale money market, i.e. those financial markets where lending and borrowing of money and securities takes place on a short-term basis, with a specific chapter on the legal and regulatory framework of the London money market. A chapter on UK payment systems follows. Parts V and VI, dealing respectively with European and international securities systems adhere to the editors’ aim to make European and international coverage broader (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

171

than that in FSL. Part V contains chapters on the regulation of European markets and exchanges, the Markets in Financial Instruments Directive 2004/39 (MiFID) and European ecommerce and emoney. Part VI contains chapters on the structure of international markets regulation, international settlement and Islamic securities exchanges. On the whole, FMEL is more likely to find a home with practitioners than with students, but the first chapter by George Walker, on financial markets and rxchanges, provides a very useful overview of the subject and should serve as an excellent point of reference for the student, new practitioner or, indeed, for the experienced practitioner seeking to confirm and supplement their understanding. The same can be said of the chapter by Michael Blair and Nigel Phipps on the regulation of European markets and exchanges. In addition to the usual challenge for any author seeking to set out and analyse rules and principles in a legal and regulatory environment which is constantly changing, anyone writing on financial markets regulation faces the challenge from regulators, exchanges and trade organisations who publish and distribute detailed and authoritative guidance and analysis free to industry participants and their advisers. This can make it difficult for those authors and publishers to justify the invitation to purchase their works, although it does not stop many, particularly the publishers of so-called “practitioner publications” (despite these often being a mere recital of rules, regulations and analysis already undertaken by the regulators, exchanges and organisations). The advantage of more general commentary, too often absent from “practitioner publications”, is that it creates an environment for the understanding of otherwise complex or arcane rules by sketching both the regulatory and the industry context. Too often, those writing practitioner texts take context for granted and assume knowledge and understanding on the part of their readers. Alternatively, they may feel that any comment on regulatory policy or the industry under regulation is beyond the province of a legal textbook or, worse still, the writer’s own area of expertise. Walker’s opening chapter avoids this. He provides an outline of the various types of markets and instrument traded on those markets—capital, money, Eurodollar, currency, financial derivatives, gold, commodity and insurance; market operation, structure, function, trading and orders; and specific UK exchanges. Similarly, Blair and Phipps in their chapter on European markets and exchanges, give an excellent overview of the process for bringing about the single European market in financial services and provide a useful critical analysis. For any UK financial services student or practitioner, an understanding of this process is vital not so much for its historical value but as an indicator of where the process is headed and the potential shape of the European and, by necessary extension, UK regulatory regime in years to come. (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

172

The Law Quarterly Review

[Vol. 124

The chapter on Islamic securities exchanges by Amr Marar and Wong Sau-Ngan also requires mention. It provides a useful overview of Islamic finance instruments and the manner in which Islamic finance institutions are regulated. The chapter shows a more general commitment to highlighting the growing importance of Islamic finance and the manner in which Islamic finance institutions are to be regulated. However, it does not deal with some of the more particular regulatory issues such as the question of whether Sukuk or Islamic bonds should be treated as collective investment schemes (see Henderson, “Testing the Water” (2007) International Financial Review 397). As the regulation of Islamic finance develops with the growth in the demand for Shari’a compliant securities, there will be a place for increased commentary on the regulation of Islamic finance and Islamic securities exchanges. As mentioned above, the editors make a point of justifying the differences between FSL and FMEL. FMEL should be viewed as supplementary to FSL and vice versa. If and when the editors come to prepare a second edition, they may give some thought to the structure of both. In particular, the chapter on European e-commerce and e-money belongs more properly in FSL than FMEL and the discussion of Islamic securities exchanges could be extended or supplemented in FSL by a discussion of the regulation of Islamic finance generally. For the time being, however, FMEL, together with FSL, remains a worthy addition to the growing number of works on financial regulation and will contribute to an area worthy of study in its own right, rather than as a mere subset of banking or corporate law. ANDREW HENDERSON.* Killing in Self-Defence. By FIONA LEVERICK. [Oxford: Oxford University Press. 2006. xxviii + 217 pp. Hardback: £53]. THIS book is a primarily theoretical work which seeks to explain why self-defence is an exculpatory defence in criminal law, and to explore the implications which such a theory might have. The author defends the word “killing” in the title on the ground that killings in self-defence must be the hardest of all to justify: lesser injuries, it seems, would also be justified, but more easily, by the same reasoning. So it seems that the book is primarily about self-defence rather than about homicide, and that might explain why relatively little is said (for example) about whether the person who kills following a drunken mistake should be guilty of murder, as in R. v Hatton [2005] EWCA Crim 2951; [2006] 1 Cr. App. R. 247. But to focus on homicide at all is rather distracting. Much of the theoretical emphasis in the work concerns the right to life. However, in * Solicitor, Dubai.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

173

explaining the commission of non-fatal offences in self-defence, a wider theory of rights of bodily integrity will be needed and no such account is offered in this book. More contentiously, when the author refers to “self-defence” in the title, she does not mean to discuss “defence of others or private defence more generally”. Instead, the focus is firmly on the (possibly selfish) decision on the part of one attacked person to prefer his or her own life or bodily integrity to that of the aggressor. In fact, the author considers the relationship between self-defence and duress and necessity in more detail. But she also makes an important assumption already, at the very beginning of the work, by saying that self-defence is the “archetypal justification”, so that if we cannot justify (killing in) self-defence, then we cannot justify anything. No doubt this seems to be so if one compares self-defence with duress, but what about (say) exercises in crime prevention by trained policemen? Perhaps they are the “archetypal justifications” instead? As it is, the author considers it unnecessary to explore the case for selfdefence (as she defines it) as an excuse. Perhaps she should have done, because the whole book depends upon the plausibility of her justificatory theory of self-defence. Leverick‘s thesis is that self-defence is justified by the right to life, which the defendant has but which the aggressor temporarily forfeits for as long as he provides a threat. The reliance on forfeiture, she says, is the only honest way in which we can explain why the defendant may save himself if his only option is to kill the aggressor. But the thesis does not persuade. In particular, Leverick does not clearly explain whether the life is temporarily forfeited to everyone, or just to the defendant; and she does not explain why the defender might not choose the most gratuitously torturous way of killing, if killing is indeed necessary. Chapter 7 is devoted to killing in defence of property, but is written on the assumption that there must be a principle of proportionality alongside one of necessity. Those who rely on forfeiture, however, need to explain and not to assume a principle of proportionality. In Ch.8 the author concludes that it is always proportionate to kill to avoid rape, even where the victim does not fear the possibility of death, because the rapist denies the humanity of the victim by using him or her as a sexual object. It is still not clear why (following a theory of forfeiture) the killing should need to be proportionate at all, but if it does, then the phrase “denies the victim‘s humanity” needs much more elaboration. It seems to explain why rape is a serious crime rather than why the rapist temporarily forfeits his right to life. Leverick’s attachment to forfeiture in order to justify self-defence stems at a deeper level from the assumption that exculpatory defences must be theorised as justifications or excuses, and not, at least sometimes, as a complex mixture of the two. But it is possible to rationalise selfdefence in some other way, e.g. as an expression of respect of the bodily (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

174

The Law Quarterly Review

[Vol. 124

integrity of others and of a commitment to prefer peaceful alternatives to disputes, without having to argue that self-defence has to be seen as either exclusively justificatory or excusatory. (See, for example, Ashworth, “Self-Defence and the Right to Life” [1975] Cambridge L.J. 282). Such a theory would also need a lot of fleshing out. Its application would depend upon examining the conduct of the various parties on a case-by-case basis (as one would expect when deciding whether actions show appropriate “respect” for communal values). It would not be conducive to hard-andfast rules—although that might be a strength rather than a weakness. It would, for example, avoid the conclusion that it is always proportionate to kill to avoid rape, regardless of the relationship between the parties, which is Leverick’s conclusion in Ch.8. It should be said that a theorist who would concentrate upon a commitment to peaceful alternatives to violence may be in line with the concerns of contemporary society. He will, for example, have much to say about the merits of the important principle in R. v Julien [1969] 1 W.L.R. 839 that the defendant must show an unwillingness to fight. But relatively little is said in Leverick‘s book of such notions as the duty owed by the state to protect its citizens, and the extent of the duty of the citizen to refrain from self-help. For the justification theorist, these considerations are typically sidelined, and discussed only at the stage of determining the necessity to act in individual cases. They are never brought to the forefront in explaining the rationale of the defence itself. We hear about forfeiture instead, because it is a “justificatory” theory, and it is perhaps too readily assumed that it is some such theory for which we are searching. Doctrinal problems also arise from writing through the lenses of justifications and excuses. For example, Leverick says that a person may use self-defence despite having provoked the confrontation in certain culpable ways which would forfeit a defence of duress, and the difference is explained on the ground that self-defence is classified as a justification and duress is an excuse. But, she says elsewhere, self-defence operates (again, categorically) as an excuse when the defendant was mistaken in anticipating a life-threatening attack, even if the mistake was a reasonable one. What, then, would Leverick say of the defendant who reasonably but mistakenly thinks that he is about to be stabbed, but had provoked the putative aggressor to some sort of aggressive display by having committed some minor assault upon him? Would he be ineligible for the defence altogether, on the ground that the mistaken actor is excused and not justified, and so he would then suffer from the harsher approach which applies to defendants who provoke the conditions of an excusatory defence? And would it have to follow that the same actor would not be a murderer but instead altogether justified if one thinks (more plausibly) that actions based on reasonable mistakes can be philosophically justified? These are the sorts of knots which pure justification and excuse theorists (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

175

(such as the author) tie around themselves, or (as in this case) fail to address altogether. Fletcher was the first theorist to realise the concerns which others might have of using justification and excuse theory as a tool of classification: see G. Fletcher, “The Right and the Reasonable” (1985) 98 Harvard L.R. 949. Fletcher, unlike Leverick, conceded that his “justificatory” theory of self-defence would deny any principle of proportionality. But he tried to defend that by invoking a general doctrine of abuse of rights, which would compensate for some of the extreme conclusions which would follow from his preferred justificatory model. In the meantime, Fletcher criticised Anglo-American law for preferring unified defences where issues of justification and excuse were “conflated”. It would seem more fruitful for writers on self-defence to decide whether to reject or to embrace Fletcher‘s preference for “structured” reasoning (where justifications are placed in a hierarchy above excuses) to “flat” reasoning (where the two are mixed) before deciding anything about the nature of self-defence. When one reads Leverick‘s account, one cannot but regret its selfimposed limitations. We need rather a treatise which reflects some of the modern scepticism upon the extent to which justification and excuse theory might have doctrinal consequences. We must also seek to explain selfdefence by comparison with the various associated forms of public and private defence. There may be different explanations for these apparently similar defences, and if so, then we will need to identify them in order to understand each of the different defences that bit better; and then at some point, an account of the relationship between the citizen and the state in the use of force will prove to be indispensable. JONATHAN ROGERS.* The Law of Assignment: The Creation and Transfer of Choses in Action. By MARCUS SMITH. [Oxford: Oxford University Press. 2007. lxxxv + 637pp. Hardback: £135]. THE commercial importance of the legal institution of assignment needs no elaboration. And, given the reality that wealth is increasingly represented not by traditional tangible realty and personalty but by intangible personalty (that is to say, choses in action), the need for the legal practitioner to get to grips with this area of the law is unlikely to diminish. Despite its obvious and growing significance, much of the law on assignment remains difficult to access, both on a practical as well as a theoretical level. Until recently, Marshall’s Assignments of Choses in Action (1950) stood alone in the secondary Commonwealth literature on the topic. This work was joined by a few others, such as Starke’s * University College, London.

(2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

176

The Law Quarterly Review

[Vol. 124

Assignments of Choses in Action in Australia (1972). However, it never rains but pours, and in the last two years, two major works on the subject have emerged. Tolhurst’s Assignment of Contractual Rights (2006) was published last year. The second, and the subject of this review, is Marcus Smith’s Law of Assignment: the Creation and Transfer of Choses in Action. Recognising that an area of law left fairly fallow for 50-odd years had now been the subject of a penetrating survey in Tolhurst’s work, Smith adopts a rather different approach which is more ambitious and, perhaps inevitably, more modest in its goals. First, unlike Tolhurst who focuses solely on the problem of assignments of contractual choses in action, Smith has attempted to deal with assignments of all choses in action. Indeed, he goes further than even Marshall as he also attempts to discuss how specific types of choses in action are assigned. It is therefore a work of considerable breadth and offers its readers something akin to the “onestop shop” typified, for example, by Snell’s Equity. It is (probably) all there, but therein lies the catch. Although Smith’s work serves as a useful starting point to prompt further research and thought into the topic, given the limits of size, it would be entirely unrealistic to expect the same degree of explication and analysis which a more narrowly-focused work might be able to bear. That said, Smith’s work is far from being a laundry list of judicial phenomena. Within the very wide landscape which Smith has set himself the task of exploring, a good deal of care has been taken to find some common principle to hold together the various modes of assignment, to explain the policy reasons which have led to the state of the rules today, as well as how the rules might be further refined and rationalised. Smith’s book is divided into five main parts. First, keeping true to its sub-title, Smith attempts to explain what choses in action are and what their nature is. The next two parts focus on the heart of the topic and deal with the transfer of choses in action and the consequences and effects of assignments. He then moves from the general to the particular, examining how assignments work in particular contexts, traversing in quick succession involuntary assignments; insurance contracts; leases of land; shares; intellectual property; and documentary intangibles and money. Lastly, Smith sets out a small section on assignments and the conflict of laws. So much done, but there may yet be more to do. Like Tolhurst and the others who have written in the field before him, Smith concentrates mainly on equitable assignments and spends comparatively little time on statutory assignments. The reader is directed to s.136 of the Law of Property Act 1925, and is told that its object was to “dispense with the necessity for joining the assignor as a party, but only in those cases where the assignor’s presence was unnecessary and therefore inconvenient” (at para.10.05). So in Smith’s view, s.136 does not affect the substantive law (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

177

of assignment, but merely provides improved machinery (at para.10.06). In support, Smith cites the cases of Re Westerton [1919] 2 Ch. 104 and Marchant v Morton, Down & Co [1901] 2 K.B. 829. But Channell J.’s reference to s.25(6) of the Judicature Act 1873 (the statutory predecessor to s.136) as being “merely machinery” (Marchant, at 832) is obiter dicta, and Sargant J.’s oft-repeated agreement with the proposition that s.25(6) of the Judicature Act 1873 was “only to alter procedure” did not just stop there (see Re Westerton, at 113). That agreement led Sargant J. to conclude that such change in procedure had substantive effects: that other requirements for relief in equity (such as consideration) might also be done away with so long as such question was regarded as being immaterial at law (ibid.). To suggest that s.136 is mere machinery might, therefore, be taking the provision and its effects too lightly. Although no chose in action may be assigned pursuant to s.136 if such chose had not been assignable in equity, Smith appears to assume that statutory assignment operates in a manner similar to its equitable predecessor. Section 136(1), on its face, purports to pass and transfer from the assignor to the assignee the legal right to, the legal and other remedies for and the power to give a good discharge for such debt or thing in action assigned thereunder. But it has never been the law that an equitable assignment effects any such transfer. This exposes the terminological problem that begins with the very title of the book. Because of concerns as to maintenance and champerty, and the original conception within the common law that obligations were personal to the obligor and obligee (outside of the real or mixed actions), the simple transfer of the benefits under a chose in action was barred at common law. This, and other parts of the common law’s history of distaste for the free transfer of obligations, is set out by Smith in Ch.5 of his book. That chapter also traces the manner in which the courts of equity overtook their more conservative common law brethren. First, the Courts of Chancery recognised that the common law’s general opposition to the free transfer of obligations had no application to equitable choses in action (Smith traces this development to the case of Warmstrey v Tanfield (1628–1629) 1 Chan. Rep. 29). This followed logically from the fact that equitable choses in action arose within equity’s exclusive jurisdiction. Secondly, as equity follows the law, equity may not effect a transfer of a common law chose in action in the same manner as it might an equitable one (and, in this regard, Smith is at one with Snell’s Equity in adhering to that maxim where the chose in action is not equitable). Smith explicitly acknowledges and accepts (at para.6.06) that equity cannot transfer what he terms to be the “legal title in a chose.” Just what such legal title is, is not made clear. But if this phrase is taken to mean, “the entitlement of the obligor to a common law chose to bring proceedings in a common law court”, it is rather odd, then, to persist in applying the language of “transfer” so (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

178

The Law Quarterly Review

[Vol. 124

pervasively in relation to assignments (whether statutory or equitable) of all forms of chose in action, be they common law or equitable in nature. Indeed, the inaptness of the equation of “transfer” and “equitable assignment” is reinforced by Smith’s reduction of equitable assignment of a legal chose in action into two primary types: equitable assignment by means of an express trust constituted over the legal chose in action to be assigned, and equitable assignment by means of the imposition of a constructive trust over that chose in action (what Marshall had termed in his book, “informal equitable assignments”). The express constitution of a trust over a legal chose in action plainly entails no disposition of any equitable interest in the chose. But the same is true of the latter. Following Westdeutsche Landesbank Girozentrale v Islington LBC [1996] A.C. 669, an equitable assignment of a legal chose in action entails not a disposition of an equitable interest, but the creation of a new one. This is a point which, again, Smith accepts (at paras 7.64–7.65), noting that this is precisely why the requirements of s.53(1)(c) of the Law of Property Act 1925 are inapplicable to such cases. Yet at para.6.10, Smith describes equitable assignments of a legal chose in action as having the effect of causing the beneficial interest in the chose to pass from the assignor to the assignee. It is only by pressing on to para.6.12 (in particular, the text to and in fn.25) that it is revealed that Smith’s use of the term “transfer” is not straightforward. The attempt to use the layman’s term of “transfer” so broadly is well intentioned but, perhaps, unhelpful. For what occurs in the case of an equitable assignment of an equitable chose in action is certainly not what is happening in the case of an equitable assignment of a common law chose in action, and neither case is truly on all fours with the effect of a statutory assignment of a chose in action (whether common law or equitable) pursuant to s.136 of the Law of Property Act 1925. The New Shorter Oxford English Dictionary (1993) traces the etymology of the verb “to transfer” from the French “transf´erer” and the Latin “transferre”, both being derived from the juxtaposition of the Latin preposition “trans” (meaning “across, over”) with “ferre” (meaning “bear, carry”). And if we keep that etymology in mind, application of that verb beyond the realm of tangible personalty ought to be done with care. Tangible property may be borne over from hand to hand, but realty and intangible personalty cannot. And though, in relation to the latter, one may readily conceive of devices that the courts (common law and/or equitable) may recognise as having an effect equivalent to transfer, there can be no transfer per se. Such transfer may only be figurative and cannot be literal. In the interest of enhancing the precision of analysis, given the elusive nature of assignments of choses in action, it may be helpful to remember just how limited the core meaning of a “transfer” really is, whatever the common usage of that word might be. (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

JANUARY 2008]

Reviews and Notices

179

One area where the decision to couch the analysis of assignment in terms of transfer may have hindered rather than aided a reader’s understanding of the complex issues underlying assignment may be found in Smith’s treatment of the conflict of laws or private international law issues commonly raised by assignments of choses in action. In the final part of his work, Smith takes up the difficult task of providing some suggestions about private international law implications arising from his analysis. The conflicts issues discussed are, in the main, Euro-centric. Most of the discussion is conducted in light of the Brussels Regulation and the Rome Convention. Where the Brussels Regulation on jurisdiction does not apply, Smith directs the reader to refer to the mainstream textbooks, and spends only seven pages at the end of the book discussing the position as to choice of law where the Rome Convention is inapplicable. This brevity is, perhaps, inevitable given that the conflicts issues are far from being the book’s core business. Nevertheless, there may be a point or two within this part that bears re-consideration. At p.598, Smith suggests that statutory assignments fall within the ambit of Art.12(1) of the Rome Convention. This is despite his recognition (at paras 10.05–10.06) that they operate as a mechanism allowing the legal estate in a chose in action to pass. If that is true, logic would impel us to take the view that statutory assignments (whether requiring registration or otherwise) are entirely distinct from equitable assignments, for equitable assignments never had and still do not have that effect. Even if an autonomous construction is placed upon the term “assignment” within the context of the Convention, it is by no means clear that such autonomous construction is sufficiently broad to encompass equitable as well as statutory assignments, particularly of legal choses in action. Certainly, the editors of Dicey, Morris & Collins on the Conflict of Laws do not deal with the question. As pointed out above, whether effected by means of a contract for value or a voluntary declaration, an equitable assignment of a legal chose takes effect insofar as the courts of equity would recognise an equitable interest in the assignee following the constitution of the contract or the declaration of the assignment. There is no “transfer” of any equitable estate, and certainly, none of the legal estate. If the argument is that statutory assignments operate similarly to equitable assignments, and ought therefore to fall within the ambit of Art.12(1) as well (leaving aside the instances where the particular sort of statutory assignment requires registration—as with the assignment of shares in an English-incorporated company), that similarity is hard to fathom, even in relation to statutory assignment regimes which do not require any form of registration. Given the unfathomably diffuse nature of its subject-matter and the lack of any contemporary precedent, it is a tremendous achievement that Smith’s book exists at all. Smith has done the legal community a huge (2008) 124 L.Q.R., JANUARY  SWEET & MAXWELL

AND

CONTRIBUTORS

180

The Law Quarterly Review

[Vol. 124

service by bringing together much of the relevant primary and secondary material which make up the large, amorphous and growing body of law on assignment and mechanisms having an equivalent effect to transfer in one fairly compact and succinctly written volume. If it inspires others to reflect on the materials and reach conclusions of their own, so much the better for all. CHEE HO THAM.*

* Singapore Management University School of Law.

(2008) 124 L.Q.R., JANUARY ď&#x203A;&#x2122; SWEET & MAXWELL

AND

CONTRIBUTORS